Iec - Ieee 62209-1528-2020

IEC/IEEE 62209-1528
®
Edition 1.0 2020-10
INTERNATIONAL
STANDARD
colour
inside
Measurement procedure for the assessment of specific absorption rate of

human exposure to radio frequency fields from hand-held and body-mounted
wireless communication devices –
Part 1528: Human models, instrumentation, and procedures
(Frequency range of 4 MHz to 10 GHz)
IEC/IEEE 62209-1528:2020-10(en)
THIS PUBLICATION IS COPYRIGHT PROTECTED
Copyright © 2020 IEC, Geneva, Switzerland
Copyright © 2020 IEEE
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form
or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing being
secured. Requests for permission to reproduce should be addressed to either IEC at the address below or IEC’s
member National Committee in the country of the requester or from IEEE.
IEC Central Office Institute of Electrical and Electronics Engineers, Inc.

3, rue de Varembé 3 Park Avenue
CH-1211 Geneva 20 New York, NY 10016-5997
Switzerland United States of America
Tel.: +41 22 919 02 11 [email protected]
[email protected] www.ieee.org
www.iec.ch
About the IEC

The International Electrotechnical Commission (IEC) is the leading global organization that prepares and publishes
International Standards for all electrical, electronic and related technologies.
About the IEEE

IEEE is the world’s largest professional association dedicated to advancing technological innovation and excellence for
the benefit of humanity. IEEE and its members inspire a global community through its highly cited publications,
conferences, technology standards, and professional and educational activities.
About IEC/IEEE publications
The technical content of IEC/IEEE publications is kept under constant review by the IEC and IEEE. Please make sure
that you have the latest edition, a corrigendum or an amendment might have been published.
IEC publications search - webstore.iec.ch/advsearchform Electropedia - www.electropedia.org

The advanced search enables to find IEC publications by a The world's leading online dictionary on electrotechnology,
variety of criteria (reference number, text, technical containing more than 22 000 terminological entries in English
committee,…). It also gives information on projects, replaced and French, with equivalent terms in 16 additional languages.
and withdrawn publications. Also known as the International Electrotechnical Vocabulary
(IEV) online.
IEC Just Published - webstore.iec.ch/justpublished
Stay up to date on all new IEC publications. Just Published IEC Glossary - std.iec.ch/glossary
details all new publications released. Available online and 67 000 electrotechnical terminology entries in English and
once a month by email. French extracted from the Terms and definitions clause of
IEC publications issued between 2002 and 2015. Some
IEC Customer Service Centre - webstore.iec.ch/csc entries have been collected from earlier publications of IEC
If you wish to give us your feedback on this publication or TC 37, 77, 86 and CISPR.
need further assistance, please contact the Customer Service
Centre: [email protected].
IEC/IEEE 62209-1528
®
Edition 1.0 2020-10
INTERNATIONAL
STANDARD
colour
inside
Measurement procedure for the assessment of specific absorption rate of

human exposure to radio frequency fields from hand-held and body-mounted
wireless communication devices –
INTERNATIONAL
ELECTROTECHNICAL
COMMISSION
ICS 17.220.20 ,(&ISBN 978-2-8322-8533-6

,(((,6%167'
Warning! Make sure that you obtained this publication from an authorized distributor.
® Registered trademark of the International Electrotechnical Commission

–2– IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
CONTENTS
FOREWORD ......................................................................................................................... 14
INTRODUCTION ................................................................................................................... 17
1 Scope ............................................................................................................................ 18
2 Normative references .................................................................................................... 18
3 Terms and definitions .................................................................................................... 18
4 Symbols and abbreviated terms ..................................................................................... 26
4.1 Physical quantities ................................................................................................ 26
4.2 Constants ............................................................................................................. 26
4.3 Abbreviated terms ................................................................................................. 27
5 Quick start guide and evaluation plan checklist .............................................................. 28
6 Measurement system specifications ............................................................................... 30
6.1 General requirements for full SAR testing ............................................................. 30
6.2 Phantom specifications ......................................................................................... 31
6.2.1 General ......................................................................................................... 31
6.2.2 Basic phantom parameters ............................................................................ 31
6.2.3 Head phantom ............................................................................................... 33
6.2.4 Flat phantom ................................................................................................. 34
6.2.5 Device-specific phantoms .............................................................................. 35
6.3 Influence of hand on SAR in head ......................................................................... 35
6.4 Scanning system requirements ............................................................................. 36
6.5 Device holder specifications .................................................................................. 36
6.6 Characteristics of the readout electronics ............................................................. 37
7 Protocol for SAR assessment ........................................................................................ 37
7.1 General ................................................................................................................. 37
7.2 Measurement preparation ..................................................................................... 37
7.2.1 Preparation of tissue-equivalent medium and system check........................... 37
7.2.2 Preparation of the wireless communication DUT ............................................ 38
7.2.3 DUT operating mode requirements ................................................................ 38
7.2.4 Positioning of the DUT relative to the phantom .............................................. 40
7.2.5 Antenna configurations .................................................................................. 57
7.2.6 Options and accessories ................................................................................ 57
7.2.7 DUTs with alternative form factor ................................................................... 57
7.2.8 Test frequencies for DUTs ............................................................................. 58
7.3 Tests to be performed for DUTs ............................................................................ 58
7.3.1 General ......................................................................................................... 58
7.3.2 Basic approach for DUT testing ..................................................................... 59
7.4 Measurement procedure ....................................................................................... 60
7.4.1 General ......................................................................................................... 60
7.4.2 Full SAR testing procedure ............................................................................ 60
7.4.3 Drift ............................................................................................................... 64
7.4.4 SAR measurements of DUTs with multiple antennas or multiple
transmitters ................................................................................................... 66
7.5 Post-processing of SAR measurement data .......................................................... 72
7.5.1 Interpolation .................................................................................................. 72
7.5.2 Extrapolation ................................................................................................. 72
7.5.3 Definition of the averaging volume ................................................................. 72
IEC/IEEE 62209-1528:2020 –3–
© IEC/IEEE 2020
7.5.4 Searching for the maxima .............................................................................. 73
7.6 Time-period averaged SAR considerations............................................................ 73
7.6.1 General ......................................................................................................... 73
7.6.2 RF conducted power ...................................................................................... 73
7.6.3 Time-period averaged SAR measurement settings for SAR
measurement methods .................................................................................. 73
7.6.4 Exposure condition and test position considerations ...................................... 74
7.6.5 Time-period averaged SAR for simultaneous transmission ............................. 74
7.6.6 TX factor assessment .................................................................................... 74
7.6.7 SAR measurements ....................................................................................... 75
7.6.8 Uncertainty in TPAS evaluations .................................................................... 75
7.7 Proximity sensors considerations .......................................................................... 76
7.7.1 General ......................................................................................................... 76
7.7.2 Procedures for determining proximity sensor triggering distances .................. 77
7.7.3 Procedure for determining proximity sensor coverage area ............................ 80
7.7.4 SAR measurement procedure involving proximity sensors ............................. 81
7.8 SAR correction for deviations of complex permittivity from targets ........................ 81
7.8.1 General ......................................................................................................... 81
7.8.2 SAR correction formula .................................................................................. 82
7.8.3 Uncertainty of the correction formula ............................................................. 83
7.9 Minimization of testing time ................................................................................... 83
7.9.1 General ......................................................................................................... 83
7.9.2 Fast SAR testing............................................................................................ 84
7.9.3 SAR test reductions ....................................................................................... 89
8 Measurement uncertainty estimation............................................................................ 100
8.1 General ............................................................................................................... 100
8.2 Requirements on the uncertainty evaluation ........................................................ 101
8.3 Description of uncertainty models ....................................................................... 102
8.3.1 General ....................................................................................................... 102
8.3.2 SAR measurement of a DUT ........................................................................ 102
8.3.3 System validation and system check measurement ...................................... 102
8.3.4 System check repeatability and reproducibility ............................................. 102
8.3.5 Fast SAR testing (relative measurement) ..................................................... 102
8.4 Parameters contributing to uncertainty ................................................................ 104
8.4.1 Measurement system errors......................................................................... 104
8.4.2 Phantom and device (DUT or validation antenna) errors .............................. 105
8.4.3 Corrections to the SAR result (if applied) ..................................................... 107
9 Measurement report .................................................................................................... 108
9.1 General ............................................................................................................... 108
9.2 Items to be recorded in the measurement report ................................................. 108
Annex A (normative) SAR measurement system verification .............................................. 112
A.1 Overview............................................................................................................. 112
A.2 System check ..................................................................................................... 112
A.2.1 Purpose ....................................................................................................... 112
A.2.2 Phantom set-up ........................................................................................... 113
A.2.3 System check antenna ................................................................................. 113
A.2.4 System check antenna input power measurement ........................................ 114
A.2.5 System check procedure .............................................................................. 115
A.2.6 System check acceptance criteria ................................................................ 116
–4– IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
A.3 System validation ................................................................................................ 116
A.3.1 Purpose ....................................................................................................... 116
A.3.2 Phantom set-up ........................................................................................... 116
A.3.3 System validation antennas ......................................................................... 116
A.3.4 Input power measurement ........................................................................... 117
A.3.5 System validation procedure ........................................................................ 117
A.4 Fast SAR testing system validation and system check ........................................ 119
A.4.1 General ....................................................................................................... 119
A.4.2 Fast SAR testing system validation .............................................................. 119
A.4.3 Fast SAR testing system check .................................................................... 120
Annex B (informative) SAR test reduction supporting information ....................................... 122
B.1 General ............................................................................................................... 122
B.2 Test reduction based on characteristics of DUT design ....................................... 122
B.2.1 General ....................................................................................................... 122
B.2.2 Statistical analysis overview ........................................................................ 122
B.2.3 Analysis results ........................................................................................... 123
B.2.4 Conclusions ................................................................................................. 126
B.2.5 Expansion to multi-transmission antennas ................................................... 126
B.3 Test reduction based on analysis of SAR results on other signal modulations ..... 126
B.3.1 General ....................................................................................................... 126
B.3.2 Analysis results ........................................................................................... 127
B.4 Test reduction based on SAR level analysis ........................................................ 128
B.4.1 General ....................................................................................................... 128
B.4.2 Statistical analysis ....................................................................................... 129
B.4.3 Test reduction applicability example ............................................................ 132
B.5 Other statistical approaches to search for the high SAR test configurations ........ 134
B.5.1 General ....................................................................................................... 134
B.5.2 Test reductions based on a DOE ................................................................. 134
B.5.3 One factor at a time (OFAT) search ............................................................. 134
B.5.4 Analysis of unstructured data ....................................................................... 134
Annex C (informative) Measurement uncertainty of results obtained from specific fast
SAR testing methods .......................................................................................................... 135
C.1 General ............................................................................................................... 135
C.2 Measurement uncertainty evaluation – contributing parameters .......................... 135
C.2.1 General ....................................................................................................... 135
C.2.2 Probe calibration and system calibration drift ............................................... 136
C.2.3 Isotropy ....................................................................................................... 136
C.2.4 Probe positioning ......................................................................................... 137
C.2.5 Mutual sensor coupling ................................................................................ 138
C.2.6 Scattering within the probe array ................................................................. 139
C.2.7 Sampling error ............................................................................................. 139
C.2.8 Array boundaries ......................................................................................... 139
C.2.9 Probe or probe array coupling with the DUT ................................................ 139
C.2.10 Measurement system immunity / secondary reception ................................. 139
C.2.11 Deviations in phantom shape ....................................................................... 140
C.2.12 Spatial variation in dielectric properties ....................................................... 140
C.2.13 Reconstruction ............................................................................................ 140
C.3 Uncertainty budget .............................................................................................. 140
Annex D (normative) SAR system validation antennas ....................................................... 143
IEC/IEEE 62209-1528:2020 –5–
© IEC/IEEE 2020
D.1 General antenna requirements ............................................................................ 143
D.2 Standard dipole antenna ..................................................................................... 143
D.2.1 Mechanical description ................................................................................ 143
D.2.2 Numerical target SAR values ....................................................................... 146
D.3 Standard waveguide ........................................................................................... 148
D.4 System validation antennas for below 150 MHz .................................................. 149
D.4.1 General ....................................................................................................... 149
D.4.2 Confined loop antenna ................................................................................. 150
D.4.3 Meander dipole antenna .............................................................................. 152
D.5 Orthogonal E-field source – VPIFA ..................................................................... 153
Annex E (normative) Calibration and characterization of dosimetric (SAR) probes ............. 157
E.1 Introductory remarks ........................................................................................... 157
E.2 Linearity .............................................................................................................. 158
E.3 Assessment of the sensitivity of the dipole sensors ............................................. 158
E.3.1 General ....................................................................................................... 158
E.3.2 Two-step calibration procedures .................................................................. 158
E.3.3 One-step calibration procedure – reference antenna method ....................... 164
E.3.4 One-step calibration procedure – coaxial calorimeter method ...................... 168
E.4 Isotropy .............................................................................................................. 170
E.4.1 Axial isotropy ............................................................................................... 170
E.4.2 Hemispherical isotropy ................................................................................ 170
E.5 Lower detection limit ........................................................................................... 175
E.6 Boundary effect................................................................................................... 176
E.7 Response time .................................................................................................... 176
Annex F (informative) Example recipes for phantom tissue-equivalent media ..................... 177
F.1 General ............................................................................................................... 177
F.2 Ingredients .......................................................................................................... 177
F.3 Tissue-equivalent medium liquid formulas (permittivity/conductivity) ................... 178
Annex G (normative) Phantom specifications ..................................................................... 180
G.1 Rationale for the phantom characteristics ........................................................... 180
G.1.1 General ....................................................................................................... 180
G.1.2 Rationale for the SAM phantom ................................................................... 180
G.1.3 Rationale for the flat phantom ...................................................................... 180
G.2 SAM phantom specifications ............................................................................... 181
G.2.1 General SAM phantom specifications........................................................... 181
G.2.2 SAM phantom shell specification ................................................................. 185
G.3 Flat phantom specifications ................................................................................. 187
G.4 Justification of flat phantom dimensions .............................................................. 188
G.5 Rationale for tissue-equivalent media ................................................................. 191
G.6 Definition of a phantom coordinate system and a DUT coordinate system ........... 193
Annex H (informative) Measurement of the dielectric properties of tissue-equivalent
media and uncertainty estimation ........................................................................................ 195
H.1 Overview............................................................................................................. 195
H.2 Measurement techniques .................................................................................... 195
H.2.1 General ....................................................................................................... 195
–6– IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
H.2.2 Instrumentation ............................................................................................ 195
H.2.3 General principles ....................................................................................... 195
H.3 Slotted coaxial transmission line ......................................................................... 196
H.3.1 General ....................................................................................................... 196
H.3.2 Equipment set-up ........................................................................................ 196
H.3.3 Measurement procedure .............................................................................. 197
H.4 Contact coaxial probe ......................................................................................... 197
H.4.1 General ....................................................................................................... 197
H.5 TEM transmission line ......................................................................................... 199
H.5.1 General ....................................................................................................... 199
H.6 Dielectric properties of reference liquids ............................................................. 201
Annex I (informative) Studies for potential hand effects on head SAR ................................ 204
I.1 Overview............................................................................................................. 204
I.2 Background......................................................................................................... 204
I.2.1 General ....................................................................................................... 204
I.2.2 Hand phantoms ........................................................................................... 205
I.3 Summary of experimental studies ....................................................................... 205
I.3.1 Experimental studies using fully compliant SAR measurement systems ....... 205
I.3.2 Experimental studies using other SAR measurement systems ..................... 205
I.4 Summary of computational studies ...................................................................... 206
I.5 Conclusions ........................................................................................................ 206
Annex J (informative) Skin enhancement factor ................................................................. 207
J.1 Background......................................................................................................... 207
J.2 Rationale ............................................................................................................ 208
J.3 Simulations ......................................................................................................... 208
J.4 Recommendation ................................................................................................ 209
Annex K (normative) Application-specific phantoms ........................................................... 211
K.1 General ............................................................................................................... 211
K.2 Phantom basic requirements ............................................................................... 211
K.3 Examples of specific alternative phantoms .......................................................... 211
K.3.1 Face-down SAM phantom ............................................................................ 211
K.3.2 Head-stand SAM phantom ........................................................................... 212
K.3.3 Wrist phantom ............................................................................................. 212
K.4 Scanning and evaluation requirements ................................................................ 213
K.5 Uncertainty assessment ...................................................................................... 213
K.6 Reporting ............................................................................................................ 213
Annex L (normative) Fast compliance evaluations using a flat-bottom phantom with a
curved corner (Uniphantom) ................................................................................................ 214
L.1 General ............................................................................................................... 214
L.2 Uniphantom ........................................................................................................ 214
L.3 Device positions for compliance testing and definitions of handset shapes ......... 214
L.3.1 General ....................................................................................................... 214
L.3.2 Handsets with a straight form factor ............................................................. 215
L.3.3 Handsets with a clamshell form factor .......................................................... 215
L.4 Testing procedure ............................................................................................... 215
IEC/IEEE 62209-1528:2020 –7–
© IEC/IEEE 2020
L.4.1 General ....................................................................................................... 215
L.4.2 Handsets with straight form factors .............................................................. 215
L.4.3 Handsets with clamshell form factors ........................................................... 216
L.5 Uncertainty of SAR measurement results using Uniphantom ............................... 217
Annex M (informative) Wired hands-free headset testing ................................................... 218
M.1 Concept .............................................................................................................. 218
M.2 Example results .................................................................................................. 219
M.3 Discussion .......................................................................................................... 220
Annex N (informative) Applying the head SAR test procedures .......................................... 221
Annex O (normative) Uncertainty analysis for measurement system manufacturers and
calibration laboratories........................................................................................................ 224
O.1 Probe linearity and detection limits ..................................................................... 224
O.2 Broadband signal uncertainty .............................................................................. 225
O.3 Boundary effect................................................................................................... 225
O.4 Field-probe readout electronics uncertainty ......................................................... 226
O.5 Signal step-response time uncertainty ................................................................. 226
O.6 Probe integration-time uncertainty ...................................................................... 227
O.6.1 General ....................................................................................................... 227
O.6.2 Probe integration-time uncertainty for periodic pulsed signals ...................... 227
O.6.3 Probe integration-time uncertainty for non-periodic signals .......................... 228
O.7 Contribution of mechanical constraints ................................................................ 228
O.7.1 Mechanical tolerances of the probe positioner (directions parallel to
phantom surface) ......................................................................................... 228
O.7.2 Probe positioning with respect to phantom shell surface .............................. 228
O.7.3 First-order approximation of exponential decay ............................................ 229
O.8 Contribution of post-processing ........................................................................... 229
O.8.1 General ....................................................................................................... 229
O.8.2 Evaluation test functions .............................................................................. 230
O.8.3 Data-processing algorithm uncertainty evaluations ...................................... 232
O.9 Tissue-equivalent medium properties uncertainty ................................................ 235
O.9.1 General ....................................................................................................... 235
O.9.2 Medium density ........................................................................................... 235
O.9.3 Medium conductivity uncertainty .................................................................. 235
O.9.4 Medium permittivity uncertainty ................................................................... 235
O.9.5 Assessment of dielectric properties measurement uncertainties ................... 235
O.9.6 Medium temperature uncertainty .................................................................. 237
Annex P (normative) Post-processing techniques .............................................................. 239
P.1 Extrapolation and interpolation schemes ............................................................. 239
P.1.1 General ....................................................................................................... 239
P.1.2 Extrapolation schemes................................................................................. 239
P.1.3 Interpolation schemes .................................................................................. 239
P.2 Averaging scheme and maximum finding ............................................................ 239
P.2.1 Volume average schemes ............................................................................ 239
P.2.2 Finding the psSAR and estimating the uncertainty ....................................... 240
Annex Q (informative) Rationale for time-period averaged SAR test procedure .................. 241
Annex R (normative) Measurement uncertainty analysis for testing laboratories ................ 242
R.1 RF ambient conditions ........................................................................................ 242
R.2 Device positioning and holder uncertainties ........................................................ 242
R.2.1 General ....................................................................................................... 242
–8– IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
R.2.2 Device holder perturbation uncertainty ......................................................... 243
R.2.3 DUT positioning uncertainty with a specific test device holder: Type A ........ 244
R.3 Probe modulation response ................................................................................ 244
R.4 Time-period averaged SAR ................................................................................. 245
R.4.1 General ....................................................................................................... 245
R.4.2 TX factor uncertainty ................................................................................... 245
R.5 Measured SAR drift ............................................................................................. 246
R.5.1 General ....................................................................................................... 246
R.5.2 Accounting for drift ...................................................................................... 246
R.6 SAR scaling uncertainty ...................................................................................... 247
Annex S (normative) Validation antenna SAR measurement uncertainty ............................ 248
S.1 Deviation of experimental antennas .................................................................... 248
S.2 Other uncertainty contributions when using system validation antennas .............. 248
Annex T (normative) Interlaboratory comparisons .............................................................. 250
T.1 Purpose .............................................................................................................. 250
T.2 Phantom set-up .................................................................................................. 250
T.3 Reference devices .............................................................................................. 250
T.4 Power set-up ...................................................................................................... 250
T.5 Interlaboratory comparison – procedure .............................................................. 251
Annex U (informative) Determination of the margin for compliance evaluation using the
Uniphantom ........................................................................................................................ 252
U.1 General ............................................................................................................... 252
U.2 Deviation of the psSAR measured using the Uniphantom from the psSAR
measured using the SAM phantom ...................................................................... 252
U.3 Determination of margin based on 95 % confidence interval................................ 253
U.4 Examples of the determination of the margin factor ............................................. 253
U.4.1 Margin for handsets with straight form factors at flat-bottom position ........... 253
U.4.2 Margin for handsets with straight form factors (except smart phones at
flat-bottom position) ..................................................................................... 255
U.4.3 Margin for smart phones at flat-bottom position ........................................... 257
U.4.4 Margin for smart phones at corner position .................................................. 259
U.4.5 Margin for handsets with clamshell form factors at corner position ............... 261
Annex V (informative) Automatic input power level control for system validation ................ 264
V.1 General ............................................................................................................... 264
V.2 Operational mechanism of AIPLC ....................................................................... 264
Annex W (informative) LTE test configurations supporting information ............................... 266
W.1 General ............................................................................................................... 266
W.2 Study 1 ............................................................................................................... 266
W.3 Study 2 ............................................................................................................... 268
W.4 Justifications of relative standard deviations ....................................................... 269
Bibliography ........................................................................................................................ 271
Figure 1 – Quick start guide .................................................................................................. 29

Figure 2 – Dimensions of the elliptical phantom .................................................................... 35
Figure 3 – Mounting of the DUT in the device holder using low-permittivity and low-loss
foam to avoid changes of DUT performance by the holder material ....................................... 37
Figure 4 – Designation of DUT reference points .................................................................... 41
IEC/IEEE 62209-1528:2020 –9–
© IEC/IEEE 2020
Figure 5 – Measurements performed by shifting a large device over the efficient
measurement area of the system including overlapping areas – in this case: six tests
performed ............................................................................................................................. 42
Figure 6 – Test positions for body-worn devices ................................................................... 43
Figure 7 – Device with swivel antenna .................................................................................. 44
Figure 8 – Test positions for body supported devices ............................................................ 45
Figure 9 – Test positions for desktop devices ....................................................................... 47
Figure 10 – Test positions for front-of-face devices ............................................................... 48
Figure 11 – Test position for hand-held devices, not used at the head or torso ..................... 49
Figure 12 – Test position for limb-worn devices .................................................................... 49
Figure 13 – Test position for clothing-integrated wireless communication devices ................. 50
Figure 14 – Possible test positions for a generic device ........................................................ 51
Figure 15 – Vertical and horizontal reference lines and reference points A and B on
two example device types: a full touch-screen smart phone (left) and a DUT with a
keypad (right) ....................................................................................................................... 53
Figure 16 – Cheek position of the DUT on the left side of SAM where the device
position shall be maintained for the phantom test set-up ....................................................... 56
Figure 17 – Tilt position of the DUT on the left side of SAM .................................................. 56
Figure 18 – An alternative form factor DUT with reference points and reference lines ........... 57
Figure 19 – Block diagram of the tests to be performed ........................................................ 60
Figure 20 – Orientation of the probe with respect to the line normal to the phantom
surface, for head and flat phantoms, shown at two different locations ................................... 64
Figure 21 – Measurement procedure for different types of correlated signals ........................ 72
Figure 22 – Positioning of the surfaces and edges of the DUT for determining the
proximity sensor triggering distance ...................................................................................... 79
Figure 23 – Positioning of the edges of the DUT to determine proximity sensor
triggering distance variations with the edge positioned at different angles from the
perpendicular position ........................................................................................................... 80
Figure 24 – Fast SAR Procedure A ....................................................................................... 87
Figure 25 – Fast SAR Procedure B ....................................................................................... 89
Figure 26 – Modified chart of Figure 19 ................................................................................. 93
Figure 27 – Use of conducted power for LTE mode selection, for Band 1 (1 920 MHz to
1 980 MHz) (MPR values are in dB) ...................................................................................... 97
Figure 28 – Use of conducted power for LTE mode selection, for Band 17 (704 MHz to
716 MHz) (MPR values are in dB) ......................................................................................... 98
Figure A.1 – Test set-up for the system check .................................................................... 114
Figure B.1 – Distribution of Tilt/Cheek ................................................................................. 124
Figure B.2 – SAR relative to SAR in position with maximum SAR in GSM mode .................. 128
Figure B.3 – Two points identifying the minimum distance between the position of the
interpolated maximum SAR and the points at 0,6 × SAR max ................................................ 130
Figure B.4 – Histogram for D min in the case of GSM 900 and iso-level at 0,6 × SAR max ..... 130
Figure B.5 – Histogram for random variable Factor 1g,1800 ................................................. 132
Figure D.1 – Mechanical details of the standard dipoles ...................................................... 145
Figure D.2 – Standard waveguide (dimensions are according to Table D.3) ........................ 148
Figure D.3 – Drawing of the CLA that corresponds to a resonant loop integrated in a
metallic structure to isolate the resonant structure from the environment ............................ 150
Figure D.4 – Mechanical details of the meander dipoles for 150 MHz .................................. 152
– 10 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure D.5 – VPIFA validation antenna ............................................................................... 155
Figure D.6 – Mask for positioning VPIFAs ........................................................................... 155
Figure E.1 – Experimental set-up for assessment of the sensitivity (conversion factor)
using a vertically-oriented rectangular waveguide ............................................................... 162
Figure E.2 – Illustration of the antenna gain evaluation set-up ............................................ 165
Figure E.3 – Schematic of the coaxial calorimeter system ................................................... 169
Figure E.4 – Set-up to assess hemispherical isotropy deviation in tissue-equivalent
medium ............................................................................................................................... 171
Figure E.5 – Alternative set-up to assess hemispherical isotropy deviation in tissue-
equivalent medium .............................................................................................................. 172
Figure E.6 – Experimental set-up for the hemispherical isotropy assessment ...................... 173
Figure E.7 – Conventions for dipole position (ξ) and polarization (θ) ................................... 174
Figure E.8 – Measurement of hemispherical isotropy with reference antenna ...................... 175
Figure G.1 – Illustration of dimensions in Table G.1 and Table G.2 ..................................... 182
Figure G.2 – Close up side view of phantom showing the ear region ................................... 184
Figure G.3 – Side view of the phantom showing relevant markings ..................................... 185
Figure G.4 – Sagittally bisected phantom with extended perimeter (shown placed on its
side as used for device SAR tests) ..................................................................................... 186
Figure G.5 – Picture of the phantom showing the central strip ............................................. 186
Figure G.6 – Cross-sectional view of SAM at the reference plane ....................................... 187
Figure G.7 – Dimensions of the flat phantom set-up used for deriving the minimal
phantom dimensions for W and L for a given phantom depth D ........................................... 189
Figure G.8 – FDTD predicted error in the 10 g psSAR as a function of the dimensions
of the flat phantom compared with an infinite flat phantom at 800 MHz ............................... 190
Figure G.9 – Complex permittivity of human tissues compared to the phantom target
properties ........................................................................................................................... 193
Figure G.10 – Example reference coordinate system for the left-ear ERP of the SAM
phantom ............................................................................................................................. 194
Figure G.11 – Example coordinate system on a DUT .......................................................... 194
Figure H.1 – Slotted line set-up........................................................................................... 196
Figure H.2 – An open-ended coaxial probe with inner and outer radii a and b,
respectively ........................................................................................................................ 198
Figure H.3 – TEM line dielectric properties test set-up [85] ................................................. 200
Figure J.1 – SAR and temperature increase (ΔT) distributions simulated for a three-
layer (skin, fat, muscle) planar torso model ......................................................................... 207
Figure J.2 – Statistical approach to protect 90 % of the population ..................................... 209
Figure J.3 – psSAR skin enhancement factors .................................................................... 210
Figure K.1 – SAM face-down phantom ................................................................................ 212
Figure K.2 – SAM head-stand phantom ............................................................................... 212
Figure K.3 – Wrist phantom ................................................................................................ 213
Figure L.1 – Cross section of the unified phantom (Uniphantom) with its dimensions .......... 214
Figure L.2 – Measurement positions of handsets with straight and clamshell form
factors ................................................................................................................................ 215
Figure L.3 – Flow chart of testing procedure for handsets with straight form factors ............ 216
Figure L.4 – Flow chart of testing procedure for handsets with clamshell form factors ......... 217
Figure M.1 – Configuration of a personal wired hands-free headset .................................... 218
IEC/IEEE 62209-1528:2020 – 11 –
© IEC/IEEE 2020
Figure M.2 – Configuration without a personal wired hands-free headset ............................ 219
Figure O.1 – Orientation and surface of averaging volume relative to phantom surface ....... 235
Figure U.1 – Categories (classes) for comparison of the measured psSAR between the
Uniphantom (SAR uni ) and the SAM phantom (SAR SAM ) ...................................................... 252
Figure U.2 – Histogram of the deviation of the 10 g psSAR of 45 handsets with straight
form factors positioned at the flat bottom of the Uniphantom ............................................... 254
Figure U.5 – Histogram of the deviation of the 1 g psSAR from 20 handsets with
straight form factors positioned at the flat bottom of the Uniphantom .................................. 257
form factors or smart phones positioned at the flat bottom of the Uniphantom ..................... 258
form factors or smart phones positioned at the flat bottom of the Uniphantom ..................... 259
form factors or smart phones positioned at the corner of the Uniphantom ........................... 260
form factors or smart phones positioned at the corner of the Uniphantom ........................... 261
Figure U.10 – Histogram of the deviation of the 10 g psSAR of 20 handsets with
clamshell form factors at the corner of the Uniphantom ....................................................... 262
clamshell form factors at the corner of the Uniphantom ....................................................... 263
Figure V.1 – Generated RF input power variations to operation time without and with
application of AIPLC ........................................................................................................... 264
Figure V.2 – The system block diagram of the AIPLC .......................................................... 265
Figure V.3 – Power variation characteristics by adjusting the amplifier or signal
generator outputs................................................................................................................ 265
Figure W.1 – Low, middle, and high channels at 2 GHz band (Band 1) ............................... 267
Figure W.2 – RF conducted power versus 10 g psSAR ....................................................... 268
Figure W.3 – 1 g SAR as a function of RF conducted power in various test conditions ........ 269
Table 1 – Evaluation plan checklist ....................................................................................... 28

Table 2 – Dielectric properties of the tissue-equivalent medium ............................................ 32
Table 3 – Area scan parameters ........................................................................................... 63
Table 4 – Zoom scan parameters .......................................................................................... 63
Table 5 – Example method to determine the combined SAR value using Alternative 1 .......... 70
Table 6 – Root-mean-squared error SAR correction formula as a function of the
maximum change in permittivity or conductivity [28] .............................................................. 83
Table 7 – Threshold values TH(f) used in this proposed test reduction protocol .................... 93
Table 8 – Divisors for common probability density functions (PDFs) .................................... 101
Table 9 – Uncertainty budget template for evaluating the uncertainty in the measured
value of 1 g or 10 g psSAR from a DUT or validation antenna (N = normal, R =
rectangular) ........................................................................................................................ 103
Table 10 – Uncertainty of Formula (8) (see 7.8.2) as a function of the maximum
change in permittivity or conductivity .................................................................................. 107
Table B.1 – The number of DUTs used for the statistical study ........................................... 123
– 12 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table B.2 – Statistical analysis results of P(Tilt/Cheek > x) for various x values .................. 124
Table B.3 – Statistical analysis results of P(Tilt/Cheek > x) for 1 g and 10 g psSAR ............ 124
Table B.4 – Statistical analysis results of P(Tilt/Cheek > x) for various antenna
locations ............................................................................................................................. 125
Table B.5 – Statistical analysis results of P(Tilt/Cheek > x) for various frequency bands ..... 125
Table B.6 – Statistical analysis results of P(Tilt/Cheek > x) for various device types ........... 126
*
Table B.7 – Distance Dmin for various “iso-level” values ..................................................... 130
Table B.8 – Experimental thresholds to have a 95 % probability that the maximum
measured SAR value from the area scan will also have a psSAR ........................................ 132
Table B.9 – SAR values from the area scan (GSM 900 band): Example 1 ........................... 133
Table B.10 – SAR values from the area scan (GSM 900 band): Example 2 ......................... 133
Table C.1 – Measurement uncertainty budget for relative SAR measurements using
Class 2 fast SAR testing, for tests performed within one frequency band and
modulation .......................................................................................................................... 141
Table C.2 – Measurement uncertainty budget for system check using Class 2 fast SAR
testing ................................................................................................................................ 142
Table D.1 – Mechanical dimensions of the reference dipoles .............................................. 146
Table D.2 – Numerical target SAR values (W/kg) for standard dipole and flat phantom ....... 147
Table D.3 – Mechanical dimensions of the standard waveguide .......................................... 148
Table D.4 – Numerical target SAR values for waveguides ................................................... 149
Table D.5 – Numerical target SAR values for CLAs ............................................................. 151
Table D.6 – Mechanical dimensions of the reference meander dipole ................................. 152
Table D.7 – Numerical target SAR value (W/kg) for meander dipole .................................... 153
Table D.8 – Dimensions for VPIFA antennas at different frequencies .................................. 154
Table D.9 – Electric properties for the dielectric layers for VPIFA antennas ........................ 155
Table D.10 – Numerical target SAR values for VPIFAs on the flat phantom ......................... 156
Table E.1 – Uncertainty analysis for transfer calibration using temperature probes ............. 160
Table E.2 – Guidelines for designing calibration waveguides .............................................. 163
Table E.3 – Uncertainty analysis of the probe calibration in waveguide ............................... 164
Table E.4 – Uncertainty template for evaluation of reference antenna gain ......................... 166
Table E.5 – Uncertainty template for calibration using reference antenna ........................... 167
Table E.6 – Uncertainty components for probe calibration using thermal methods .............. 170
Table F.1 – Suggested recipes for achieving target dielectric properties, 30 MHz to
900 MHz ............................................................................................................................. 178
Table F.2 – Suggested recipes for achieving target dielectric properties, 1 800 MHz to
10 000 MHz ........................................................................................................................ 179
Table G.1 – Dimensions used in deriving SAM phantom from the ARMY 90th percentile
male head data (Gordon et al. [61]) .................................................................................... 183
Table G.2 – Additional SAM dimensions compared with selected dimensions from the
ARMY 90th percentile male head data (Gordon et al. [61])—specialist head
measurement section .......................................................................................................... 183
Table G.3 – Parameters used for calculation of reference SAR values in Table D.2 ............ 190
Table H.1 – Parameters for calculating the dielectric properties of various reference
liquids ................................................................................................................................. 202
Table H.2 – Dielectric properties of reference liquids at 20 °C ............................................ 203
Table J.1 –psSAR correction factors ................................................................................... 209
IEC/IEEE 62209-1528:2020 – 13 –
© IEC/IEEE 2020
Table N.1 – SAR results tables for example test results in GSM 850 band .......................... 221
Table N.2 – SAR results tables for example test results in GSM 900 band .......................... 222
Table N.3 – SAR results tables for example test results in GSM 1800 band ........................ 222
Table N.4 – SAR results tables for example test results in GSM 1900 band ........................ 223
Table O.1 – Parameters for the reference function f 1 in Formula (O.12) .............................. 231
Table O.2 – Reference SAR values from the distribution functions f 1 , f 2 , and f 3 ................. 232
Table O.3 – Example uncertainty template and example numerical values for
permittivity ( ε r′ ) and conductivity (σ) measurement ............................................................. 237
Table S.1 – Uncertainties relating to the deviations of the parameters of the standard
waveguide from theory ........................................................................................................ 248
Table S.2 – Other uncertainty contributions relating to the dipole antennas specified in
Annex D. ............................................................................................................................. 249
Table S.3 – Other uncertainty contributions relating to the standard waveguides
specified in Annex D ........................................................................................................... 249
Table U.1 – Summary of information to determine the margin for handsets with straight
Table U.2 – Summary of information to determine the margin for handsets with straight
form factors, including slide-type and bar handsets (except smart phones), positioned
at the flat bottom of the Uniphantom ................................................................................... 256
Table U.3 – Summary of information to determine the margin for the smart phones
positioned at the flat bottom of the Uniphantom .................................................................. 258
Table U.4 – Summary of information to determine the margin for smart phones
positioned at the corner of the Uniphantom ......................................................................... 260
Table U.5 – Statistical analysis results of P(Tilt/Cheek > x) for various device types ........... 261
Table U.6 – Summary of information to determine the margin for handsets with
clamshell form factors positioned at the corner of the Uniphantom ...................................... 262
Table W.1 – Relative standard deviation of α found in Study 1 (without MPR) ..................... 267
Table W.2 – Maximum relative standard deviation of α found in Study 2 (with MPR) ........... 269
– 14 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
INTERNATIONAL ELECTROTECHNICAL COMMISSION
____________
MEASUREMENT PROCEDURE FOR THE ASSESSMENT OF SPECIFIC

ABSORPTION RATE OF HUMAN EXPOSURE TO RADIO FREQUENCY
FIELDS FROM HAND-HELD AND BODY-MOUNTED WIRELESS
COMMUNICATION DEVICES –

FOREWORD
1) The International Electrotechnical Commission (IEC) is a worldwide organization for standardization comprising
all national electrotechnical committees (IEC National Committees). The object of IEC is to promote
international co-operation on all questions concerning standardization in the electrical and electronic fields. To
this end and in addition to other activities, IEC publishes International Standards, Technical Specifications,
Technical Reports, Publicly Available Specifications (PAS) and Guides (hereafter referred to as “IEC
Publication(s)”). Their preparation is entrusted to technical committees; any IEC National Committee interested
in the subject dealt with may participate in this preparatory work. International, governmental and non-
governmental organizations liaising with the IEC also participate in this preparation.
IEEE Standards documents are developed within IEEE Societies and Standards Coordinating Committees of the
IEEE Standards Association (IEEE SA) Standards Board. IEEE develops its standards through a consensus
development process, approved by the American National Standards Institute, which brings together volunteers
representing varied viewpoints and interests to achieve the final product. Volunteers are not necessarily
members of IEEE and serve without compensation. While IEEE administers the process and establishes rules
to promote fairness in the consensus development process, IEEE does not independently evaluate, test, or
verify the accuracy of any of the information contained in its standards. Use of IEEE Standards documents is
wholly voluntary. IEEE documents are made available for use subject to important notices and legal disclaimers
(see http://standards.ieee.org/IPR/disclaimers.html for more information).
IEC collaborates closely with IEEE in accordance with conditions determined by agreement between the two
organizations. This Dual Logo International Standard was jointly developed by the IEC and IEEE under the
terms of that agreement.
2) The formal decisions of IEC on technical matters express, as nearly as possible, an international consensus of
opinion on the relevant subjects since each technical committee has representation from all interested IEC
National Committees. The formal decisions of IEEE on technical matters, once consensus within IEEE Societies
and Standards Coordinating Committees has been reached, is determined by a balanced ballot of materially
interested parties who indicate interest in reviewing the proposed standard. Final approval of the IEEE
standards document is given by the IEEE Standards Association (IEEE SA) Standards Board.
3) IEC/IEEE Publications have the form of recommendations for international use and are accepted by IEC
National Committees/IEEE Societies in that sense. While all reasonable efforts are made to ensure that the
technical content of IEC/IEEE Publications is accurate, IEC or IEEE cannot be held responsible for the way in
which they are used or for any misinterpretation by any end user.
4) In order to promote international uniformity, IEC National Committees undertake to apply IEC Publications
(including IEC/IEEE Publications) transparently to the maximum extent possible in their national and regional
publications. Any divergence between any IEC/IEEE Publication and the corresponding national or regional
publication shall be clearly indicated in the latter.
5) IEC and IEEE do not provide any attestation of conformity. Independent certification bodies provide conformity
assessment services and, in some areas, access to IEC marks of conformity. IEC and IEEE are not responsible
for any services carried out by independent certification bodies.
6) All users should ensure that they have the latest edition of this publication.
7) No liability shall attach to IEC or IEEE or their directors, employees, servants or agents including individual
experts and members of technical committees and IEC National Committees, or volunteers of IEEE Societies
and the Standards Coordinating Committees of the IEEE Standards Association (IEEE SA) Standards Board,
for any personal injury, property damage or other damage of any nature whatsoever, whether direct or indirect,
or for costs (including legal fees) and expenses arising out of the publication, use of, or reliance upon, this
IEC/IEEE Publication or any other IEC or IEEE Publications.
8) Attention is drawn to the normative references cited in this publication. Use of the referenced publications is
indispensable for the correct application of this publication.
IEC/IEEE 62209-1528:2020 – 15 –
© IEC/IEEE 2020
9) Attention is drawn to the possibility that implementation of this IEC/IEEE Publication may require use of
material covered by patent rights. By publication of this standard, no position is taken with respect to the
existence or validity of any patent rights in connection therewith. IEC or IEEE shall not be held responsible for
identifying Essential Patent Claims for which a license may be required, for conducting inquiries into the legal
validity or scope of Patent Claims or determining whether any licensing terms or conditions provided in
connection with submission of a Letter of Assurance, if any, or in any licensing agreements are reasonable or
non-discriminatory. Users of this standard are expressly advised that determination of the validity of any patent
rights, and the risk of infringement of such rights, is entirely their own responsibility.
International Standard IEC/IEEE 62209-1528 has been prepared by IEC technical committee
106: Methods for the assessment of electric, magnetic and electromagnetic fields associated
with human exposure, in cooperation with the International Committee on Electromagnetic
Safety of the IEEE Standards Association, under the IEC/IEEE Dual Logo Agreement.
This first edition of IEC/IEEE 62209-1528 cancels and replaces IEC 62209-1:2016,
IEC 62209-2:2010, IEC 62209-2:2010/AMD1:2019 and IEEE Std 1528-2013. This edition
constitutes a technical revision.
This edition includes the following significant technical changes with respect to the previous
edition:
a) extension of the frequency range down to 4 MHz and up to 10 GHz;

b) testing of devices with proximity sensors;
c) application specific phantoms;
d) device holder specifications;
e) fast SAR testing procedures;
f) test reduction procedures;
g) LTE assessment procedure;
h) revision of validation clause, including validation antennas;
i) revision of SAR assessment procedure;
j) time-average SAR measurement procedure;
k) uncertainty analysis;
This publication is published as an IEC/IEEE Dual Logo standard.
This publication contains attached files in the form of the Fast SAR Wizard described in
7.9.2.2 as well as CAD files for the SAM phantom. These files are available at:
http://www.iec.ch/dyn/www/f?p=103:227:0::::FSP_ORG_ID,FSP_LANG_ID:1303,25.
These files are intended to be used as a complement and do not form an integral part of the
publication.
The text of this standard is based on the following IEC documents:
FDIS Report on voting

106/514/FDIS 106/520/RVD
Full information on the voting for the approval of this standard can be found in the report on
voting indicated in the above table.
International Standards are drafted in accordance with the rules given in the ISO/IEC
Directives, Part 2.
– 16 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The IEC Technical Committee and IEEE Technical Committee have decided that the contents
of this publication will remain unchanged until the stability date indicated on the IEC web site
under "http://webstore.iec.ch" in the data related to the specific publication. At this date, the
publication will be
• reconfirmed,
• withdrawn,
• replaced by a revised edition, or
• amended.
IMPORTANT – The 'colour inside' logo on the cover page of this publication indicates
that it contains colours which are considered to be useful for the correct
understanding of its contents. Users should therefore print this document using a
colour printer.
IEC/IEEE 62209-1528:2020 – 17 –
© IEC/IEEE 2020
INTRODUCTION
The objective of this document is to provide procedures for measuring the human exposure
from devices intended to be used at a position near the human head or body. It was
developed to provide procedures to evaluate electromagnetic field (EMF) exposures due to
radio frequency (RF) transmitting devices used next to the ear, in front of the face, mounted
on the body, operating in conjunction with other RF-transmitting and non-transmitting devices
or accessories (e.g. belt-clips), or embedded in garments. The types of devices dealt with
include but are not limited to mobile telephones, cordless telephones, cordless microphones,
and radio transmitters in personal computers. The applicable frequency range is from 4 MHz
to 10 GHz. The document defines:
– measurement system requirements (Clause 6),

– SAR measurement protocols (Clause 7),
– SAR measurement uncertainty evaluation (Clause 8), and
– reporting requirements (Clause 9).
At the time this document was developed, two computational and measurement joint IEC/IEEE
projects dealing with millimetre-wave power density assessment were under development,
covering the frequency range from 6 GHz to 300 GHz. Hence there is an overlap of frequency
between this document, which deals with SAR, and the other joint IEC/IEEE projects dealing
with power density from 6 GHz to 10 GHz. The IEC/IEEE joint working group was aware of
this fact and believed that it would give the flexibility of using whatever metrics suitable for the
considered case of compliance assessment.
– 18 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
MEASUREMENT PROCEDURE FOR THE ASSESSMENT OF SPECIFIC
ABSORPTION RATE OF HUMAN EXPOSURE TO RADIO FREQUENCY
FIELDS FROM HAND-HELD AND BODY-MOUNTED WIRELESS
COMMUNICATION DEVICES –

1 Scope
This document specifies protocols and test procedures for the reproducible and repeatable
measurement of the conservative exposure peak spatial average SAR (psSAR) induced inside
a simplified model of the head and the body by radio-frequency (RF) transmitting devices,
with a defined measurement uncertainty. These protocols and procedures apply to a
significant majority of the population, including children, during the use of hand-held and
body-worn wireless communication devices. These devices include single or multiple
transmitters or antennas, and are operated with their radiating structure(s) at distances up to
200 mm from a human head or body. This document is employed to evaluate SAR compliance
of different types of wireless communication devices used next to the ear, in front of the face,
mounted on the body, operating in conjunction with other RF-transmitting, non-transmitting
devices or accessories (e.g. belt-clips), or embedded in garments. The applicable frequency
range is from 4 MHz to 10 GHz. Devices operating in the applicable frequency range can be
tested using the phantoms and other requirements defined in this document.
The device categories covered include, but are not limited to, mobile telephones, cordless
microphones, and radio transmitters in personal, desktop and laptop computers, for
multi-band operations using single or multiple antennas, including push-to-talk devices. This
document can also be applied for wireless power transfer devices operating above 4 MHz.
This document does not apply to implanted medical devices.
2 Normative references
The following documents are referred to in the text in such a way that some or all of their
content constitutes requirements of this document. For dated references, only the edition
cited applies. For undated references, the latest edition of the referenced document (including
any amendments) applies.
IEC 62209-3:2019, Measurement procedure for the assessment of specific absorption rate of
human exposure to radio frequency fields from hand-held and body-mounted wireless
communication devices – Part 3: Vector measurement-based systems (Frequency range of
600 MHz to 6 GHz.
ISO/IEC Guide 98-3:2008, Uncertainty of measurement – Part 3: Guide to the expression of

uncertainty in measurement (GUM:1995)
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
ISO, IEC, and IEEE maintain terminological databases for use in standardization at the
following addresses:
• IEC Electropedia: available at http://www.electropedia.org/

IEC/IEEE 62209-1528:2020 – 19 –
© IEC/IEEE 2020
• ISO Online browsing platform: available at http://www.iso.org/obp
• IEEE Dictionary Online: available at http://dictionary.ieee.org
3.1
accessory
optional component or attachment that can be used in conjunction with the device under test
(DUT)
EXAMPLE 1 Accessories for holding, affixing, or otherwise carrying, wearing or attaching the device, as well as
providing a specific use condition on the body of the user (e.g. a belt-clip, wrist-strap or any other body strap, or
lanyard for wearing the device using a necklace).
EXAMPLE 2 Electronic accessories for performing specialized tasks or which provide additional features
(e.g. GPS modules, outboard printers, MP3 players, cameras or viewing devices).
EXAMPLE 3 Electronic accessories providing audio or video input or output (e.g. headsets, microphones,
cameras).
EXAMPLE 4 Accessories providing enhanced RF capability to the device (e.g. original, replacement, or auxiliary
antennas).
EXAMPLE 5 Batteries and related DC power components.
EXAMPLE 6 Combinations of accessories, where two or more of the above are combined within one component.
3.2
axial isotropy
maximum deviation of the SAR measured when rotating the probe around its major axis while
it is exposed to a wave impinging from a direction coinciding with its major axis
3.3
basic restriction
human exposure limits for compliance with time-varying electric, magnetic, and
electromagnetic fields measured inside the body that are based on established adverse health
effects
Note 1 to entry: Within the scope of this document, the physical quantity used as a basic restriction is the specific
absorption rate (SAR).
3.4
body-worn device
device containing one or more wireless transmitters or transceivers which is positioned in
close proximity to a person’s torso or limbs (excluding the head) by means of a carrying
accessory to support intended use operations
Note 1 to entry: This document uses the term body-worn synonymously for the term body-mounted, which was
employed in earlier versions.
3.5
body-supported device
device containing one or more wireless transmitters or transceivers, with intended use that
includes transmitting with any portion of the device being held directly against a user’s body
3.6
boundary effect
change in the sensitivity of an electric-field probe when the probe tip is located close (less
than one probe-tip diameter) to media boundaries
Note 1 to entry: This effect is caused by distortion of the scattered field at the probe tip due to the nearby
phantom dielectric surface. This effect can be compensated if the probe orientation with respect to the phantom
surface is known.
– 20 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
3.7
cable
wire that is necessary for the device functionality in the intended operational configuration
3.8
channel
RF channel
specific sub-division of the transmit frequency range according to the operating parameters of
a particular wireless technology
Note 1 to entry: The number of RF channels and the corresponding channel bandwidth can vary among individual
wireless technologies. For the purpose of this document, SAR measurements are performed at specific channels;
for example, the high, middle, and low channels of the transmission band.
3.9
conducted power
power delivered by the power amplifier of the device to a matched load
3.10
conservative exposure
estimate of the psSAR, including uncertainties as specified in this document, representative of
and slightly higher than that expected to occur in the head or body of a significant majority of
the human population during intended use of a wireless transmitting device
3.11
constant-envelope mode
transmission mode where the signal amplitude does not vary with time
3.12
correlated signals
electromagnetic fields, associated to distinct signal waveforms, yielding non-zero time-domain
correlation integral at some time instant
Note 1 to entry: Details are specified in IEC TR 62630 [1].
3.13
desktop device
device placed or mounted on a desk, table, or similar supporting structure, with antennas that
are intended to be operated at or closer than 200 mm from the human body
3.14
device holder
fixture constructed of low-loss dielectric material that is used to hold the device under test in
the required test position during SAR measurement
3.15
device under test
DUT
device containing one or more wireless transmitters or transceivers that is tested according to
the methods of this document
Note 1 to entry: A device under test may be further categorized as a body-worn, body-supported, desktop, front-
of-face, hand-held, limb-worn, clothing-integrated, or generic device.
3.16
duty factor
proportion of time that a transmitter transmits over a specified period
IEC/IEEE 62209-1528:2020 – 21 –
© IEC/IEEE 2020
3.17
fast SAR testing
usage of special techniques, methods, or algorithms to decrease the SAR measurement time
compared to full SAR testing
Note 1 to entry: Class 1 fast SAR testing uses hardware that complies with all of the requirements of this
document for full SAR testing. However, it uses faster testing procedures than those that comply with full SAR
testing procedures.
Note 2 to entry: Class 2 fast SAR testing uses hardware and testing procedures that do not comply with the
requirements of this document for full SAR testing.
3.18
frequency band
transmitting frequency range associated with a specific wireless operating mode
Note 1 to entry: The frequency band is usually referred to using rounded figures, however the actual frequency
allocation can be slightly different; e.g. the GSM 850 MHz band actually uses 824 MHz to 849 MHz and 869 MHz to
894 MHz, and the GSM 900 MHz band actually uses 880 MHz to 915 MHz and 925 MHz to 960 MHz. In other
cases, the official 3GPP nomenclature might be used, for instance LTE Band 12 (uplink 699 MHz to 716 MHz;
downlink 729 MHz to 746 MHz).
3.19
front-of-face device
hand-held device containing one or more wireless transmitters or transceivers operated in
close proximity to the face
Note 1 to entry: Front-of-face device types include PTT devices, two-way radios, and devices equipped with an
optical camera.
3.20
full SAR testing
usage of methods, procedures and specific hardware which fully comply with all of the
normative requirements in this document, except those specified in 7.9.2
3.21
generic device
device that cannot be categorized as belonging to any of the specific device categories
3.22
hand-held device
portable device which is operated in a user’s hand at a distance of over 200 mm from the
head or torso during its intended use
3.23
handset
wireless communication device intended to be held in the hand and operated next to the ear,
consisting of an acoustic output or earphone and an acoustic input or microphone, and
containing a radio transmitter and receiver
Note 1 to entry: For synonymous terms such as “mobile phone”, the terms “mobile” and “portable” have specific
but generic meanings in IEC 60050 – mobile: capable of operating while being moved (IEC 60050-151:2001, 151-
16-46); portable: capable to be carried by one person (IEC 60050 151:2001, 151-16-47). The term “portable” often
implies the ability to operate when carried on the user. These definitions are used interchangeably in various
wireless regulations and industry specifications, in some cases referring to types of wireless communication
devices and in other cases to intended use.
3.24
head-mounted device
device containing one or more wireless transmitters or transceivers that is positioned in close
proximity to a person’s head or neck by means of an accessory, if necessary, to support its
intended use operation
Note 1 to entry: This definition excludes hand-held devices.

– 22 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
3.25
hemispherical isotropy
maximum deviation of the measured SAR when rotating the probe around its major axis with
the probe exposed to reference waves impinging from the half space in front of the probe tip,
at varying incidence angles relative to the axis of the probe
3.26
host
equipment that needs to be operated with a wireless device or module during testing to
provide transmitting functionality according to its intended use
3.27
intended use
use conditions for which a product is intended according to the full range of available
functions, in accordance with the specifications, instructions, and information provided by the
manufacturer in the user documentation
3.28
limb-worn device
device containing one or more wireless transmitters or transceivers, with intended use that
includes being strapped onto the arm or leg of the user while transmitting
Note 1 to entry: Limb-worn device types include wrist-worn, ankle-worn, and forearm-worn devices. Some
regulatory jurisdictions specify different requirements for devices worn on limbs away from the wrists or ankles.
3.29
linearity error
maximum deviation of a measured quantity from the expected linear dependency over the
measurement range
3.30
measurement drift
continuous or incremental change in the measured quantity over time, due to changes in
metrological properties of a measuring instrument
3.31
measurement uncertainty
3.31.1
uncertainty evaluation
<Type A> evaluation of uncertainty by the statistical analysis of series of observations
(measurements)
3.31.2
uncertainty evaluation
<Type B> evaluation of uncertainty by means other than the statistical analysis of series of
observations (measurements)
3.31.3
standard uncertainty
estimated standard deviation of a measurement result, equal to the positive square root of the
estimated variance
3.31.4
combined uncertainty
estimated standard deviation of the measurement result obtained by combining the individual
standard uncertainties of both Type A and Type B evaluations using the usual “root-sum-
squares” method of combining standard deviations that were obtained by taking the positive
square root of the estimated variances
IEC/IEEE 62209-1528:2020 – 23 –
© IEC/IEEE 2020
3.31.5
expanded uncertainty
quantity defining an interval about the result of a measurement that is expected to encompass
a distribution of values within a specified confidence interval that could reasonably be
attributed to the measured quantity
3.32
multi-band
<wireless communication device> capable of operating in more than one frequency band
3.33
operating mode
wireless protocol or standard used by a device to communicate with another terminal or
device over the wireless network
Note 1 to entry: The operating mode includes all parameters to establish the communication over the wireless
network. The specific operating mode parameters include, but are not limited to, communication standard,
frequency band, channel, signal modulation, communications protocol, data rate, bandwidth, number and positions
of time slots, and number and positions of resource blocks. Some operating modes do not have some of these
parameters.
Note 2 to entry: Modern terminals can have multiple operating modes for individual or simultaneous transmission.
Examples of operating modes include wireless protocols such as GSM, EDGE, EVDO, GPRS, CDMA, WCDMA,
LTE, 5G NR, Bluetooth® 1, and Wi-Fi® 2. Multiple operating modes can transmit in one or more frequency bands,
e.g. a DUT could support both multiple wireless technologies in a frequency band, and also allow multiple wireless
transmission modes within a wireless technology, such as GSM, GPRS, EDGE, or WCDMA and HSPA, or 5G NR.
3.34
output power
power at the output of the RF transmitter when the antenna, or a load with the same input
impedance as the antenna, is connected to it
3.35
primary peak SAR value
largest interpolated local SAR value determined in an area scan measurement
3.36
secondary peak SAR value
in an area scan measurement, interpolated local SAR maximum that is smaller than the
primary peak SAR value
3.37
peak spatial-average SAR
psSAR
maximum SAR averaged within a local region based on a specific averaging mass, e.g. any
1 g or 10 g of tissue in the shape of a cube
Note 1 to entry: In this document, the terms psSAR (over 1 g or 10 g) and the terms 1 g SAR and 10 g SAR are
used interchangeably.
3.38
penetration depth
depth at which the SAR of an associated incident plane wave, penetrating into a
homogeneous lossy medium, is reduced to 1/e 2 of its value just beneath the surface of the
medium
___________
1 Bluetooth® is a trademark of the Bluetooth Special Interest Group (SIG). This information is given for the
convenience of users of this document and does not constitute an endorsement by IEC.
2 Wi-Fi® is a trademark of the Wi-Fi Alliance. This information is given for the convenience of users of this
document and does not constitute an endorsement by IEC.
– 24 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Note 1 to entry: For a plane-wave incident normally on a planar half-space, the E-field penetration depth δ is
given as:
1
−
  2  2
1  µ0εr′ε0   σ  
δ
=  1 +  ωε ′ε  − 1 
ω  2   r 0  
  
where
ω is the angular frequency;
μ0 is the magnetic permeability of vacuum, in henry per metre;
ε0 is the dielectric permittivity of vacuum, in farad per metre;
ε r′ is the relative dielectric permittivity of the lossy medium, unitless;
σ is the electrical conductivity of the lossy medium, in siemens per metre.
3.39
phantom
physical model with an equivalent human anatomy and comprised of a tissue-equivalent
medium with dielectric properties specified in this document
3.40
pinna
auricle
cartilaginous projecting portion of the outer ear, consisting of the helix, lobule, and anti-helix
3.41
power control algorithm
protocol of the DUT to set and adjust the maximum output power of the transmitter over a
specified averaging period
3.42
production variation
range in output power among DUT production samples when set to transmit at the maximum
level
3.43
push-to-talk device
hand-held radio transceiver that enables a user to operate a mechanical switch to toggle
between radio transmission and reception (simplex operating mode)
EXAMPLE Two-way radio.
3.44
probe isotropy
property pertaining to the degree to which the response of an electric field probe or magnetic
field probe is independent of the polarization and direction of propagation of the incident wave
Note 1 to entry: See also axial isotropy (3.2) and hemispherical isotropy (3.25).
3.45
proximity sensor
capacitive sensor or combination of sensors in the DUT utilized for the detection of user
proximity for the purpose of limiting transmitter power in order to ensure compliance with RF
exposure limits
IEC/IEEE 62209-1528:2020 – 25 –
© IEC/IEEE 2020
3.46
readout electronics
measurement system component that connects to the E-field probe and provides an analog-
to-digital conversion of the measured values to the post-processor of the measurement
system
3.47
response time
time required by the measuring equipment to reach 90 % of its final value after a step
variation of the input signal
3.48
scanning system
automatic positioning system capable of placing the measurement probe at specified positions
according to requirements of this document
3.49
sensitivity
ratio of the magnitude of the system response (e.g. voltage) to the magnitude of the quantity
being measured (e.g. electric field strength squared)
3.50
separation distance
distance between the device under test and the outside surface of the phantom
Note 1 to entry: The separation distance represents the distance during intended use.
3.51
specific absorption rate
SAR
measure of the rate at which energy is absorbed by the human body when exposed to a radio
frequency electromagnetic field
Note 1 to entry: The SAR in the tissue-equivalent medium can be determined by the rate of temperature increase
or by E-field measurements, according to the following formulas:
σE 2
SAR =
ρ
∂T
SAR = ch
∂t t =0
where:
SAR is the specific absorption rate in watts per kilogram;
E is the RMS value of the electric field strength in the tissue-equivalent medium in volts per metre;
σ is the electrical conductivity of the tissue-equivalent medium in siemens per metre;
ρ is the mass density of the tissue-equivalent medium in kilograms per cubic metre;
ch is the specific heat capacity of the tissue-equivalent medium in joules per kilogram per kelvin;
∂T is the initial time derivative of temperature in the tissue-equivalent medium in kelvin per second.
∂t t =0
3.52
test configuration
set of DUT parameters applicable to the SAR measurement, consisting of the position of the
DUT on the applicable phantom, the DUT use configuration (e.g. antenna position), and
operating mode
– 26 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
3.53
time-period averaged SAR
TPAS
SAR averaged over a specified averaging period when a power control algorithm is applied
3.54
time-averaged SAR
SAR measured over a sufficiently long integration time to ensure a stable result over the duty
factor and other characteristics of the modulated signal
3.55
tune-up specification
output power tolerance when setting a sample of the DUT to transmit at the maximum output
power level
3.56
TX factor
ratio of the average output power to the peak output power of the DUT over a specified time-
averaging period
4 Symbols and abbreviated terms
4.1 Physical quantities
The International System of Units (SI) is used throughout this document.
Symbol Quantity Unit Dimensions

α Attenuation coefficient reciprocal metre 1/m
P Average (temporal) absorbed power watt W
J Current density ampere per square metre A/m 2
σ Electric conductivity siemens per metre S/m
E Electric field strength volt per metre V/m
f Frequency hertz Hz
ρ Mass density kilogram per cubic metre kg/m 3
δ Penetration depth metre m
µ Permeability henry per metre H/m
joule per kilogram per

ch Specific heat capacity J/(kg K)
kelvin
ε Permittivity farad per metre F/m
SAR Specific absorption rate watt per kilogram W/kg
T Temperature kelvin K
λ Wavelength metre m
NOTE In this document, temperature is quantified in degrees Celsius, as determined by: T (°C) = T (K) – 273,15 K.
4.2 Constants
Symbol Physical constant Magnitude
c0 Speed of light in vacuum 2,998 × 10 8 m/s
ε0 Permittivity of free space 8,854 × 10 −12 F/m
μ0 Permeability of free space 4π × 10 −7 H/m
IEC/IEEE 62209-1528:2020 – 27 –
© IEC/IEEE 2020
4.3 Abbreviated terms
5G NR Fifth generation new radio
AIPLC Automatic input power level control
APS Absolute psSAR
CAD Computer aided design
CDMA Code division multiple access
CLA Confined-loop antenna
CW Continuous Wave
DCS Digital Cellular Service
DOE Design of experiments
DUT Device under test
EDGE Enhanced Data rates for GSM Evolution
E-field Electric field strength
ERP Ear reference point
FDTD Finite-difference time-domain
GPRS General Packet Radio Service
GSM Global System for Mobile Communications
H-field Magnetic field strength
ICES International Committee on Electromagnetic Safety
ICNIRP International Commission on Non-Ionizing Radiation Protection
LE Left ear
LTE Long Term Evolution
MIMO Multiple input multiple output
MPR Maximum power reduction
OFAT One-factor-at-a-time
OFDM Orthogonal frequency-division multiplexing
OFDMA Orthogonal frequency-division multiple access
PCS Personal Communications Service
PDF Probability density function
psSAR Peak spatial-average SAR
PTT Push-to-talk
RE Right ear
RF Radio frequency
RMS Root mean square
RSS Root sum square
SAM Specific anthropomorphic mannequin
SAR Specific absorption rate
STBC Space-time block coding
TDMA Time division multiple access
TX Transmission
UMTS Universal Mobile Telecommunications System
VPIFA Vertical Planar Inverted F-Antenna
WCDMA Wideband code division multiple access
WiMax Worldwide Interoperability for Microwave Access
WLAN Wireless local area network
– 28 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
5 Quick start guide and evaluation plan checklist
The quick start guide shown in Figure 1 and the evaluation plan checklist shown in Table 1
provide a high level overview of the SAR measurement procedures of this document. For the
details of each step, refer to the relevant clause specified in Figure 1 and Table 1.
Table 1 – Evaluation plan checklist
Stage Evaluation steps Action

Evaluation – What type of device is being evaluated? Complete checklist
plan checklist – Review of antennas and supported wireless technologies, based on evaluation
single or multiple combinations. plan.
Planning of
the – What are the test positions applicable?
measurement
configurations – Working under laboratory conditions and following good
laboratory practice and documentation, prepare to conduct the
necessary evaluation.
Preliminary – A liquid dielectric properties measurement shall be performed Should be conducted

actions and and recorded. *For array systems, this is typically provided by before starting a project
measurement the manufacturer and repeated.
system check – System check needs to be performed to verify that the system
components are working correctly.
DUT set-up The protocol for SAR assessment defines all the permutations of Ensure the DUT is
operational conditions that should be tested. The essential steps configured in the correct
include: way and radiating
– Preparing the DUT by establishing a connection to a base appropriately when
station simulator. positioned against the
phantom.
– Configuring the DUT to operate on the appropriate test
frequencies.
– Positioning the DUT in relation to the phantom.
Uncertainty – Guidelines and approximation formulas are provided, enabling Determine uncertainty
evaluation the estimation of each individual uncertainty component. and complete the
– The uncertainty budget shall cover the appropriate frequency uncertainty table.
range with regards to equipment used in the SAR system.
SAR – This stage is the actual procedure for the measurement of Determining highest
measurements SAR. SAR. Ensure the
procedures – The procedure is iterative to ensure that the highest value of rationale for test
the psSAR of a DUT is captured. reduction or fast SAR
testing is recorded and
– This is done by a systematic process looking at all device reported if used.
positions, configurations and operating modes in all frequency
bands accordingly.
– Measurement systems can incorporate a single probe or array
probes. Either fast SAR testing or test reduction techniques
may be used. Some accepted techniques may be used to
either reduce or accelerate the measurement process.
Reporting – The final report describes the results of the evaluations, Prepare the final
SAR results provides enough technical details to allow for repeatability of measurement report
the evaluations performed, and reports the results by along with addressing
comparison with the relevant limit. the requirements.
– The measurement report shall contain enough details and
information for demonstrating compliance to the requirements
of this document.
IEC/IEEE 62209-1528:2020 – 29 –
© IEC/IEEE 2020
Figure 1 – Quick start guide

– 30 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
6 Measurement system specifications
6.1 General requirements for full SAR testing
A SAR measurement system consists of the phantom (a SAM phantom representing the
human head, or a flat phantom representing the human body, or both) filled with tissue-
equivalent medium, electronic measurement instrumentation, a scanning system, and a DUT
holder. Procedures for full SAR testing are described in 7.4. SAR tests performed in
accordance with this document shall conform to the following requirements.
a) SAR shall be measured using a probe that is automatically positioned to measure the
internal E-field distribution in the phantom. From the measured E-field values, the SAR
distribution and the psSAR value shall be calculated.
b) The phantom shall be filled with the required tissue-equivalent medium, representing the
dielectric properties of tissues in the human head or body.
c) The tissue-equivalent medium shall be a liquid of low viscosity to allow free movement of
the probe within it.
d) For the tissue-equivalent media, a density ρ of 1000 kg/m 3 shall be used.
e) At least three reference points shall be defined on the phantom by the phantom
manufacturer to be used when aligning the scanning system with the phantom. These
points shall be visible to the user, enclosing at least 80 % of the phantom top surface and
each point shall be at least 20 cm from any other point.
f) Both the ambient temperature and tissue-equivalent medium temperature shall be in the
range of 18 °C to 25 °C, inclusive; see O.9.6 to determine the temperature uncertainty.
g) Prior to the measurements of tissue-equivalent medium dielectric properties and SAR, the
DUT, test equipment, medium, and phantom shall have been kept in the laboratory long
enough for their temperatures to have stabilized (i.e. they shall not have been recently
moved from another area with a different ambient temperature, such as a refrigerator or
storage area).
NOTE Depending on the shape, size, volume, and initial temperature of the medium and the container,
24 hours or more might be required for temperature stabilization.
h) The temperature of the tissue-equivalent medium during the SAR measurements shall be
within 2 °C (or a temperature difference corresponding to a 5 % change in either
permittivity or conductivity, if this is smaller) of the temperature at which the dielectric
properties were measured. If the temperature change exceeds this tolerance, the dielectric
properties shall be re-measured.
i) The effect of reflections from cables, test equipment, or other reflectors shall be
determined by the SAR system check procedure described in Clause A.2, by comparing
the SAR measured with and without the reflectors present, or where necessary with the
judicious placement of absorbing materials and/or the use of ferrite beads on cables.
j) SAR measurements of DUTs shall only be performed when the effects of reflections,
secondary RF transmitters, etc., result in a psSAR (for 1 g or 10 g mass, whichever is
applicable to the test) less than 0,012 W/kg by measuring the psSAR at (approximately)
0,4 W/kg (used to establish the 3 % low detection limit, see 8.4.1.7). When the effect of
cables and reflectors is more than 0,012 W/kg, ferrite beads, RF absorbers, and other
mitigation techniques shall be applied to reduce the SAR error.
If the preceding limit cannot be achieved, a value higher than 3 % (0,012 W/kg) may be
considered in the uncertainty budget in the “RF ambient conditions–reflections” row of the
applicable tables in Annex C, provided it can be demonstrated that the SAR contribution
due to reflections determined by the system check procedure is less than 10 % of the SAR
measured for the DUT.
k) The requirement on reflections shall be verified at least every year, or whenever the
system check shows unexpected results.
l) During testing, the DUT shall not be connected to any wireless network, except a base
station simulator in the laboratory.
IEC/IEEE 62209-1528:2020 – 31 –
© IEC/IEEE 2020
m) System validation according to the protocol specified in Clause A.3 shall be done at least
once per year, and additionally when a new measurement system is put into operation,
and whenever modifications have been made to the measurement system (such as a new
software version, different type or version of readout electronics, or different probes). The
transmit sources used for system validation shall be designed and validated to meet the
requirements specified in Annex D. Additional transmit sources of dipole and waveguide
types at specific frequencies not presently included in Table D.1, Table D.2, Table D.3,
Table D.4, Table D.5, Table D.6, Table D.7, Table D.8, Table D.9, and Table D.10 may be
used as sources provided they meet the requirements specified in Annex D. In this
document, flat phantoms shall be used for system check and system validation.
n) The measurement system shall be validated as a complete system.
o) The probe shall be calibrated in a tissue-equivalent medium required for the calibration
frequency and temperature range, according to the methodology specified in Annex E.
Calibration of the probe separately from the measurement system is allowed, provided that
the same electrical interface characteristics specified for the probe and readout
electronics are used during measurements. The probe(s) shall be calibrated together with
an identical amplifier, measurement device, and data acquisition system.
p) The lower detection limit shall be at most 0,01 W/kg, and the maximum detection limit
shall be at least 100 W/kg.
q) The probe isotropy shall be less than the expanded uncertainty for the axial isotropy
component using the procedures of Table 9 and 8.4.1.5, or ±0,2 dB (±5 %), whichever is
less, as determined in the tissue-equivalent medium.
r) The outermost diameter of the probe tip shall not exceed 8 mm in the vicinity of the
measurement elements, for frequencies up to and including 2 GHz.
s) For frequencies above 2 GHz, the probe tip diameter shall not exceed λ/3, where λ is the
wavelength in the medium. The probe tip diameter may be larger if it can be demonstrated
that the E-field from any potential field distribution can be measured with an uncertainty of
less than ±15 % (k = 2) at the distances from the surface of the phantom as listed in
Table 3 and Table 4, or at the distances recommended by the SAR measurement system
manufacturer (whichever is less).
t) The system linearity shall be within ±0,5 dB over the SAR range from 0,12 W/kg to
100 W/kg.
u) Where this document explicitly specifies performance characteristics for the measurement
system or a part of the measurement system, the manufacturer of the system or sub-
system, or the system integrator, shall document the conformity with the provisions of this
document.
6.2 Phantom specifications
6.2.1 General
Subclause 6.2 specifies phantoms to be used for psSAR measurements of the DUT.
Subclause 6.2.3 defines the head phantom to be used for evaluating exposure in the head,
next to the ear. Subclause 6.2.4 defines the flat phantom to be used for evaluating exposure
in the torso or other body regions. If the standard head or flat phantoms are not applicable,
due to physical constraints of the scanning system or intended use conditions of the DUT, the
specific phantoms described in 6.2.5 may be used.
6.2.2 Basic phantom parameters
The phantom shell material shall be resistant to damage or reaction with the chemicals used
in the tissue-equivalent medium. The phantom shell shall be made of low-loss and low-
permittivity material, having loss tangent tan δ ≤ 0,05 and relative permittivity:
• εr′ ≤ 5, for f ≤ 3 GHz

• 3 ≤ εr′ ≤ 5, for f > 3 GHz
– 32 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Target dielectric properties of the tissue-equivalent medium in the phantom are specified in
Table 2 for discrete frequencies ranging between 4 MHz and 10 GHz. For other frequencies
within this range, the nominal dielectric properties shall be obtained by linear interpolation
between the higher and lower tabulated figures. Example recipes for preparing tissue-
equivalent media designed to produce the dielectric properties at some of the frequencies in
the 4 MHz to 10 GHz range are provided in Annex F. The rationale for the medium dielectric
properties is given in Clause G.5. The liquid temperature specifications are given in 6.1.
NOTE 1 Nominal values are target dielectric properties with a specific tolerance requirement.
To minimize reflections from the upper surface of the tissue-equivalent medium, the depth of
the medium should be at least 15 cm. Medium depth of less than 15 cm may be used only if it
is demonstrated (e.g. using numerical simulations) that the effect on psSAR is less than 1 %
during system validation (see Annex A). If it is more than 1 % but less than 3 %, uncertainty
for the highest value from the system validation shall be added to the uncertainty budget.
The dielectric properties shall be measured and compared with the values given in Table 2
using linear interpolation. This measurement can be performed using the equipment and
procedures specified in Annex H. The measured dielectric properties, not the values of
Table 2, shall be used in the SAR calculations.
NOTE 2 The dielectric properties (permittivity and conductivity) of the medium have been specified to result in
conservative exposure (see Clause G.5 for the rationale used to derive the medium dielectric properties).
NOTE 3 See 7.2.1 for the allowable variations between the measured and the target dielectric properties specified
in Table 2.
Table 2 – Dielectric properties of the tissue-equivalent medium
Real part of the

complex relative Penetration depth
Frequency Conductivity, σ
permittivity, ε r′ (E-field), δ
MHz S/m mm
4 55,0 0,75 293,0
13 55,0 0,75 165,5
30 55,0 0,75 112,8
150 52,3 0,76 62,0
300 45,3 0,87 46,1
450 43,5 0,87 43,0
750 41,9 0,89 39,8
835 41,5 0,90 39,0
900 41,5 0,97 36,2
1 450 40,5 1,20 28,6
1 800 40,0 1,40 24,3
1 900 40,0 1,40 24,3
1 950 40,0 1,40 24,3
2 000 40,0 1,40 24,3
2 100 39,8 1,49 22,8
2 450 39,2 1,80 18,7
2 600 39,0 1,96 17,2
3 000 38,5 2,40 14,0
3 500 37,9 2,91 11,4
4 000 37,4 3,43 10,0
4 500 36,8 3,94 9,7
IEC/IEEE 62209-1528:2020 – 33 –
© IEC/IEEE 2020
Real part of the
complex relative Penetration depth
Frequency Conductivity, σ
permittivity, ε r′ (E-field), δ
MHz S/m mm
5 000 36,2 4,45 1,5
5 200 36,0 4,66 8,4
5 400 35,8 4,86 8,1
5 600 35,5 5,07 7,5
5 800 35,3 5,27 7,3
6 000 35,1 5,48 7,0
6 500 34,5 6,07 6,7
7 000 33,9 6,65 6,4
7 500 33,3 7,24 6,1
8 000 32,7 7,84 5,9
8 500 32,1 8,46 5,3
9 000 31,6 9,08 4,8
9 500 31,0 9,71 4,4
10 000 30,4 10,40 4,0
NOTE For convenience, permittivity and conductivity values are linearly interpolated for frequencies
that are not a part of the original data from Drossos et al. [2]. They are shown in italics in Table 2. The
italicized values are linearly interpolated (below 5 800 MHz) or extrapolated (above 5 800 MHz) from
the non-italicized values that are immediately above and below these values.
6.2.3 Head phantom
The specific anthropomorphic mannequin (SAM) phantom shall be used for evaluating
exposure in the head. The rationale for choosing the specific head phantom model (i.e. SAM)
for this document is based on the following criteria.
a) The psSAR shall provide conservative exposure.

b) The test results shall not unnecessarily overestimate the maximum SAR expected in
actual users.
c) The phantom shall allow stable and repeatable device positioning for psSAR
measurements, and be effective for verifying measurement repeatability according to
system check procedures, and measurement reproducibility through inter-laboratory
comparisons.
d) The phantom shall be practical for routine SAR evaluation use.
e) The phantom shall support these criteria for contemporary and future wireless device
designs, and be unbiased with respect to any particular design or shape.
f) Based on the presently available science, literature, and experience, the design of the
SAM phantom that meets the above criteria is a function of at least the following
parameters:
1) size and shape of the head phantom shell;
2) dielectric properties and material homogeneity of the tissue-equivalent media and the
phantom shell;
3) ear pinna size, shape, location, and material properties;
4) exclusion of the hand for measuring SAR in the head (Annex I, 6.3).
– 34 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
6.2.4 Flat phantom
a) For evaluating exposure in the body, the following requirements apply.
1) The flat phantom shall be used.
2) Its shell shall be constructed in the form of an open-top container with a flat bottom.
3) Any flat phantoms used shall be large enough to allow the measurement of SAR in 1 g
and 10 g volumes with an influence from the container shape of less than 1 %
(Clause G.4).
b) The flat phantom shall have the following shape and dimensions (Figure 2).
1) Except as specified in b) 2), the phantom flat bottom wall shall have the shape of an
ellipse with a length of 600 mm and a width of 400 mm.
2) For frequencies above 150 MHz, and for separation distances less than or equal to
25 mm from the outer surface of the bottom of the phantom shell, phantoms with other
shapes and smaller dimensions are allowed as follows:
i) between 150 MHz and 800 MHz, the phantom flat bottom wall may have any shape
that encompasses an ellipse with length of 0,6 λ 0 and width of 0,4 λ 0 , where λ 0 is
the wavelength in air;
ii) between 800 MHz and 10 GHz, the phantom may have any shape flat bottom wall
that encompasses an ellipse with length of 225 mm and width of 150 mm.
c) The thickness of the bottom-wall of the flat phantom shall be 2,0 mm, with a tolerance of
±0,2 mm. When filled with the required depth of medium, any sagging of the bottom
surface of the phantom within the projection of the DUT shall be determined and
considered in the uncertainty budget (see 8.4.2.4). If the preceding dimensional
requirements are met, the effect of the shape and thickness on the repeatability of SAR
measurement results is less than 1 % (see Clause G.4). Effects on the SAR due to the
influence of the deviation from the required shell parameters and thickness shall be
included in the uncertainty estimation (see 8.4.2.3).
NOTE Use of a flat phantom as a standard phantom for SAR evaluations of body-worn and body-supported
devices is intended to represent maximum coupling between the DUT and the human body. Such coupling to the
phantom, while maintaining prescribed distances, is likely to produce conservative exposure. Annex J provides
more details about this topic.
IEC/IEEE 62209-1528:2020 – 35 –
© IEC/IEEE 2020
Dimensions in millimetres
Figure 2 – Dimensions of the elliptical phantom
6.2.5 Device-specific phantoms
Device specific phantoms, alternative to those specified in 6.2.3 or 6.2.4, may be used if the
flat phantom or SAM does not support the exposure conditions, or if either of the following
applies:
a) the device is intended for operation on different parts of the body other than the head or
torso (e.g. the wrist);
b) the physical constraints imposed on the scanning system by the phantom do not allow
psSAR for the DUT to be determined with the required accuracy.
Examples of alternative device-specific phantoms are specified in Annex K and Annex L. The
scientific or technical rationale and specifications of these phantoms are provided in the
aforementioned annexes. Alternative device positioning procedures can also be required.
These, and the rationale for using alternative specific phantoms, shall be reported.
6.3 Influence of hand on SAR in head
During normal operation, the head and hand are in the near-field of the DUT when used next
to the ear, and hence both absorb energy. For extremities such as the hand, a higher SAR
limit is allowed (e.g. 4 W/kg averaged over 10 g of tissue in ICNIRP RF exposure guidelines
[3] and IEEE Std C95.1-2019 [4]). Numerical and experimental studies have shown that the
SAR in the hand at the power levels used by typical DUTs is not expected to exceed those
limits (Francavilla et al. [5], Gandhi et al. [6], Jensen and Rahmat-Samii [7], [8], Kuster et al.
[9], Watanabe et al. [10]). Furthermore, a practical phantom for SAR measurement in the
hand is currently unavailable. Therefore, SAR measurement in the hand is not addressed in
this document.
– 36 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The influence of a hand holding a DUT to the head during SAR measurements was
considered in IEEE Std 1528-2013 [11]. Earlier work by Balzano et al. [12] and Kuster et al.
[9] reported that the presence of the hand either decreased or had no significant effect on the
head SAR; however numerical results by Meyer et al. [13] showed one case with 7 % increase
in head SAR due to the hand. These deviations in head SAR were concluded to be within the
conservativeness provided by the SAM phantom. Based on these studies, the exclusion of the
hand from test procedures would lead to head SAR overestimation in the majority of
situations, as confirmed by more recent research results [14], [15], [16] on DUT SAR levels.
For these reasons, hand models are not considered in this document.
The state of dosimetric research on the effect of the hand on head SAR produced by DUTs is
specified in Annex I. This research shows that there are cases when the SAR produced by
DUTs in the SAM phantom can increase as well as decrease significantly due to the hand
holding the DUT, for specific DUT designs, operating bands, and hand grips. These initial
findings have shown that the SAM phantom alone can still overestimate head SAR in a
statistically significant number of cases, compared to when the hand is present for the
measurement. Nevertheless this deserves further investigations, which may potentially
warrant future changes in this document.
6.4 Scanning system requirements
The SAR probe scanning system shall be able to scan the required measurement regions of
the phantom that are within the projections of a DUT in order to evaluate the area and zoom
scans. The tolerance of the probe tip positioning at a measurement point in the phantom shall
be ≤ 0,2 mm. Accuracy of the probe positioning system in relation to the phantom shall be
verified according to the phantom reference points defined by the phantom manufacturer. The
mechanical structure of the scanning system shall not interfere with the SAR measurements.
The scanning range shall cover at least 90 % of the phantom dimensions in all directions.
6.5 Device holder specifications
The device holder shall permit the DUT to be positioned according to the definitions given in
7.2.4. It shall be made of low loss and low permittivity material(s): loss tangent ≤ 0,05, and
relative permittivity ≤ 5.
The DUT is preferably mounted on a piece of low relative permittivity (< 1,2) and low loss
foam before being mounted into the device holder, to avoid any direct contact between the
device holder and the DUT (Figure 3). If this is not possible, the device holder shall provide
the minimum amount of contact to the DUT to give a secure hold and maintain the required
position during the measurement. In cases where a default relative positioning cannot be
achieved, e.g. due to the device holder interaction with buttons and sensors on the DUT, then
minimal position offsets in a predefined direction should be applied to achieve the required
DUT test position. The details shall be provided in the measurement report.
The device holder shall ensure precise and repeatable positioning of the DUT for SAR
measurement. The positioning uncertainties shall be estimated following the procedures
specified in 8.4.2.5 and 8.4.2.6. If the highest psSAR is determined according to step 3 of the
test protocol in 7.3.2, then the device holder uncertainty effect on the reported SAR value is
considered to be negligible.
IEC/IEEE 62209-1528:2020 – 37 –
© IEC/IEEE 2020
NOTE The DUT may be fixed on top of the foam (left), or held down using a thin strap (right).
Figure 3 – Mounting of the DUT in the device holder using low-permittivity and
low-loss foam to avoid changes of DUT performance by the holder material
6.6 Characteristics of the readout electronics
The probe output is processed by the readout electronics and associated instrumentation that
combines the voltages from the probe’s sensors to provide an output that is proportional to
the amplitude-squared of the E-field incident on the sensors. Detector diodes at the dipole
feed-point are used to rectify the sensor voltages. The rectified signals are transmitted
through resistive (RF-transparent) lines to the readout electronics system. For a continuous-
wave signal at low field strength levels, the probe output is proportional to the square of the
amplitude of the incident E-field; at higher signal levels (above the diode compression point),
the output is not linearly proportional to |E| 2 , but becomes proportional to |E|. This signal
compression will lead to an underestimation of the actual SAR at high field strength conditions
if it is not compensated correctly through probe calibration. Also, amplifiers in the readout
electronics can deviate from an ideal linear response and introduce additional uncertainty. For
uncertainties associated with the probe readout electronics, see 8.4.1.6.
7 Protocol for SAR assessment
7.1 General
All measurements should be carried out following good laboratory practice, e.g. in accordance
with ISO/IEC 17025 [17], and any other national requirements for device certification. This
document does not contain methods to verify the parameters of the DUT operating mode
(e.g. modulation, data rate, time slots, resource blocks; see 3.33).
7.2 Measurement preparation
7.2.1 Preparation of tissue-equivalent medium and system check

a) The measured dielectric properties of the tissue-equivalent media shall be applied in the
SAR evaluation. The dielectric properties shall be measured within 24 hours before the
SAR measurements, and for every 48 hours of continuous use. However, in the situation
where a laboratory documents compliance with Table 2 and the requirements of 6.2.2
using measurement intervals of up to a maximum of one week, less frequent dielectric
properties measurements are acceptable. When the test series takes longer than
48 hours, the medium properties shall also be measured at the end of the test series.
– 38 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
b) Tissue-equivalent media shall yield measured relative permittivity and conductivity values
within ±5 % of the target values at frequencies at which the SAR is measured. When the
method of 7.8 is used to correct the measured SAR for the deviations in permittivity and
conductivity, the tolerance may be relaxed to a maximum of ±10 % from the targets. When
out-of-tolerance dielectric properties measurements are made during a measurement
interval longer than 48 hours, all SAR data measured since the last dielectric properties
measurement shall be discarded.
c) For fast SAR testing systems that are hermetically sealed, the manufacturer shall provide
either a procedure for the accurate measurement of the dielectric properties, or
documentation of the dielectric properties of the medium (with corresponding medium
temperature and ambient temperature) that were measured for the specific phantom when
it was manufactured, or during system recalibration.
d) The temperature of the tissue-equivalent medium shall be measured and recorded at the
time of the dielectric properties measurement, and prior to the system check and SAR
measurements.
e) The system check shall be performed at each frequency band to be tested, in order to
detect possible drift over short time periods and other measurement errors in the system.
A system check according to the procedures of Clause A.2 shall be completed within 24
hours before performing SAR measurements for a DUT.
f) The system check procedure shall be performed on the same SAR measurement system,
with the same SAR probe(s) and same tissue-equivalent medium required by the SAR
evaluation of the device for each frequency band tested. To limit power drift during the
system check, automatic input power level control may be implemented, as specified in
Annex V.
7.2.2 Preparation of the wireless communication DUT

a) Any radiating element(s), battery, and accessories shall be those specified by the
manufacturer, and documented in the measurement report. The battery shall be fully
charged before each measurement, and the DUT positioned without external connections
or cables that are not required for normal operation.
b) The RF output power and frequency (channel) shall be controlled by a wireless link to a
base station or network simulator for supported operating modes. When a wireless link
cannot be established with a base station simulator, test software and internal test codes
may be used.
NOTE 1 Some national regulatory agencies might require that verification be undertaken by the device
manufacturer for tests performed on prototypes, to ensure that the commercial version has the exact same
mechanical and electrical characteristics as the tested prototype.
c) The DUT shall be set to transmit at the highest time-averaged output power level, within
the range of the product’s tune-up specifications and production variations.
For testing of wireless power transfer devices, other preparation procedures should be
applied, as specified in IEC TR 62905 [18].
NOTE 2 For devices that can monitor and control their transmit power to maintain the time-average SAR below
the applicable SAR limit, see 7.6.
7.2.3 DUT operating mode requirements
7.2.3.1 Requirements on DUT output power and signal parameters
The objectives of 7.2.3.1 are as follows.
• To determine that SAR of the DUT is evaluated at its maximum power level, taking into
consideration device maximum output power tolerance and production variations.
• To correctly apply probe linearity compensation for time-varying signals, and to ensure
that the signal is sampled correctly.
IEC/IEEE 62209-1528:2020 – 39 –
© IEC/IEEE 2020
The power requirements are as follows.
a) The maximum time-averaged power levels of the DUT shall be verified by the device
manufacturer, e.g. by conducted power tests with a fully charged battery, to support SAR
scaling.
b) All applicable operating modes shall be considered for testing, although testing of some
operating modes might not be required (7.9.3).
c) A DUT operating in constant-envelope modes shall be tested at the maximum output
power of its CW-equivalent (carrier) signal.
d) A DUT operating in pulsed-envelope modes shall be tested using the same pulse
characteristics (duty factor, pulse period). If the DUT is capable of transmitting several
modes by varying the pulse characteristics (e.g. transmission in multiple slots) in a
wireless technology, the signal with the highest time-average output power per frequency
band shall be tested, after considering the maximum output power of the DUT for each
mode.
e) If the pulsed-envelope mode has a long pulse period such that the lowest frequency
component of the signal envelope is below the low-pass filter cutoff frequency of the SAR
measurement acquisition sub-system, the integration time and/or sample acquisition
period shall be chosen to accurately measure the RMS of the signal, e.g. by using an
integral number of pulse periods to avoid aliasing.
f) A DUT in operating modes where the envelope of the signal varies randomly with time
(e.g. CDMA, OFDM) shall be tested in the mode resulting in the maximum time-averaged
output power transmitted by the device for each wireless technology. The probe shall be
calibrated over its dynamic range, and the probe output shall include probe linearity
compensation. The linearity compensation in general depends on the operating mode, as
reported by Di Nallo and Faraone [19] and by Nadakuduti et.al [20].
g) If the DUT cannot operate for the duration of the SAR measurement at its highest time-
averaged output power level, within the range of the product’s tune-up specifications and
production variations, the following shall be applied.
1) The DUT may be operated at a lower output power or lower duty factor, and the SAR
shall be scaled to the required highest time-averaged output power level. However, the
time-average output power shall be within the tolerance range specified for the device.
2) The SAR of the device shall be shown to increase linearly with conducted power.
3) Devices that operate with dual-output amplifiers, where both amplifiers are not
required when operating at low power levels, or those that are designed to operate
intermittently at very low duty factors, might require additional considerations to ensure
SAR can be scaled linearly.
NOTE Some regulatory authorities might impose additional requirements on the maximum time-averaged output
power reduction allowed for device SAR testing.
The procedure for power scaling is given in 7.2.3.2.
7.2.3.2 SAR scaling procedure for DUT power or signal variations
SAR scaling is the extrapolation of the SAR of a DUT determined with a tested operating
mode (mod X) to a SAR of the same device in the same exposure test position and frequency
channel with a different operating mode (mod Y ). The difference can be in the power level,
operating mode, or both. The time-averaged RF output power ratio of mod X and mod Y shall
both be determined either by measuring the average power, or by numerical integration of the
power envelope if the signals are sufficiently well known. SAR scaling is possible if:
a) the same RF amplifier stage is used for mod X and mod Y ;

b) the same antenna is used for mod X and mod Y , and MIMO or other antenna diversity
techniques are not applied;
c) the probe modulation response uncertainty has been evaluated for the modulated signal
mod X (8.4.2.7), and the SAR has been determined for mod X ;
– 40 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
d) the time-averaged RF output power ratio (R p ) of mod X and mod Y after the RF amplifier
stage is known, according to Formula (1):
Pmax, mod Y
Rp = (1)
Pmod X
where P max,modY corresponds to the highest time-averaged output power setting. If the two
signals differ in operating mode, then the test signal will be set at maximum time-averaged
output power: P modX = P max,modX.
The factor R p shall be determined by experimental means (e.g. measurement using an
average power meter).
e) the RF carrier frequency of mod X is the same as for mod Y ;
f) the RF signal bandwidth ratio R m of mod X and mod Y satisfies Formula (2):
BWmod Y
=Rm − 1 × 100 ≤ 30% (2)
BWmod X
g) the channel bandwidths of mod X and mod Y are each within 5 % of the carrier frequency.
If the preceding requirements are fulfilled, a scaling of the SAR from mod X and mod Y shall be
performed according to Formula (3), and scaling uncertainty shall be determined as specified
in 8.4.3.2:
SARmod=
Y Rm ⋅ SARmod X (3)
When the approach in 7.2.3.2 is used, it shall be stated in the measurement report that points
a) through g) are satisfied.
For devices supporting PTT modes or other technology specific TDD operations, the
maximum duty factor shall be assessed.
7.2.4 Positioning of the DUT relative to the phantom
7.2.4.1 Test distance and positioning relative to the flat phantom
7.2.4.1.1 General
Subclause 7.2.4.1 describes how the DUT, with or without accessories, shall be positioned,
oriented, and configured relative to the phantom for SAR measurement. The specified test
position(s) is (are) applicable for device-to-phantom surface separations up to and including
200 mm.
Subclause 7.2.4.1 describes positioning procedures for the following device types:
– devices that are large relative to phantom surface area (7.2.4.1.4);

– devices used with body-worn accessories (7.2.4.1.5);
– devices with hinged or swivel antenna (7.2.4.1.6);
– devices used next to or against the body without an accessory (7.2.4.1.7);
– desktop device (7.2.4.1.8);
– front-of-face device (7.2.4.1.9);
– hand-held usage of the device, not at the head or torso (7.2.4.1.10);
– limb-worn device (7.2.4.1.11);
– clothing-integrated device (7.2.4.1.12);
IEC/IEEE 62209-1528:2020 – 41 –
© IEC/IEEE 2020
– generic device (7.2.4.1.13);
NOTE The figures provided in 7.2.4.1 are for illustration purposes only, and are not to scale or representative of
the areas that need SAR measurement.
7.2.4.1.2 DUT-to-phantom separation distance
When performing measurements against the flat phantom, the DUT-to-phantom separation
distance may be set according to the following hierarchy.
a) Regulatory requirement: When there is a national regulatory requirement that specifies the
DUT separation distance to the phantom, the DUT shall be positioned according to this
requirement.
b) “Intended use distance” specified by the manufacturer: When there is no regulatory
requirement, the intended use condition or distance specified by the manufacturer shall be
used. This information shall be acquired from the user documentation accompanying the
DUT.
c) Default separation distance: When there is neither a regulatory requirement nor an
intended use distance specified by the manufacturer, the DUT shall be measured with
each accessible face in direct contact with the surface of the phantom (see also 7.2.4.1.13
generic device).
The stated separation distance may be determined with the carry accessory attached to the
DUT, when such condition is tested; see 7.2.4.1.5.
Care should be taken to ensure precise positioning at the applicable separation distances, for
example using calibrated gauge blocks or similar means to ensure repeatability.
7.2.4.1.3 DUT reference points
General guidance to identify the orientation of a DUT, or when placed in a carry accessory for
positioning to the flat phantom, is as follows. P1, P2, P3, and P4 are defined as the midpoint
on each edge of the surface facing the phantom as shown in Figure 4. Line P1-P2 and Line
P3-P4 shall be parallel to the phantom surface such that the distance between P1 and the
phantom surface is equal to the distance between P2 and the phantom surface. Similarly, the
distance between P3 and the phantom surface shall be equal to the distance between P4 and
the phantom surface. The separation distance is specified as the distance between the
phantom shell and the closest point on the DUT when positioned as specified in the preceding
sentences. This general guidance may need to be adapted for specific designs, such as
devices or carry accessories with curved surfaces, that require the device to be positioned
differently for testing. Clear details of the actual test positions used shall be fully documented
in the measurement report.
Figure 4 – Designation of DUT reference points

– 42 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.2.4.1.4 Positioning for devices that are large relative to phantom surface area
If the DUT is larger than the measurable region of the flat phantom specified in 6.2.4, the
following applies.
a) The DUT shall be shifted such that the SAR distribution of the DUT can be covered by
multiple area scans. When the DUT is shifted, the RF coupling between the DUT and
phantom could change, resulting in a different SAR distribution compared to that using a
larger phantom. A larger phantom, however, might not be representative of the significant
majority of exposure situations involving people (see NOTE in 6.2.4).
b) To limit differences in the measured SAR due to coupling variations, two successive scans
shall have an overlapping area of at least one third the size of the combined area, as
shown in Figure 5.
c) In the overlapping area, the SAR distribution shall be similar such that the SAR difference
at any point relative to the DUT geometry is less than the expanded uncertainty for
repeatability. Otherwise, the DUT shall be shifted by a smaller amount such that the SAR
difference is within the uncertainty.
Shifting is not needed if the radiating structures are small compared to the DUT and the
phantom and/or the first area scan shows that the SAR distribution is entirely captured within
the scanning area. In the situation where shifting was omitted, the rationale shall be clearly
documented in the measurement report.
Figure 5 – Measurements performed by shifting a large device

over the efficient measurement area of the system including
overlapping areas – in this case: six tests performed
7.2.4.1.5 Devices used with body-worn accessories
A device within the scope of this document intended for use with a body-worn accessory
approved or specified by the device manufacturer shall be assessed according to the
following requirements.
a) If the user instructions provided by the manufacturer specify the intended use with a carry
accessory (belt-clip, holster, carry-case or similar), the device shall be placed in that carry
accessory and positioned in the intended use conditions against the flat phantom.
b) For carry accessories constructed from non-conductive materials that are capable of
holding the DUT at varying distances to the phantom, the carry accessory providing the
closest separation distance is expected to produce the highest SAR; therefore, testing of a
carry accessory providing larger separation distances is not necessary unless the carry
accessory contains conductive materials.
c) If there are multiple carry accessories that hold the DUT at the same distance, at least one
carry accessory should be tested. Note that all carry accessories containing conductive
materials shall be tested.
IEC/IEEE 62209-1528:2020 – 43 –
© IEC/IEEE 2020
d) For carry accessories that do not contain conductive materials (e.g. no metal parts), it is
acceptable to substitute the carry accessory with an air-gap that keeps the DUT at a
distance from the phantom surface no greater than the distance provided by the carry
accessory.
e) The device shall be positioned in the intended use configuration with the body-worn
accessory where the distance to the outer surface of the phantom is established by the
specific accessory (Figure 6).
f) When evaluating devices that are not supplied with body-worn accessories but users can
acquire body-worn accessories on their own, body-worn accessory SAR shall be
measured with a separation distance according to the conditions of 7.2.4.1.2.
g) The surface of the device pointing towards the flat phantom should be positioned at the
surface of the phantom according to 7.2.4.1.3; however, not all devices have a flat
surface. Therefore, the details of the device position, e.g. the definition of the distance and
the physical relationship between the device and the phantom according to the body-worn
accessories, shall be documented in the measurement report.
h) When it is unclear if multiple surfaces can operate in a body-worn accessory facing the
user, the DUT shall be tested with all applicable surfaces placed at the most conservative
separation distance from the flat phantom. The details of the device position, attached to
the accessory, shall be documented in the measurement report. If testing of some (but not
all) surfaces is not required, this shall be documented with an associated rationale in the
measurement report.
Figure 6 – Test positions for body-worn devices
7.2.4.1.6 Devices with hinged or swivel antenna(s)
For devices that employ one or more external antennas with variable positions (e.g. antenna
extended, retracted, rotated), the following applies.
a) The DUT shall be positioned in accordance with 7.2.4.1.2 and 7.2.4.1.3.

b) When it is unclear if certain antenna positions are applicable, tests shall be performed for
the antenna positions and polarizations that are expected to be used by the users
(Figure 7); for example with the antenna extended and retracted such as to obtain the
highest exposure condition.
c) For antennas that can be rotated through one or two planes, a preliminary investigation
across rotation ranges should be made and documented in the measurement report, to
determine the highest exposure scenario(s) so that only that position(s) need(s) to be
tested.
d) For devices with multiple detachable antennas, each antenna should be considered
separately to determine SAR test requirements; for example, if the antennas are
substantially similar (equivalent), certain test reduction considerations may be applied
(7.9.3.3).
– 44 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Subclause 7.2.4.1.6 may be applied in conjunction with test configurations specified in other
subclauses, as appropriate for the device.
Figure 7 – Device with swivel antenna
7.2.4.1.7 Devices used next to or against the body without an accessory
A typical example of a device used next to the body is a wireless enabled laptop or a tablet
device containing a peripheral plug-in radio transmitter that can be supported on the user’s
body. Other devices that fall into this category include tablets of equivalent sizes or larger
than a netbook computer (> 20 cm diagonal display), credit card transaction authorization
terminals, and point-of-sale and/or inventory terminals. The example in Figure 8 a) shows a
tablet form factor portable computer for which SAR should be separately assessed with each
surface and edge containing an antenna positioned against the flat phantom in accordance
with the conditions of 7.2.4.1.2. Some tablets might limit display orientations so that the edges
with antennas are not used next to the user; therefore, such edges might not need testing.
To assess these types of use conditions, the following applies.
a) The DUT shall be positioned with its base (bottom surface) and applicable edges against
the flat phantom. The orientations and other use positions can be host device specific.
Therefore, these should be determined according to the transmitter and antenna locations
on the host device.
NOTE 1 RF transmitters that can operate in different hosts might require additional consideration.
b) For antennas incorporated in the display screen of a laptop, notebook or netbook
computers, tests shall be performed with the display screen open at a 90° angle as seen
in Figure 8 b) (left side), or at a more conservative angle allowing the antennas to be
closer to the phantom for assessing the user exposure conditions.
c) Devices in the scope of 7.2.4.1.7 with a display screen that should be open during normal
operation generally do not need to be tested, provided the antenna(s) on the screen is
sufficiently far away from the user’s body according to the maximum output power of the
antenna.
d) Where a screen mounted antenna is present, the SAR measurement for bystander
exposure shall be performed with the screen positioned parallel to the flat phantom as
shown in Figure 8 b) (right side).
NOTE 2 A separation distance of 2,5 cm is considered to be conservative for covering a wide range of
expected bystander exposure conditions.
e) Body-supported devices with an external power supply (e.g. AC adapter) supplemental to
the battery shall be tested with and without the external power supply attached.
f) For devices that employ an external antenna with variable positions, the procedures for
swivel and detachable antenna should be used; see 7.2.4.1.6 and Figure 7.
IEC/IEEE 62209-1528:2020 – 45 –
© IEC/IEEE 2020
a) Tablet form factor portable computer
b) Portable computer with external-antenna plug-in radio card (left side)

or with internal antenna located in screen section (right side)
c) Wireless credit card transaction authorization terminal
Figure 8 – Test positions for body supported devices
7.2.4.1.8 Desktop device
A typical example of a desktop device is a wireless enabled desktop computer placed on a

table or desk at closer than 200 mm to users while being used.
– 46 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
To assess this type of device the following applies.
• The DUT shall be positioned against the flat phantom in accordance with the conditions of
7.2.4.1.2.
• For devices that employ an external antenna with variable positions, tests shall be
performed for all antenna positions specified. Figure 7 and Figure 9 illustrate the SAR test
positions for desktop device.
Due to the physical design, some device surfaces might not require testing, e.g. the base of a
desk standing device.
IEC/IEEE 62209-1528:2020 – 47 –
© IEC/IEEE 2020
Figure 9 – Test positions for desktop devices
7.2.4.1.9 Front-of-face device
A typical example of a front-of-face device is a two-way radio that is held at a close distance
from the face of the user while transmitting. Other devices that fall into this category include
wireless-enabled still cameras and video cameras capable of sending data to a network or
other device (Figure 10 a)).
– 48 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
To assess this type of device, the following apply:
• The DUT shall be positioned at a test separation distance to the phantom surface in
accordance with the conditions of 7.2.4.1.2.
• For a device with intended use that requires contact with the user’s face (e.g. device with
an optical viewfinder), such a device shall be placed directly against the phantom
(Figure 10 a), right side).
a) Still cameras and video cameras
b) Two-way radios
Figure 10 – Test positions for front-of-face devices

IEC/IEEE 62209-1528:2020 – 49 –
© IEC/IEEE 2020
7.2.4.1.10 Hand-held usage of the device, not at the head or torso
When SAR measurement is necessary for hand-held devices that do not transmit while at the
head or torso, a flat phantom may be used. To assess this type of device, the device shall be
placed directly against the flat phantom as shown in Figure 11, for the sides of the device that
are in contact with the hand for the intended use.
NOTE Concerning a measurement phantom representing the hand, there are practical difficulties in specifying a
unique hand holding position that is applicable to all devices. Additional studies are needed for devising a
representative method for evaluating SAR in the hand of hand-held devices (whether or not they are hand-operated
devices). Future versions of this document are intended to contain a test method based on scientific data and
rationale.
Figure 11 – Test position for hand-held devices, not used at the head or torso
7.2.4.1.11 Limb-worn device
Limb-worn devices are strapped to the arm or leg of the user while transmitting. These are
similar to a body-worn device.
To assess this type of device, the following applies.
• The test positions of 7.2.4.1.5 shall be applied.

• The device shall be placed against the phantom such that the measured SAR is a
conservative exposure (e.g. by opening or removing the strap as shown in Figure 12, if
applicable).
• When the flat phantom or a curved section of the SAM phantom is not applicable, specific
phantoms may be considered (e.g. see Annex K and Annex L).
Figure 12 – Test position for limb-worn devices
7.2.4.1.12 Clothing-integrated device
A typical example of a clothing-integrated device is a wireless communication device

integrated into a jacket to provide voice communications through an embedded speaker and
microphone. This category also includes head-mounted devices with integrated wireless
communication devices.
– 50 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
To assess this type of device, the following applies.
• All wireless or RF transmitting components shall be positioned against the flat phantom (or
specific phantoms, see Annex K and Annex L) that correspond to the intended use
conditions when they are integrated into the clothing (Figure 13).
• Devices integrated in head-mounted devices may be tested using the SAM phantom (if
appropriate) or specific phantoms (6.2.3 or 6.2.5).
Figure 13 – Test position for clothing-integrated wireless communication devices
7.2.4.1.13 Generic device
A device that cannot be categorized as any of the device types in 7.2.4.1.5 through 7.2.4.1.12
is considered a generic device; i.e. it is represented by a closed box incorporating at least one
internal RF transmitter and antenna. The generic device principle may be applied to any
device. In situations when a transmitter is added to the host device so that the host and
transmitter operate as a single device, the generic category may be considered only if the
provisions of 7.2.4.1.5 to 7.2.4.1.12 are not applicable. When the antenna or the attached RF
transmitter is external to the host and the positioning of the antenna or attached RF
transmitter is independent of positioning of the host (e.g. transmitter or antenna is attached by
a cable and 7.2.4.1.6 does not apply), it shall be assessed using the generic device
procedures.
The DUT distance hierarchy specified in 7.2.4.1.2 shall be used to assess generic devices.
Testing of all six faces of the DUT (see Figure 14) might not be required; justification shall be
provided when omitting testing of some faces. The surface of the generic device pointing
towards the flat phantom shall be according to 7.2.4.1.3.
IEC/IEEE 62209-1528:2020 – 51 –
© IEC/IEEE 2020
Figure 14 – Possible test positions for a generic device
7.2.4.2 Positioning relative to the head phantom
7.2.4.2.1 General
This document specifies two DUT test positions against the head phantom – the “cheek”
position and the “tilt” position. These two test positions are specified in 7.2.4.2.2 and
7.2.4.2.3. The DUT shall be tested in both of these positions, on the left and right sides of the
SAM phantom.
In some cases, such as asymmetric DUTs or other form factors, the DUT positioning
procedures of 7.2.4.2.2 and 7.2.4.2.3 cannot be used. In these cases, adapted alignment
procedures may be applied. The resulting test positions shall replicate the intended use
conditions as closely as possible.
Clear details of the actual test positions used, including adapted alignment procedures (where
needed), shall be fully documented in the measurement report.
7.2.4.2.2 Definition of the cheek position
The cheek position is established using steps a) to j) as follows.
a) Configure the DUT for voice operation, if necessary. For example, for a DUT with a flip,
swivel, or slide cover piece, open the cover if this is consistent with voice operation. If the
DUT can also be used with the cover closed, both configurations shall be tested.
b) Define two imaginary lines on the DUT, the vertical centreline and the horizontal line,
relative to the DUT in vertical orientation as shown in Figure 15.
c) The vertical centreline passes through two points on the front side of the DUT: the
midpoint of the width w t of the DUT at the level of the acoustic output (Point A in
Figure 15), and the midpoint of the width w b at the bottom of the DUT (Point B). The
horizontal line is perpendicular to the vertical centerline, and passes through the centre of
the acoustic output (Figure 15). The two lines intersect at Point A. Note that for many
DUTs, Point A coincides with the centre of the acoustic output. However, the acoustic
output could be located elsewhere on the horizontal line. Also note that the vertical
centreline is not necessarily parallel to the front face of the DUT, especially for clamshell
DUTs, DUTs with flip cover pieces, and other irregularly shaped DUTs.
– 52 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
d) Position the DUT close to the surface of the phantom such that Point A is on the (virtual)
extension of the line passing through points RE (right-ear ear reference point) and LE
(left-ear ear reference point) on the phantom (see Figure 16a) and Figure 16b)). The plane
determined by the vertical centreline and the horizontal line of the DUT shall be parallel to
the sagittal plane of the phantom.
e) Translate the DUT towards the phantom along the line passing through RE and LE until
the DUT touches the ear (see Figure 16c)).
f) Rotate the DUT around the (virtual) LE-RE Line until the DUT vertical centreline is in the
reference plane (see Figure 16d)).
g) Rotate the DUT around its vertical centreline until the plane established by the DUT
vertical centreline and horizontal line is parallel to the N-F line (see Annex G), and then
translate the DUT towards the phantom along the LE-RE line until DUT Point A touches
the ear at the ERP (ear reference point) (see Figure 16e)).
h) While keeping Point A on the line passing through RE and LE and maintaining the DUT in
contact with the pinna, rotate the DUT about the N-F line until any point on the DUT is in
contact with a phantom point below the pinna (cheek) (see Figure 16f)). The physical
angles of rotation shall be documented.
i) While keeping DUT Point A in contact with the ERP, rotate the DUT around a line
perpendicular to the plane established by the DUT vertical centreline and horizontal line
and passing through DUT Point A, until the DUT vertical centreline is in the reference
plane (see Figure 16g)).
j) Verify that the cheek position is correct as follows:
1) the N-F line is in the plane established by the DUT vertical centreline and horizontal
line;
2) DUT Point A touches the pinna at the ERP;
3) the DUT vertical centreline is in the reference plane.
IEC/IEEE 62209-1528:2020 – 53 –
© IEC/IEEE 2020
Key
w t Width of the DUT at the level of the acoustic output

w b Width of the bottom of the DUT
A Midpoint of the width w t of the DUT at the level of the acoustic output
B Midpoint of the width w b of the bottom of the DUT
Figure 15 – Vertical and horizontal reference lines and reference

points A and B on two example device types: a full touch-screen
smart phone (left) and a DUT with a keypad (right)
– 54 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
NOTE The reference points for the right-ear ear reference point (RE), left-ear ear reference point (LE), and
mouth (M), which establish the reference plane for DUT positioning, are indicated. This device position shall be
maintained for the sagittal phantom test set-up shown in Figure G.4.
a) Phone position 1 – cheek position
b) One possible DUT position against the head after applying 7.2.4.2.2 c)
NOTE The black arrows show the direction of translation of the DUT for 7.2.4.2.2 d).
c) DUT position after applying 7.2.4.2.2 d)

IEC/IEEE 62209-1528:2020 – 55 –
© IEC/IEEE 2020
NOTE The curved black arrows show the direction of rotation of the DUT for 7.2.4.2.2 e).
d) DUT position after applying 7.2.4.2.2 e)
NOTE The curved black arrows show the direction of rotation of the DUT for 7.2.4.2.2 f).
e) DUT position after applying 7.2.4.2.2 f)
NOTE The curved black arrows show the direction of rotation of the DUT for 7.2.4.2.2 g)
f) DUT position after applying 7.2.4.2.2 g)

– 56 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
NOTE The curved black arrows show the direction of rotation of the DUT for 7.2.4.2.2 h).
g) DUT position after applying 7.2.4.2.2 h)
Figure 16 – Cheek position of the DUT on the left side of SAM where
the device position shall be maintained for the phantom test set-up
7.2.4.2.3 Definition of the tilt position
The tilt position is established using steps a) through d) as follows.
a) Repeat steps a) through j) of 7.2.4.2.2 to place the DUT in the cheek position
(see Figure 16).
b) While maintaining the orientation of the DUT, move the DUT away from the pinna along
the line passing through RE and LE far enough to allow a rotation of the DUT away from
the cheek by 15°.
c) Rotate the DUT around the horizontal line by 15° (see Figure 17).
d) While maintaining the orientation of the DUT, move the DUT towards the phantom on a
line passing through RE and LE until any part of the DUT touches the ear. The tilt position
is obtained when the contact is on the pinna. If the contact is at any location other than
the pinna, e.g. an extended antenna in contact with the back of the head phantom, the
angle of the DUT shall be reduced. In this case, the tilt position is obtained if any part of
the DUT is in contact with the pinna and a second point on the DUT is in contact with the
phantom, e.g. the antenna in contact with the back of the head.
Key
M Mouth reference point

LE Left-ear ear reference point
RE Right-ear ear reference point
This device position shall be maintained for the phantom test set-up.
Figure 17 – Tilt position of the DUT on the left side of SAM

IEC/IEEE 62209-1528:2020 – 57 –
© IEC/IEEE 2020
7.2.5 Antenna configurations
For devices that employ one or more external antennas with variable positions (e.g. extended,
retracted, or rotated antennas), these shall be positioned in accordance with the user
instructions provided by the manufacturer. If no intended antenna position is specified, tests
shall be performed with the antenna(s) oriented to obtain the highest exposure condition while
maintaining the device in the positions of 7.2.4. For antennas that can be extended, testing
shall be performed with the antenna fully extended and fully retracted. The antenna
configurations shall be documented in the measurement report. Transmit diversity antennas
shall be tested independently for SAR.
7.2.6 Options and accessories
Other accessories that affect the RF output power or RF current distribution of the DUT when
used next to a person shall be tested according to the intended use conditions specified by
the manufacturer; see 7.2.4.1.5 for next to body positioning and 7.2.4.2 for next to head
positioning. For example, a) optional antennas, b) optional battery packs that change the
wireless communication device performance or SAR, etc., and c) wires connected during
intended use. Low-power NFC or receive-only wireless power transfer accessories generally
do not require SAR measurement, but their influence on the SAR of other transmitters shall be
considered.
7.2.7 DUTs with alternative form factor
The basic principles identified and specified in 7.2.4 may be applied to devices with similar
form factors. Wireless headsets (e.g. connected by Bluetooth) are examples of head-mounted
devices that may be evaluated by applying these principles.
The geometry and coordinate mapping system shown in Figure 18 is applicable. Point A,
being the acoustic output, is located at the mid-point of the width, and point B, being the
primary microphone, is located at the bottom of the device (the end closest to the mouth).
Note that for the purpose of applying the positioning procedures, the DUT is considered to be
a conventional bar type (rectangular, cuboid) form factor.
Prior to using 7.2.7, consideration shall be made of the available operating modes and the
maximum operating power levels, because some devices might not require testing.
All details relating to alternative form factor DUTs shall be fully documented in the
measurement report, including diagrams or photographs. Sound engineering practice shall be
applied to implement the mapping of an alternate form factor device.
Figure 18 – An alternative form factor DUT with reference points and reference lines
– 58 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.2.8 Test frequencies for DUTs
A DUT shall be compliant with applicable exposure standards at all transmitting frequencies.
However, testing at every channel is impractical and unnecessary. The purpose of 7.2.8 is to
define a practical subset of channels where SAR measurements are to be performed. This
subset of channels is chosen to give a characterization of a DUT for comparison with
applicable exposure standards.
Before performing testing using the specific channels required by this document, the
maximum output power of the channels shall be verified to determine that the selected
channels are producing the highest rated output of the device. The process used to establish
the channels for testing purposes shall be documented in the measurement report.
For each operating mode of a wireless technology used by the DUT:
a) tests shall be performed at the channel producing the highest rated output power;
b) when the width of the transmit frequency band (∆f = f high − f low ) exceeds 1 % of its centre
frequency f c , the channels at the lowest and highest frequencies of the transmit band shall
also be tested;
c) when the width of the transmit frequency band exceeds 10 % of its centre frequency,
Formula (4) shall be used to determine the number of channels, N c , to be tested:
 (
N c = 2 × roundup 10 × fhigh − flow ) fc  + 1

(4)
where
fc is the centre frequency channel of the transmission band in Hz;
f high is the highest frequency channel of the transmission band in Hz;
f low is the lowest frequency channel of the transmission band in Hz;
Nc is the number of channels.
NOTE 1 The function roundup(x) rounds its argument x to the next highest integer. Thus, the number of
channels, N c , will always be an odd number.
NOTE 2 National regulatory agencies might have different requirements on the number of channels to be
tested per transmission band, according to frequency allocations and other wireless technology requirements.
One example is given in Formula (5):
 100 f
Nc = round 
(high − flow ) fc 
0,2 
 (5)
   

fc  100  
The number of channels, N c , should be equally spaced and shall meet the preceding
requirements.
Probe calibrations shall be valid for all test frequencies and medium dielectric properties at
those frequencies. Multiple probe calibration points and different tissue-equivalent media to
cover the entire frequency band can be required for devices with substantially large
transmission bands.
7.3 Tests to be performed for DUTs
7.3.1 General
The DUT test procedure encompasses the following two main processes:
– the selection of the test configurations to be measured using the test reduction techniques
discussed in 7.9.3 (optional), and
IEC/IEEE 62209-1528:2020 – 59 –
© IEC/IEEE 2020
– the SAR evaluation of these test configurations using the full SAR testing procedure
discussed in 7.4.2; optionally, fast SAR testing conforming to the requirements set forth in
7.9.2 may be employed.
In all cases where test reductions (7.9.3) have been applied, the device including accessory
combinations that are excluded from SAR measurement and the rationale for test reduction
shall be clearly documented in the measurement report.
7.3.2 Basic approach for DUT testing
In order to determine the highest psSAR of a DUT, all required device test positions,
configurations and operating modes shall be tested for each frequency band according to the
following steps 1 through 3. For devices capable of simultaneous transmission, apply the
appropriate procedure specified in 7.4.4; the test process is shown in Figure 19. Application
examples are given in Annex N.
For LTE modes, the procedure of 7.9.3.6.3 shall be used, rather than 7.3.2.
Step 1: The measurement procedure specified in 7.4 shall be performed at the channel
producing the highest rated output power for each transmit antenna used:
a) for all device test positions (as specified in 7.2.4),

b) for all use configurations for each device position in a), e.g. with device slide or cover
open and closed, or antenna extended and retracted, and
c) for all operating modes (see 3.33), for each device position in a) and for each
configuration in b), in each frequency band.
Step 2: For the test configuration providing the highest psSAR determined in step 1, and for
each test in a), b) and c), perform all tests specified in 7.4 at the other applicable test
frequency channels, e.g. lowest and highest channels (7.2.8). In addition, for each device
position, configuration, and operating mode where the psSAR value determined in steps 1 a),
b), and c) is greater than or equal to half of the applicable SAR limit, testing of all channels of
7.2.8 is required; otherwise it is not required.
Step 3: If the highest psSAR from steps 1 and 2 is larger than the applicable SAR limit
reduced by the expanded measurement uncertainty; and if, for this test configuration, the DUT
was in direct contact with parts of the device holder having relative permittivity greater than
1,2, this test configuration shall be tested again after mounting the device on a piece of foam,
as specified in 6.5 and Figure 3. Both values shall be reported.
Step 4: Examine all data and report the measured psSAR from step 1, step 2, and step 3
according to regulatory requirements.
– 60 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure 19 – Block diagram of the tests to be performed
7.4 Measurement procedure
7.4.1 General
Subclause 7.4 describes the procedure for evaluating the psSAR of the DUT, including the
minimum number of drift measurements to be performed. If the drift is high for a particular
operating mode and frequency channel, more drift measurements can be necessary, as
specified in 7.4.3.
7.4.2 Full SAR testing procedure
The following procedure shall be performed for each test configuration (Figure 19) specified in
7.3. Table 3 provides the measurement parameters used in the area scan, and Table 4 for the
zoom scan.
IEC/IEEE 62209-1528:2020 – 61 –
© IEC/IEEE 2020
a) Measure the local SAR at a test point within 5 mm of the inner surface of the phantom
where the measured local SAR exceeds the lower detection limit of the measurement
system. Preferably, the test point will be near the expected peak SAR location, and still
within the said distance from the phantom surface. As explained in step f), a comparative
measurement shall be made by the system at the same point after completion of the SAR
measurement.
b) Measure the two-dimensional SAR distribution within the phantom (i.e. the area scan).
Table 3 provides the measurement parameters required for the area scan.
1) The area over which the SAR measurement is performed shall cover at least an area
larger than the projection of the DUT, including its antenna. For some DUTs, the area
projected onto the phantom can be relatively large, such that the probe might not reach
all points. In this case, rotated phantoms may be used, and the area may be assessed
by multiple overlapping area scans. The measurement resolution and spatial resolution
for interpolation shall be selected to allow identification of the local peak locations to
within one-half of the linear dimension of the corresponding side of the zoom-scan
volume.
2) For the flat phantom, the boundary of the measurement area shall not be closer than
20 mm from the phantom side walls.
c) From the area-scan SAR distribution, identify the position of the maximum SAR value. In
addition, identify the positions of any local maxima with SAR values within 2 dB of the
maximum value, and that will not be within the zoom scan of other peaks. Additional zoom
scans shall be measured for such peaks only when the primary peak is within 2 dB of the
SAR compliance limit (e.g. 1 W/kg for 1,6 W/kg 1 g limit, or 1,26 W/kg for 2 W/kg 10 g
limit).
d) Measure the three-dimensional SAR distribution at each of the local maxima locations
identified in step c) (i.e. the zoom scan).
1) For frequencies at or below 3 GHz, the following procedure shall be applied (see
Table 4).
i) The minimum size of the zoom scan volume shall be 30 mm by 30 mm by 30 mm.
ii) The horizontal grid step shall be 8 mm or less.
iii) The grid step in the vertical direction shall be 5 mm, or less if uniform spacing is
used.
iv) If variable spacing is used in the vertical direction, the maximum spacing between
the two closest measured points to the phantom shell (M1 and M2, see Figure 20)
shall be 4 mm or less, and the spacing between the farther points shall increase by
a factor of 1,5 or less.
v) For other parameters, see Table 4 and Figure 20.
2) For frequencies above 3 GHz, the following procedure shall be applied.
i) the minimum size of the zoom scan volume may be reduced to 22 mm by 22 mm by
22 mm.
ii) The horizontal grid step shall be (24 / f [GHz]) mm or less.
iii) If uniform spacing in the vertical direction is used, the grid step in the vertical
direction shall be (10 / (f [GHz] − 1)) mm or less.
iv) If variable spacing is used in the vertical direction, the maximum spacing between
the two measured points closest to the phantom shell shall be (12 / f [GHz]) mm or
less, and the spacing between farther points shall increase by a factor of 1,5 or
less.
v) For other parameters, see Table 4 and Figure 20.
3) If the highest SAR 1 g or 10 g cube is touching the boundary of a zoom-scan volume,
the entire zoom scan shall be repeated with the new centre located at the maximum
psSAR location indicated by the preceding zoom scan measurement. It is also
acceptable to expand the zoom scan during measurement until the 1 g or 10 g cube is
no longer touching the boundary of the zoom-scan volume.
– 62 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
4) If the zoom scan measured as specified in the preceding paragraphs complies with
both i) and ii), or if the psSAR is below 0,1 W/kg, no additional measurements are
needed.
i) The smallest horizontal distance from the local SAR peaks to all points 3 dB below
the SAR peak shall be larger than the horizontal grid steps in both x- and
y-directions (∆x, ∆y). This shall be checked for the measured zoom scan plane
conformal to the phantom at the distance z M1 . The minimum distance shall be
recorded in the SAR test report.
ii) The ratio of the SAR at the second measured point (M2) to the SAR at the closest
measured point (M1) at the x-y location of the measured maximum SAR value shall
be at least 30 % (Figure 20). This ratio (in %) shall be recorded in the SAR test
report.
NOTE 1 The evaluation of the zoom scan is typically done by the post-processor, interpolation and
extrapolation, and without reconstruction of the field. More focused induced SAR distributions (e.g. for
more localized sources such as capacitively-coupled sources) require a denser grid, such that the same
integration and extrapolation algorithms can be used for the same assessment uncertainty.
NOTE 2 The minimum ratio of 30 % is derived from the plane wave penetration depth at 6 GHz.
5) If one or both of step d) 4) i) and ii) are not met, the zoom scan measurement shall be
repeated using a finer resolution while keeping the other zoom scan parameters
compatible with Table 4. New horizontal and vertical grid steps shall be determined
from the measured SAR distribution so that d) 4) i) and ii) are met. Compliance with
d) 4) i) and ii) shall be demonstrated for the new measured zoom scan. The size of the
higher-resolution zoom scan and the other parameters of Table 4 shall apply. The
closest point to the phantom shell shall be 2 mm or less for graded grids, and the
grading factor shall be 1,5 or less.
6) Uncertainties due to field distortion between the media boundary and the dielectric
enclosure of the probe should also be minimized, which is achieved if the distance
between the phantom surface and physical tip of the probe is larger than the probe tip
diameter. Other methods may utilize correction procedures to compensate for
boundary effects, which can enable high precision measurements at closer than half
the probe tip diameter [22]. For all measurement points, the angle of the probe normal
to the phantom surface shall be less than 5° for flat phantoms, and less than 30° for
head phantoms. If these angles cannot be achieved, additional uncertainty evaluation
is required (see Clause O.3).
e) Use the post-processing (i.e. interpolation and extrapolation) procedures specified in 7.5
to determine psSAR values.
f) Measure the local SAR at exactly the same test point location as in step a). The SAR drift
of the DUT may be estimated by the difference between the two measured local SAR
values from steps a) and f). See 7.4.3 for more information on addressing SAR
measurement drift.
IEC/IEEE 62209-1528:2020 – 63 –
© IEC/IEEE 2020
Table 3 – Area scan parameters
DUT transmit frequency being tested

Parameter
f ≤ 3 GHz 3 GHz < f ≤ 10 GHz
Maximum distance between the measured points
(geometric centre of the sensors) and the inner a
5±1 δ ln(2)/2 ± 0,5
phantom surface (z M1 in Figure 20 in mm)
60 / f, or half of the
20, or half of the
Maximum spacing between adjacent measured corresponding zoom
corresponding zoom scan
points in mm (see O.8.3.1) b scan length,
length, whichever is smaller
whichever is smaller
Maximum angle between the probe axis and the 5° (flat phantom only) 5° (flat phantom only)
phantom surface normal (α in Figure 20) c 30° (other phantoms) 20° (other phantoms)
Tolerance in the probe angle 1° 1°
a δ is the penetration depth for a plane-wave incident normally on a planar half-space.
b See Clause O.8 on how Δx and Δy may be selected for individual area scan requirements.
c The probe angle relative to the phantom surface normal is restricted due to the degradation in the
measurement accuracy in fields with steep spatial gradients. The measurement accuracy decreases
with increasing probe angle and increasing frequency. This is the reason for the tighter probe angle
restriction at frequencies above 3 GHz.
Table 4 – Zoom scan parameters
DUT transmit frequency being tested

Parameter
f ≤ 3 GHz 3 GHz < f ≤ 10 GHz
Maximum distance between the closest
measured points and the phantom surface a
5 δ ln(2)/2
(z M1 in Figure 20 and Table 3, in mm)
Maximum angle between the probe axis and the 5° (flat phantom only) 5° (flat phantom only)
phantom surface normal (α in Figure 20) 30° (other phantoms) 20° (other phantoms)
Maximum spacing between measured points in b
8 24 / f
the x- and y-directions (Δx and Δy, in mm)
For uniform grids:
Maximum spacing between measured points in
5 10 / (f − 1)
the direction normal to the phantom shell
(Δz 1 in Figure 20, in mm)
For graded grids:
Maximum spacing between the two closest
4 12 / f
measured points in the direction normal to the
phantom shell (Δz 1 in Figure 20, in mm)
For graded grids:
Maximum incremental increase in the spacing
1,5 1,5
between measured points in the direction normal
to the phantom shell (R z = Δz 2 /Δz 1 in Figure 20)
Minimum edge length of the zoom scan volume
30 22
in the x- and y-directions (L z in O.8.3.2, in mm)
Minimum edge length of the zoom scan volume
in the direction normal to the phantom shell 30 22
(L h in O.8.3.2 in mm)
Tolerance in the probe angle 1° 1°
a δ is the penetration depth for a plane-wave incident normally on a planar half-space.
b This is the maximum spacing allowed, which might not work for all circumstances.
– 64 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
a) Head phantom
b) Flat phantom
NOTE M1 to M8 are example measurement points used for extrapolation to the surface. The maximum of the
angle α between the evaluation axis and the surface normal line is given in Table 3 and Table 4. The distance z M1
is from the phantom shell to the first measurement point M1, and its maximum value is given in Table 3 and
Table 4. The distances ∆z i (i = 1, 2, 3, …) are the distances from measurement points M i to M i+1 . For uniform grids,
all ∆z i are equal. For graded grids, ∆z i+1 > ∆z i . The term R z = ∆z i+1 /∆z i is a ratio with a maximum value given in
Table 4. The z-direction corresponds to the vertical direction, the x-direction is horizontal and the y-direction is
horizontal into the page.
Figure 20 – Orientation of the probe with respect to the line normal to the phantom
surface, for head and flat phantoms, shown at two different locations
7.4.3 Drift
a) The SAR measurement drift of the DUT is accounted for by the first and last step of the
measurement process specified in 7.4.2 by using the following two methods.
1) As the preferred method, local SAR measurements are made by the SAR
measurement system prior to performing the area scan. A second measurement is
made by the system at the same point after completion of the SAR measurement. The
measurement is done within the tissue-equivalent medium at a reference point where
the SAR values for the first and second measurements shall exceed the lower
detection limit of the measurement system. The distance from the reference point to
the inner phantom surface, in the normal direction from the inner phantom surface,
shall be less than or equal to 5 mm.
2) Alternatively, and if the preferred method in a) is not sensitive enough, conducted
power measurements may be made on the device at the antenna port using equipment
capable of measuring RF power, prior to device placement for SAR test. A second RF
power measurement shall be made after the SAR test has completed.
IEC/IEEE 62209-1528:2020 – 65 –
© IEC/IEEE 2020
b) For either of case a) 1) or a) 2), the drift is recorded as the percentage difference of the
second reference measurement, Ref secondary (SAR or conducted power), from the first
reference measurement, Ref primary , as described in Formula (6):
Ref secondary − Refprimary

drift 100 % ×
= (6)
Refprimary
The SAR drift during the measurement of the wireless device shall be within ±5 %. Some
devices could have significant fluctuations in output power that are a characteristic of the
normal operating behaviour of the device, and therefore are not classified as undesirable
power drift. In this case, other methods such as SAR scaling shall be considered to ensure
that an accurate and conservative exposure is obtained.
c) When the SAR drift cannot meet the 5 % threshold while performing the SAR testing in
accordance with 7.4.2, a drift measurement for the longest intended measurement
evaluation time shall be performed without recharging the battery. This is done by
performing a measurement in accordance with a) 1) or a) 2) continuously over the
evaluation time (at least once every 5 s). This time sweep measurement shall be
performed at each frequency band for the operating mode having the highest time-
averaged output power. When the difference between the maximum and minimum in the
time sweep is less than 5 % of the average value, or when the difference is less than 10 %
and the SAR primarily decreases during the time sweep (it does not increase by more than
2 % at any time during the time sweep), it is sufficient to perform reference measurements
at the beginning of the area scan and at end of the last zoom scan, as specified in 7.4.2 a)
and f). Otherwise, additional reference measurements shall be taken during the zoom
scan, and the zoom scan measurements shall be corrected prior to performing
extrapolation, integration, and averaging.
d) Linear interpolation between the reference measurements is performed prior to the
correction. The SAR values measured during the zoom scan shall be corrected by the
difference between the interpolated values and the first reference value measured before
the area scan. The time between reference measurements during the zoom scan shall be
sufficiently small such that correction of the time sweep curve specified in the preceding
paragraph is conservative for all points.
e) When the SAR drift is within 5 %, it can be treated either as an uncertainty (i.e. random
error) or a systematic offset. When the drift is larger than 5 %, the measurement drift shall
be considered a systematic offset rather than an uncertainty. If treated as an uncertainty,
the absolute value of the drift is recorded in the uncertainty table; also, the drift is not
added to the assessed SAR value. The uncertainty value reported in the uncertainty
budget shall either be the value corresponding to the highest SAR measurement drift
reported, or the maximum allowed (i.e. 5 %). If treated as a systematic offset, apply
compensation to the measured SAR; i.e. add the absolute difference to the determined
SAR value if the drift is either negative or positive, as described in Formula (7):
= SARmeasured × (1+ drift / 100 % )

SARcompensated (7)
In this case, it is not necessary to record the drift in the uncertainty budget (i.e. u i = 0 %).
To maintain a conservative exposure, drifts are not subtracted from the assessed SAR. If
different assessed operating modes of the device exhibit different drift ratios, all the
corresponding measured SAR values can be compensated by the same ratio, provided
that the applied drift ratio is the largest one detected during the SAR testing over all
operating modes of the device. The uncertainty shall be estimated assuming a rectangular
probability distribution
– 66 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.4.4 SAR measurements of DUTs with multiple antennas or multiple transmitters
7.4.4.1 General
DUTs with multiple antennas or multiple transmitters (with single or multiple antennas)
transmitting simultaneously require special test considerations. The methods to combine the
fields in order to determine the combined SAR distribution differ depending on whether the
corresponding RF transmitters emit waveforms that are correlated or uncorrelated in time. The
field summation method and the associated measurement instrumentation requirements for
correlated signal waveforms are different from those for uncorrelated signals (see
IEC TR 62630 [1]).
A procedure that is suitable for assessing the SAR in the head from a wired personal hands-
free headset is described in Annex M.
7.4.4.2 SAR measurements for non-correlated signals
7.4.4.2.1 General
The following procedures are applicable to devices incorporating multiple operating modes
that are intended to operate simultaneously.
a) For multiple frequencies (f 1 , f 2 , etc.) that are separated by more than the valid frequency
range of the probe calibration or the tissue-equivalent medium, whichever is the smallest,
the SAR cannot normally be assessed simultaneously using the same probe and medium.
The valid frequency range of probe calibration is typically narrow (e.g. ±50 MHz to
±100 MHz) for electric field probes in most systems currently in use. Also, because
electric field probes used in present systems typically have a DC voltage at the output, the
probe cannot distinguish between signals at different frequencies. The valid frequency
range of the tissue-equivalent medium refers to the frequency range over which the
dielectric properties are within tolerance of the target values (see Table 2 and [2]). Due to
these limitations, the SAR values shall first be assessed separately, then combined
arithmetically.
b) In the case of multiple antennas transmitting different operating modes at the same
frequency, measurements can be made with both signals transmitting simultaneously.
However, this is not necessary when the psSAR is measured separately for each
operating mode then summed as specified in Alternative 1 (see 7.4.4.2.2), because that
method provides a conservative exposure for the combined SAR. For the case of multiple
antennas transmitting correlated signals (e.g. certain MIMO configurations), see 7.4.4.3.
c) In 7.4.4.2, a test combination is designated as a particular combination of device positions
(left cheek, right tilt, etc.), device configurations (e.g. antenna position), and device
accessories (battery). Subclauses 7.4.4.2.2 through 7.4.4.2.5 specify alternative
evaluation procedures for simultaneous transmission in different frequency bands. The
procedures also apply to simultaneous transmission in the same frequency band, when
required by device test set-up conditions or test mode software configurations. The
following prerequisites apply for the alternative methods.
1) The area scan, zoom scan and psSAR are evaluated separately at each frequency (in
accordance with 7.4.2) with the transmission at that frequency turned on, and
transmission at the other frequencies turned off.
2) The SAR data from different frequencies or antennas are combined only when the test
combination is the same for those frequency bands or antennas, and the test
combination is intended for simultaneous operation.
d) Different alternative methods may be used for different test combinations. The alternatives
are summarized as follows:
1) Alternative 1: Summation of psSAR values – simplest, but most conservative method
to find an upper bound; always applicable (7.4.4.2.2).
2) Alternative 2: Selection of the highest assessed psSAR value – simple method;
applicable when SAR distributions have little to no overlap (7.4.4.2.3).
IEC/IEEE 62209-1528:2020 – 67 –
© IEC/IEEE 2020
3) Alternative 3: Calculation of combined volumetric SAR from existing area and/or
zoom scans – accurate and fast method; always applicable (7.4.4.2.4).
4) Alternative 4: Volumetric scanning – most accurate method; always applicable
(7.4.4.2.5).
e) The DUT measurement is deemed to fully comply with the requirements of this document
when it meets the requirements of one of these alternative evaluation procedures.
f) The uncertainty according to Clause 8 shall be assessed, and documented in the
measurement report.
Alternative 1 is the most conservative and computationally simple, and it requires no

additional SAR measurement. Alternatives 2 and 3 successively reduce the degree of over-
estimation; however, they require greater computation and test analysis. Alternative 4
provides the least overestimation, and requires the most effort.
7.4.4.2.2 Alternative 1: Evaluation by summation of psSAR values
This procedure is applied to determine an upper bound for the combined SAR in a
conservative manner when the same maximum output power of each transmitter or antenna is
used for both standalone and simultaneous transmission. Note that the different psSAR
values being summed can be at different spatial locations. This procedure is always
applicable, and note that it will overestimate the combined SAR. The following procedures
shall be applied using full SAR testing that complies with all of the normative requirements of
this document. Fast SAR testing may be used to identify the highest SAR test configurations
for each frequency band, as specified in 7.9.2.
a) For a test combination where simultaneous operation is intended, add the psSAR values
for each antenna and frequency band where simultaneous operation is intended (see
NOTE 1, NOTE 2, and NOTE 3).
b) Check if the maximum summed SAR value is within 3 dB of the applicable SAR limit. If so,
ensure that all the required test frequency channels in 7.2.8 have been measured in all
frequency bands and for all antennas at which simultaneous operation is intended. Repeat
steps a) and b) to determine simultaneous transmission psSAR values.
c) The maximum summed SAR value in steps a) and b) is the combined-SAR.
NOTE 1 The SAR value at each frequency band corresponds to the frequency channel tested in that band at
which the measured psSAR is the highest. In 7.2.8 the appropriate subset of frequencies to be measured for each
frequency band is specified, and procedures for measuring at fewer frequencies than this subset are provided in
7.3. For example, if the SAR has been measured at the lowest, middle, and highest channels in a band, and the
highest SAR is at the lowest frequency channel, the psSAR at the lowest frequency channel is used as the SAR
value for that frequency band. If only the SAR at the middle frequency channel has been measured, in accordance
with the procedures in 7.3, then the psSAR at the middle frequency channel is used as the SAR value for that
frequency band.
NOTE 2 An acceptable variation of 7.4.4.2.2 a) is to add the highest psSAR values applicable to all simultaneous
transmission combinations, regardless of the test configuration. To do so, the highest psSAR for one frequency
band (among all test configurations at that frequency band) is added to the highest psSAR value at the other
frequency band (among all test configurations), and so on for the other frequency bands where simultaneous
operation is intended. Each of the test combinations that is considered using this method is then evaluated in
7.4.4.2.2 b) and c). This method is more conservative than the method of 7.4.4.2.2 a).
NOTE 3 Having identified the maximum SAR test combination, it is acceptable to perform volumetric scanning on
that test combination, according to 7.4.4.2.5, to obtain a more realistic estimate of the maximum combined SAR.
The volumetric scan can be used if the test position and device configuration are the same for the different
frequency bands or antennas.
– 68 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.4.4.2.3 Alternative 2: Evaluation by selection of highest assessed maximum
psSAR values
This procedure gives an estimate of the combined SAR when the separately measured zoom
scan SAR distributions have little or no overlap. The maxima are then separated to such an
extent that the maximum psSAR value of each distribution will not increase by more than 5 %
when the SAR distributions from all the other simultaneous operating modes are added. This
alternative is only applicable if the highest psSAR is less than 70 % of the compliance limit, as
calculated from the zoom scans at each frequency. This procedure shall be applied using full
SAR testing that complies with all normative requirements of this document.
a) Measure the psSAR at each frequency separately according to 7.4.2. The area scans shall
be performed in the same plane at each frequency. The distance z M1 for all area scans
shall be less than or equal to the smallest z M1 value specified in Table 4 for the
frequencies of interest. The probe tip diameter shall comply with the requirements of 6.1 r)
at all of the frequencies of interest, and probe calibration shall comply with the
requirements in this document.
b) The separate area scans shall be interpolated such that the overlapping area has the
same grid. The resolution of the interpolated grid shall be 1 mm or better. Find the peak
value in each of the area scans.
c) The overlapping area shall contain all SAR peaks.
d) For all measured area scans, create a new SAR distribution by adding the interpolated
area scans spatially, i.e. point-by-point, within the overlapping area.
e) When the peak value in the new SAR distribution created in step c) does not exceed the
highest of the separate maximum peak SAR values found in step b) by more than 5 %, the
combined SAR is selected as the highest of the separate psSAR values calculated from
the zoom scan, as calculated in step a).
7.4.4.2.4 Alternative 3: Evaluation by calculated volumetric SAR data
This procedure uses existing area scans and zoom scans, in combination with interpolation
and extrapolation, for generation of volumetric SAR data, and it is a quick way of obtaining
the combined SAR. Different algorithms to accomplish this have been presented in [23], [24],
[25], [26], and [27]. The uncertainty of the method used shall be well documented according to
the procedures in Clause 8, and shall be recorded. This method is always applicable. This
procedure shall be applied using full SAR testing that complies with all of the normative
requirements of this document.
a) For a test combination where simultaneous operation is intended, calculate the volumetric
SAR distribution over a region corresponding to the area scan for each frequency band
where simultaneous operation is intended.
b) Add the volumetric SAR distributions of all frequency bands spatially, using interpolation
according to 7.5.1. For each frequency band where simultaneous operation is intended,
this step shall be performed for each measured frequency channel, according to the
requirements of 7.2.8 and the procedures of 7.3 (see NOTE 1 of 7.4.4.2.2).
c) Use post-processing procedures specified in 7.5 and Annex P to determine the psSAR
values from the SAR distributions of step b).
d) Check if the maximum psSAR value is within 3 dB of the compliance limit. If so, ensure
that all of the required test frequency channels in 7.2.8 have been measured in all
frequency bands at which simultaneous operation is intended, and repeat steps a) to c).
7.4.4.2.5 Alternative 4: Evaluation by volumetric scanning
This procedure is the most accurate way of assessing the combined SAR, and it is always
applicable. The SAR data are combined for each test configuration (see 3.52) where
simultaneous transmission is intended. This procedure shall be applied using full SAR testing
that complies with all of the normative requirements of this document.
IEC/IEEE 62209-1528:2020 – 69 –
© IEC/IEEE 2020
a) For a test combination where simultaneous operation is intended, ensure that the zoom
scan has been measured according to 7.4.2 at all test frequency channels specified in
7.2.8, for each frequency band at which simultaneous operation is intended.
b) Determine a volumetric grid that encompasses the zoom scans at the test frequencies
determined in step a) over all the frequency bands at which simultaneous operation is
intended. If the zoom scans at frequencies f 1 , f 2 , etc., are so far apart that the volumetric
grid is very large, resulting in very long measurement times, an acceptable variation of this
is to identify all zoom scan locations for each frequency channel in step a) then apply the
alternative procedure in step c).
c) At each frequency channel determined in step a), measure the volumetric grid found in
step b). This volumetric grid measurement adheres to all of the requirements of 7.4.2
steps c) and d), except that the volumetric grid is larger than the zoom scan. When it was
decided in step b) to use zoom scan locations instead of volumetric grid, at each
frequency channel determined in step a), measure the zoom scan for the other
frequencies at exactly the same locations as for each previously measured zoom scan in
step a). The measurement is performed with the operating mode at that frequency turned
on, and the operating modes at the other frequencies turned off.
d) Add the SAR distributions obtained in step c) spatially to obtain a summed SAR
distribution. Calculate the maximum combined SAR from the summed distribution, using
the post-processing procedures (interpolation, extrapolation, and averaging) specified in
7.5 to determine the psSAR. When volumetric scans are performed for each frequency,
these shall be summed, and the maximum peak is determined based on the total
distribution. In case only zoom scans are performed in step c), the zoom scans at each
peak location in each frequency band are combined, and the highest one is identified to
compute psSAR.
The DUT should be fixed on the phantom when the equivalent media are changed so that the
summation of the SAR distributions is as accurate as possible. If the battery of the device
needs to be recharged, the charger cable shall be attached to the DUT while it remains
positioned on the phantom. The cable shall only be attached when the battery is being
charged between SAR measurements, and shall be detached during testing.
7.4.4.2.6 Example calculation of the combined SAR using Alternative 1
Subclause 7.4.4.2.6 describes an example of the application of the summation of psSAR

values method (Alternative 1). It is for illustrative purposes only. Table 5 shows an example of
a DUT having two antennas and four operating modes, where 10 g psSAR values from single-
band measurements are shown in the third column and third row for the four operating modes
and antennas. The SAR values are shown in the third row for Antenna 1, and in the third
column for Antenna 2. The cells to the bottom and right of these values show the combined
SAR values. In this example, the cells that are not marked with "–" will be the only ones to be
evaluated. The four operating modes consist of two voice modes and two data modes, each in
two frequency bands. The 10 g psSAR values have all been measured at the centre channel
of the respective frequency bands for single-band operation (one operating mode and antenna
is transmitting, while the other band and antenna are switched off). The combined SAR values
are computed using Alternative 1 (i.e. simple summation of the 10 g psSAR values in row 3
and column 3). For this example, the DUT supports only one voice mode at a time, and one
data mode at a time, and it supports simultaneous transmission of voice and data. Therefore,
a voice mode is not combined with another voice mode, for example. Also, SAR
measurements on the left side of the head are not combined with SAR measurements on the
right side. However, it is intended for this example that the two antennas could transmit
simultaneously.
– 70 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table 5 – Example method to determine the combined SAR value using Alternative 1
Antenna 1
Voice Voice Data Data

Antenna 2 Band 1 Band 2 Band 1 Band 2
Left Right Left Right Left Right Left Right
0,285 0,250 0,333 0,315 0,512 0,489 0,593 0,574
Voice Left 0,141 – – – – 0,653 – 0,734 –

Band 1 Right 0,120 – – – – – 0,609 – 0,694
Voice Left 0,131 – – – – 0,643 0,724 –

Band 2 Right 0,130 – – – – – 0,619 – 0,704
Data Left 0,220 0,505 – 0,553 – – – – –

Band 1 Right 0,213 – 0,463 – 0,528 – – – –
Data Left 0,225 0,510 – 0,558 – – – – –

Band 2 Right 0,216 – 0,466 – 0,531 – – – –
7.4.4.3 SAR measurements for correlated signals
DUTs with multiple antennas transmitting correlated signals simultaneously represent a

particular case of devices, such as MIMO transmitters with digital beam-forming capabilities,
and require special test considerations. Signals of such DUTs can be classified based on the
change of relative phases in a typical communication session. In general, there are two types
of signals that can be found in most recent generation multi-antenna transmitters.
The first are signals with relative phases unchanged for a relatively long duration compared
with symbol duration; these are denoted as Type 1 signals for the purposes of this document.
Type 1 signals are used in phased array antenna systems, where the relative phases of
signals fed to the antennas are controlled to form the radiation pattern of the array antenna
toward a specific direction. In different operating environments, the relative phases can be
changed to obtain different desired radiation patterns. As soon as the transmitting direction is
determined and the pattern is formed, the relative phases will be fixed for a specific duration,
and will only change when the radiation pattern is configured to another form. The duration
that the relative phases are kept unchanged is usually relatively long compared with the
duration of a symbol in typical communication sessions.
On the other hand, the second type of signals are those with relative phases that vary quickly
over a relatively short period; these are denoted as Type 2 signals for the purposes of this
document. Type 2 signals can be found in systems utilizing MIMO techniques. The relative
phases of the signals will be changed from symbol to symbol, using the function of space-time
block coding (STBC) in MIMO schemes. The relative phases of signals are changed from
symbol to symbol according to the STBC, but beam-forming is not used during typical
communication sessions.
IEC/IEEE 62209-1528:2020 – 71 –
© IEC/IEEE 2020
As explained in IEC TR 62630 [1], correlated signals are those that can only be transmitted at
the same carrier frequency, and SAR depends on the relative phase(s) between the signals.
Therefore the psSAR cannot be precisely evaluated using scalar E-field probes from a single
measurement with the multiple transmitters set to fixed relative phase conditions if those
phases are subject to change during the normal operations of the device. Instead, for precise
SAR evaluation, repeated measurements corresponding to all phase combinations between
the transmitters are needed. This is a rather time consuming evaluation, and might not be
practical unless SAR systems that enable fast SAR testing are used, combined with software
control of the DUT to cycle through all possible phase combinations of the signals transmitted
simultaneously at their highest time-averaged output powers. In general, it is also possible to
accurately evaluate psSAR of each individual transmitter transmitting separately at the
highest time-averaged output power using single SAR measurements; however this approach
requires complex vector E-field measurements (i.e. measurements of magnitude and phase of
all three E-field components), and is therefore less practical.
The alternative method utilizing conventional SAR systems is based on SAR measurements
for each transmitter transmitting separately at the highest time-averaged output power, then
combining the individual SAR results as described in IEC TR 62630 [1]. This approach leads
to much faster SAR measurement times, but provides only an upper bound of the SAR, thus
potentially overestimating the results. IEC TR 62630 [1] describes two methods of combining
the SAR from individual measurements using conventional scalar E-field probes.
The two methods described in IEC TR 62630 [1] can be implemented using conventional SAR
measurement systems, and require only a limited number of SAR scans equal to the number
of transmitters. The first method is based on combining the magnitudes of the E-field values,
and the second is based on magnitudes of the individual E-field components. Especially for
the Type 1 signals or unspecified signals, the second approach based on combining the
individual E-field components should be used, because it leads to lower potential SAR
overestimation, and many available SAR systems readily provide the required input data for
the post-processing described in [1]. The measurement procedure for the two types of
correlated signals is described in Figure 21. For Type 2 signals, the approach of time-
averaged SAR measurements (see NOTE) should be used, which requires only use of the
measurement procedure specified in 7.4, with conventional scalar probes.
NOTE Another method is to measure time-averaged SAR in a specific averaging period using conventional scalar
SAR measurement systems. This method can only be applicable to the devices generating Type 2 correlated
signals. A signal property is assumed such that the combination of relative phases that produces the maximum
SAR occurs only infrequently, and when it does occur it is in a very short time before being dominated by other
phase combinations that produce lower SAR. Because SAR is determined by electromagnetic (EM) energy
absorbed in the exposed tissue-equivalent medium, introducing time-averaging procedure is suitable for the
evaluation of SAR with such rapidly fluctuating EM fields. To evaluate the time-averaged SAR from such devices,
all transmitting antennas are turned on as they are during normal communications at the same time, and
instantaneous SAR values are measured at each measurement point within a sampling-time interval, then averaged
over a time-duration appropriate for capturing the signal characteristics.
This method has the benefit of using conventional scalar SAR probes, and furthermore, provides measured SAR
that is close to actual SAR. Also the method requires only a single time-averaging measurement at each
measurement point in accordance with the procedure of 7.4, regardless of the number of transmitting antennas.
The averaging time for a multi-antenna system, however, is longer than that for conventional single-transmitting
antenna measurements. Also, when the number of antennas increases, the averaging time would be slightly longer
to obtain converged SAR results, because the time-fluctuation of the instantaneous SAR increases along with the
increase in the number of antennas. The averaging time also depends on the probe sampling rates of the SAR
measurement systems. The averaging time is reduced as the sampling rate is increased, and vice versa. For
example, with a SAR measurement system having a measurement rate of 1 250 samples per second, the averaging
time of 1 s provides a standard deviation of the instantaneous SAR values due to multiple transmitting antennas
that is kept under 2 %. In such a case, total measurement uncertainty evaluated by using typical values and the
uncertainty budget in Clause 8, the uncertainty for multiple transmitting antennas results is of the same level
compared with that of single antenna cases (typically at 8,7 %).
– 72 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure 21 – Measurement procedure for different types of correlated signals
7.5 Post-processing of SAR measurement data
7.5.1 Interpolation
The resolution of the measurement grid is not as fine as what is required by the interpolation
requirements for the area and zoom scan to compute the psSAR. Interpolation shall be
applied to the measurement points. Examples of interpolation schemes are given in Annex P.
7.5.2 Extrapolation
The electric field probes used to measure SAR usually contain three orthogonal dipoles in
close proximity and embedded in a protective cover/case. The measurement (calibration)
point is situated a few millimetres from the tip of the probe and this offset shall be taken into
account when identifying the position of the measured SAR. Because of probe boundary
effect errors and probe sensor offset, SAR is not measured at the surface of the phantom.
The measured points closest to the surface are extrapolated to estimate the highest SAR at
the phantom surface. Examples of extrapolation schemes are given in Annex P.
7.5.3 Definition of the averaging volume
The averaging volume shall be in the shape of a cube with side dimension that establishes a
1 g or 10 g mass. The side length of the 1 g cube shall be 10 mm, and the side length of the
10 g cube shall be 21,5 mm. The orientation of the cubical volume is given in Annex P.
IEC/IEEE 62209-1528:2020 – 73 –
© IEC/IEEE 2020
7.5.4 Searching for the maxima
The cubical averaging volume shall be shifted throughout the interpolated and extrapolated
zoom-scan volume at the inner surface of the phantom in the vicinity of the local maximum
SAR, using considerations such as those given in Annex P. The cube with the highest psSAR
shall not be at the edge/perimeter of the zoom-scan volume. In case it is, the zoom-scan
volume shall be shifted and the measurements shall be repeated.
7.6 Time-period averaged SAR considerations
7.6.1 General
Subclause 7.6 specifies the test procedures for assessing the time-period averaged SAR
(TPAS) when the DUT uses time-period power control. The applicable time-averaging period
is specified in international guidelines (e.g. ICNIRP) or national regulations. This test
procedure differs from the general SAR test procedures specified in 7.4 that are applicable for
DUTs that do not have power control algorithms. The rationale for assessment of TPAS is
given in Annex Q.
NOTE Some national SAR regulations might have varying definitions of time averaging requirements for their SAR
limit.
A procedure and its requirements for assessing the TPAS are provided in 7.6.3 to 7.6.7. The
procedure may be applied if the following requirements are met.
a) The power control algorithms in the DUT shall not be alterable or accessible by the user.
b) The DUT manufacturer shall provide sufficient documentation about the power control
algorithms and their implementation such that the measurements result in a conservative
exposure.
If the preceding requirements are not met, the SAR shall be assessed at the maximum output
power level as described in 7.4.
Alternative procedures to 7.6.2 to 7.6.7 may be applied if justified and documented.
7.6.2 RF conducted power
RF conducted power measurements may be used to determine TX factors.
The power detection, integration time, and sample averaging settings of the power meter
ensure that the uncertainty component for TPAS is within the assessed value in 7.6.8.
Conducted power measurements shall also be accurate for the respective operating mode.
Linear (e.g. milliwatt) readings shall be used.
7.6.3 Time-period averaged SAR measurement settings for SAR measurement

methods
7.6.3.1 General
Subclause 7.6.3 provides requirements for TPAS measurement, and for calculation of the TX
factor for periods up to six minutes.
7.6.3.2 Array SAR systems
For Class 2 fast SAR testing (see 7.9.2), the measurement of SAR time-average shall be
performed over the averaging period, and the measurement of SAR full-power shall be performed
over a time period long enough that the measurement is stable. Neither the DUT nor the
sensors shall be moved during the averaging period. The sampling rate of the measurement
shall be enough to ensure that the uncertainty component for TPAS is within the assessed
value in 7.6.8.
– 74 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.6.3.3 Scanning SAR systems
For scanning probe SAR systems, including single probe SAR systems and Class 1 fast SAR
testing systems (see 7.9.2), the following apply.
a) Both SAR full-power and SAR time-average are local SAR measurements.
b) The probe shall be positioned at the location of the peak SAR from an area scan
performed in accordance with the requirements of 7.4.2.
c) The SAR level at this location shall be above the noise level.
d) The probe or probe array shall be in the same position for both the SAR full-power and
SAR time-average measurements.
e) SAR time-average shall be measured over the defined time period.
f) SAR full-power shall be measured over a time period long enough that the measurement is
stable.
g) The sampling rate of the measurement shall be enough to ensure that the uncertainty
component for TPAS is within the assessed value in 7.6.8.
7.6.4 Exposure condition and test position considerations
TPAS shall be assessed separately for each frequency band and operating mode, except
when it can be demonstrated that the same TX factor can be applied to multiple frequency
bands and/or wireless operating modes. Two approaches are considered, depending on
whether the device is capable of detecting and distinguishing its usage position reliably (e.g.
detecting head and body-worn positions according to audio routing to the earpiece of a device
during a voice call).
a) If the different usage positions are not detected and distinguished by the device, the
maximum TPAS among all applicable test/use positions and conditions, including all head,
body-worn, and other positions, are used to determine compliance. This is the most
conservative and simple approach that applies to all use conditions.
b) Otherwise, the SAR test positions may be grouped as follows.
1) Head test positions: The maximum TPAS among all head test positions (left cheek, left
tilt, right cheek, and right tilt), corresponding to voice calls via the earpiece, is used to
determine head exposure SAR compliance.
2) Body-worn and other test positions: The maximum TPAS among all applicable body-
worn and other test positions, based on the applicable separation distances for the
corresponding use positions, when there is no voice call via the earpiece, is used to
determine non-head SAR compliance.
7.6.5 Time-period averaged SAR for simultaneous transmission
Due to the complexity of combining simultaneous transmission with TPAS, the following
requirements apply to determine the highest psSAR:
a) the TX factor assessment in 7.6.6 shall be applied to a maximum of one of the

simultaneous transmissions;
b) all simultaneous transmissions aside from the one in a) shall be assessed at the maximum
output power;
c) the combined SAR of all simultaneous transmitters shall be assessed according to 7.4.4.
7.6.6 TX factor assessment
The TX factor is used to scale the measured SAR at the maximum output power (see 7.6.7)
down to the time-period averaged SAR. It is separately applied for each operating mode
(e.g. GSM, WCDMA, LTE, 5G NR), and for each applicable frequency band, except when it
can be demonstrated that the same TX factor can be applied to multiple frequency bands
and/or wireless operating modes.
IEC/IEEE 62209-1528:2020 – 75 –
© IEC/IEEE 2020
The following steps shall be performed to determine TX factors.
a) The power control algorithm shall be initially disabled.

b) The DUT shall be under the control of a radio communication test set (call-box).
c) The radio communication test set shall be set to request maximum TX power and duty
factor from the DUT according to wireless technology requirements.
d) When using a SAR measurement system for TX factor assessment, the DUT shall be
positioned in accordance with 7.6.4.
e) Measure the conducted power (see 7.6.2) or local SAR. The conducted power or SAR data
shall be recorded and documented in the measurement report. This value shall be noted
as P full-power or SAR full-power , respectively for power or SAR measurement.
f) Enable the power control algorithm. Ensure that the DUT position against the phantom is
maintained during this step.
g) Measure the conducted power or local SAR over at least the time-averaging period.
Record the average conducted power or SAR over the time-averaging period. This value
shall be noted as P time-average or SAR time-average .
h) The TX factor is P time-average / P full-power or SAR time-average / SAR full-power .
7.6.7 SAR measurements
SAR measurement shall be performed for all required test positions and conditions as
specified in 7.2.4. The power control algorithms in the DUT shall be disabled during these
SAR measurements.
There are two options for performing the SAR measurements.
– Option 1: During the SAR measurements, transmit power shall be at the maximum
specified for the DUT (7.2.3.1). The TX factors obtained in step h) of 7.6.6 shall be applied
to the maximum full power SAR results obtained, i.e. be multiplied by the TX factor. These
calculated TPAS results are compared directly against the applicable SAR limit to
determine compliance.
– Option 2: Transmit power shall be limited to the time-period averaged power level
P time-average , determined in step h) of 7.6.6, in order that the measured SAR corresponds
to the TX factor. When TX factors are determined using local SAR measurement in step h)
of 7.6.6, and it is demonstrated that SAR and power are linearly scalable for the wireless
configuration used in the DUT (7.2.3.2), the time-period averaged power level P time-average
is determined by scaling the maximum power (7.2.3.1) used to establish the TX factor.
These TPAS results are compared directly against the applicable SAR limit to determine
compliance.
7.6.8 Uncertainty in TPAS evaluations
The uncertainty related to TPAS calculation shall be assessed according to general guidance
given in Clause 8, and documented in the measurement report.
The additional uncertainty related to TPAS assessment is due to the sampling rate of the
measurement. The TPAS uncertainty can be neglected depending on method selected to
determine the TX factor (i.e. using either 7.6.3 SAR measurements or 7.6.2 conducted power
measurements).
The TPAS uncertainty is different from signal modulation uncertainty (MOD) and data
acquisition errors (DAE) specified in 8.4.
– 76 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
If the combined uncertainty of the SAR measurement method and the time-averaged
calculation (either SAR based or conducted-power based) exceeds the upper uncertainty limit
specified in this document, a scaling factor (U TPAS − U SAR ) shall be applied to the
measurement result, where U TPAS is uncertainty for TPAS assessment, and U SAR is the
uncertainty for SAR measurements at full power and maximum duty cycle. The measurement
system manufacturer shall be consulted for details of the uncertainty calculation.
7.7 Proximity sensors considerations
7.7.1 General
Proximity sensors are used to reduce the output power of the DUT based on the distance to
users and nearby objects. When the device is operating at closer than the triggering distance
of the proximity sensor, the maximum output power of a DUT is reduced to ensure SAR
compliance. At distances larger than the sensor triggering distance, the proximity sensor is
disabled and the maximum output power is restored. Depending on the antenna dimension
and its location, a DUT could incorporate one or more proximity sensors to ensure sensor
coverage.
The procedure specified in 7.7 only applies to proximity sensors designed to cover a single
transmitting antenna. The procedures apply to a single sensor or to multiple sensors where
the sensor responses are combined to provide a single unique and unconditional triggering.
The procedure is not intended for multiple sensors that trigger more than one antenna.
However, the procedure may be applied separately to multiple antennas, provided each
antenna is triggered independently.
NOTE 1 The procedure is not intended for conditional triggering for different use conditions. Neither does the
procedure cover simultaneous transmission.
NOTE 2 The procedure specified in 7.7 only applies to single-step power reduction with one proximity sensor
triggering distance. If multi-step power reduction is applied using multiple separation distances, regulatory
authorities can be consulted to determine the testing procedures and requirements.
NOTE 3 The procedure specified in 7.7 has been established for devices used next to the body, where the flat
phantom is applicable for SAR testing. The procedure does not consider test cases on the head or application-
specific phantoms described in Annex K, where changes in body contours would require more complex test
considerations to validate proximity sensor operating characteristics.
NOTE 4 DUTs typically ensure that if the proximity sensor is not functional then they will default to reduced output
power.
Subclause 7.7 shall be used to determine the triggering distance and coverage area of
proximity sensors to ensure that:
a) power reduction is triggered consistently and reliably at the specified distance from the
DUT to the user (7.7.2);
b) the proximity sensor covers a sufficient area or region surrounding the peak SAR location
to ensure consistent triggering from all applicable DUT orientations with respect to the
user (7.7.3);
c) the test separation distance for SAR testing at full power is determined. This distance is
the smaller of the triggering distance for perpendicular DUT orientation (7.7.2.2) and the
triggering distance for tilted DUT orientation (7.7.2.3) minus 1 mm. The minus 1 mm
spacing is generally required to ensure all production units are compliant. If more
production variation in triggering distance is expected, it shall be taken into account
accordingly.
d) any hysteresis effects in the triggering distance are accounted for.
The procedures shall be conducted separately for all antennas implementing proximity
sensor-based power reduction. Each implemented proximity sensor sensitivity setting shall be
tested once using one of the corresponding operating mode and frequency band
combinations. If the implemented proximity sensor sensitivity settings are not known, the
procedures shall be performed for all operating mode and frequency band combinations.
IEC/IEEE 62209-1528:2020 – 77 –
© IEC/IEEE 2020
Capacitive sensors are typically used for proximity sensing. The presence of approaching
nearby objects in the vicinity of the sensor can be detected by changes in sensor capacitance.
Depending on the antenna dimension and extent of the SAR distribution, one or more sensors
could be required to ensure adequate sensor coverage for the corresponding SAR
distribution. The procedures for determining sensor coverage are specified in 7.7.3.
NOTE 5 Proximity sensors can use techniques other than capacitive sensing. The procedures in 7.7 have been
written for capacitive sensors only, which will require adaptation or might not be suitable for testing other sensing
technologies. As appropriate, the regulatory authorities can be consulted to determine the testing and validation
requirements for other sensing techniques.
7.7.2 Procedures for determining proximity sensor triggering distances
7.7.2.1 General
Proximity sensors are typically incorporated on the flat surfaces or along edges of a device
where the antenna is located so that the triggering of output power reduction can
appropriately limit the SAR. The procedures in 7.7.2 are used to meet the following
requirements.
a) The minimum triggering distance shall be determined and used to establish a test
separation distance for the SAR compliance measurements at full power, when power
reduction is inactive.
b) The minimum triggering distance shall account for variations due to hysteresis and
variations across production units due to implementation tolerances.
c) For sensors that are located along the edge of a device, the minimum triggering distance
shall cover the various angles and orientations that the edge can be positioned relative to
the user to determine the test separation distance for SAR measurement.
d) All proximity sensors in the DUT shall be tested in actual use conditions as if the device
was normally operated by the user, including multiple proximity sensors used to cover an
extended SAR distribution of an antenna. Any interactions between the proximity sensors
shall be taken into account during the testing.
e) When proximity sensors are required to cover different use conditions – for example,
sensors that wrap around the edge of a device to support both edge and surface detection
– all applicable use conditions shall be tested.
f) Each antenna that is triggered to lower its output power shall be tested separately.
Procedures shall be conducted for all operating mode and frequency band combinations
that apply to the antenna.
g) The phantoms and tissue-equivalent media specified in this document shall be used to
determine the minimum triggering distance of the sensor. The dielectric properties of the
tissue-equivalent medium shall be within the tolerances specified in Table 2 at the
frequency band of interest for each proximity sensor test.
h) The full and reduced maximum output power corresponding to triggering disabled and
enabled conditions shall be verified. This may be done either by using built-in test
software, by measuring the conducted power, or by measuring the radiated power. When
test software is used to identify triggering, conducted or radiated power measurement
shall be performed separately to verify the actual power levels. Conducted power
measurement is preferred over radiated power measurements.
i) When conducted power measurement is used, the losses of the RF cables shall be
accounted for in order to verify the maximum power required for SAR measurement in the
corresponding triggering states.
j) When radiated power measurement is used, the maximum output power levels for the
triggering conditions shall be identified unambiguously by taking into account the effects of
RF propagation and environmental perturbation. Local SAR at the peak SAR location may
be recorded to identify the change in triggering condition associated with the step change
in device output power. The measured local SAR can be correlated with that at the same
location in the 1 g or 10 g SAR compliance measurement.
– 78 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
k) Any cables and connectors used for the conducted power measurement shall be
configured and positioned such that neither the sensing characteristics of the proximity
detection nor the SAR levels and distributions (for local SAR) are perturbed compared to
without such cables and connectors.
l) The SAR test report shall include tabulated data of the measured output power and sensor
states versus distance of the DUT to the phantom.
7.7.2.2 Triggering distance for perpendicular DUT orientation
The procedure for determining the triggering distance of proximity sensors (e.g. single
proximity sensor or combination of proximity sensors as specified in 7.7.2.1) incorporated on
the flat surface or edge of a device, approaching the phantom (user) perpendicularly is
covered in 7.7.2.2. The variation in triggering distance due to tilted orientations of the DUT is
covered in 7.7.2.3.
For normal approach of the DUT to the phantom, the procedure is as follows.
a) Each antenna within the DUT that requires proximity sensing shall be tested in the
applicable frequency bands using a flat phantom specified in 6.2.4.
b) The surface or edge of the DUT that contains the sensor shall be positioned
perpendicularly to the flat phantom, at an initial distance that is at least 10 mm beyond the
distance where the proximity sensor is triggered.
NOTE This triggering distance procedure is only applicable to SAR measurements in a flat phantom. This
generally does not apply to the SAM phantom or other hand-held use conditions.
c) The surface or edge of the DUT shall be moved toward the phantom in step sizes of 5 mm
until the maximum output power is reduced.
d) The surface or edge of the DUT shall be moved away from the phantom by at least 5 mm
and until the maximum output power is returned to the higher level.
e) The surface or edge of the DUT shall again be moved toward the phantom, but in 1 mm
steps until it is at least 5 mm past the triggering point or touching the phantom.
f) If the DUT is not touching the phantom, it shall be moved in step sizes of 3 mm until it is
touching the phantom to confirm that the sensor remains triggered and maximum power
stays reduced.
g) The DUT shall be moved away from the phantom, using the same step sizes as those in
steps c) to f), to determine the distance when triggering is released. The movement should
continue until it is at least 50 mm beyond the point that triggers the return of normal
maximum power.
h) The smallest separation distance determined by the sensor triggering procedure for
movements to and from the phantom shall be the minimum triggering distance for
perpendicular movement of the corresponding surface or edge towards the user.
IEC/IEEE 62209-1528:2020 – 79 –
© IEC/IEEE 2020
Figure 22 – Positioning of the surfaces and edges of the DUT

for determining the proximity sensor triggering distance
7.7.2.3 Triggering distance for tilted DUT orientation
For sensors located on the edge of a device, in addition to perpendicular triggering, influence
of the DUT edge to the triggering distance shall also be evaluated with the device edge
positioned at different angles from normal to the phantom (Figure 22). The procedure to
determine triggering variations due to the edge of a device oriented at various angles to the
phantom is as follows.
a) Position the DUT edge at the minimum triggering distance determined in 7.7.2.2.
b) Ensure that the proximity sensor is triggered when the DUT is in the perpendicular
orientation specified in 7.7.2.2.
c) Rotate the DUT as shown in Figure 23. The angle shall vary by ±40° from the
perpendicular orientation in increments of 10°. At each of these angles, ensure that the
proximity sensor is still triggered.
d) Reduce the edge distance to the phantom by 1 mm and go back to step b) if sensor
triggering is released or normal maximum output power is restored in step c).
e) The smallest separation distance determined in steps a) and b) is the sensor triggering
distance for non-perpendicular edge triggering conditions.
f) The smallest triggering distance determined for the edge in perpendicular and rotated
conditions shall be the minimum triggering distance for that edge.
– 80 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure 23 – Positioning of the edges of the DUT to determine

proximity sensor triggering distance variations with the edge
positioned at different angles from the perpendicular position
7.7.3 Procedure for determining proximity sensor coverage area
This procedure verifies that the coverage area of the proximity sensors contains the
applicable psSAR cube. This ensures that the sensor provides adequate triggering relevant to
the recorded SAR at the reduced power level.
Proximity sensors cover an area or region surrounding the peak SAR location associated with
the SAR distribution of an antenna. Depending on the implementation, the sensor can be co-
located or spatially offset from the antenna.
The surface(s) or edge(s) where the sensor coverage area test is applied are determined as
follows.
– All surfaces and edges that require proximity sensing to trigger power reduction shall be
tested, except when the antenna and sensor are both co-located with the psSAR cube.
– Two edges of the DUT shall be tested if the sensor or applicable antenna is located within
50 mm of the corner of the two adjacent edges and testing of both edges is required in
7.2.4.1.
NOTE 50 mm is considered sufficiently far away from the corner so that the antenna/sensor is not influenced
by the adjacent edge; 50 mm also corresponds to the typical size of the significant portions of the SAR
distribution for antennas in devices for which the procedure outlined in 7.7.3 has been developed.
If the procedure to determine sensor coverage area is amended (i.e. the procedure is
inadequate due to concerns relating to antenna placement, extent of the SAR distribution or
sensor offset), any amendment shall be documented.
Depending on the SAR distribution of the antenna and sensor offset, if reducing the triggering
distance extends the coverage area to include the peak SAR location, the smaller triggering
distance may be considered to avoid sensor coverage problems.
The test of the coverage area of the proximity sensor is as follows.
a) Orient the DUT such that the applicable surface or edge (determined above) is in contact
with the flat phantom. Align the long axis of the DUT surface or edge with the long axis of
the phantom.
b) Move the DUT away from the phantom by the minimum triggering distance determined in
7.7.2.2. The surface or edge shall be maintained parallel to the flat phantom shell.
IEC/IEEE 62209-1528:2020 – 81 –
© IEC/IEEE 2020
c) Establish the initial DUT position by moving the DUT lateral to the phantom until both the
antenna and proximity sensor are outside the perimeter of the phantom. The surface or
edge shall be maintained parallel to the flat phantom shell. If the precise location of the
antenna or proximity sensor is not known, ensure that the entire surface or edge is outside
the perimeter of the phantom.
d) Ensure that the proximity sensor is not triggered in the initial DUT position. If it is, check
the set-up to ensure that other objects in the vicinity are not triggering the proximity
sensor.
e) Move the DUT lateral to the phantom until the proximity sensor is triggered. The surface
or edge shall be maintained parallel to the flat phantom shell. Record this DUT location
with respect to the phantom edge.
f) Rotate the DUT 90°. The axis of rotation is normal to the flat phantom surface.
g) Move the DUT back to the initial position in step c).
h) Repeat steps d) and e). Record the location of the DUT where the proximity sensor is
triggered.
i) Repeat steps f), g), and h) so that the triggering is tested at 0°, 90°, 180°, and 270°
orientations of the applicable surface or edge. Four DUT locations (one for each
orientation) are recorded.
j) Determine the coverage area as the region between the four DUT locations recorded
above.
k) Determine the location of the psSAR with respect to the DUT surface or edge.
Ensure that the psSAR cube is fully contained within the coverage area. If this is not the
case, the reduced power SAR test in 7.7.4 shall not be applied without additional
justification.
7.7.4 SAR measurement procedure involving proximity sensors
Two different maximum output power levels are applied according to the triggering conditions
of the proximity sensor. SAR measurements shall be performed for the two different maximum
output power state and test distance combinations.
a) Full power: SAR tests shall be performed at the test separation distance corresponding to
the minimum sensor triggering distance among all triggering and coverage tests of 7.7,
minus 1 mm, as specified in 7.7.2 and 7.7.3.
b) Reduced power: SAR tests shall be performed at the closest intended use distance or at
the closest distance required by the regulator.
SAR measurements at these two power and distance combinations are enough to ensure
compliance for the use conditions requiring proximity sensing and power reduction at the
applicable device to user distance.
The test set-up photos, required power measurement, and triggering distance results, for
supporting sensor triggering and sensor coverage, shall be documented in the measurement
report.
7.8 SAR correction for deviations of complex permittivity from targets
7.8.1 General
In this document, dielectric properties of the tissue-equivalent medium used for SAR
measurement are selected to give conservative exposure. Deviations of the dielectric
properties from the target values can lead to measurement uncertainty. One way to reduce
measurement uncertainty is to keep the dielectric properties of the tissue-equivalent media
within a tight tolerance of the targets (e.g. within ±5 %). However, it can be difficult to find
suitable and stable recipes with dielectric properties that are close to the targets, particularly
at frequencies above 2 GHz. There are three solutions to this problem:
a) modify the target dielectric properties to match those of available recipes;

– 82 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
b) widen the tolerance (without correcting the SAR for the deviation in dielectric properties);
c) allow a wider tolerance, and correct the SAR for the deviation of the measured dielectric
properties from the target values.
The third solution is the best, because changing the targets restricts the applicability of the
document to particular recipes, and simply widening the tolerance increases the measurement
uncertainty.
The methodology used to determine the SAR correction is specified in 7.8.2. The
methodology was conducted over a frequency range of 4 MHz to 10 GHz. The methodology
was also studied for permittivity and conductivity ranges of ±20 % from the target values in
Table 2, but ranges of ±10 % have been selected for this document, as specified below. Given
that the change in dielectric properties influences the conversion factor of the probe, this
influence will be small if a ±10 % range is used.
7.8.2 SAR correction formula
From Douglas et al. ([28], [29]), a linear relationship was found between the percentage
change in SAR (denoted ΔSAR) and the percentage change in the permittivity and
conductivity from the target values in Table 2 (denoted Δε r and Δσ, respectively). This linear
relationship agrees with the results of Kuster and Balzano [30] and Bit-Babik et al. [31]. The
relationship is given by:
∆SAR = cε ∆εr + cσ ∆σ (8)
where
cε = ∂ ( ΔSAR ) ∂ ( Δε ) is the coefficient representing the sensitivity of SAR to permittivity

where SAR is normalized to output power;
cσ = ∂ ( ΔSAR ) ∂ ( Δσ ) is the coefficient representing the sensitivity of SAR to conductivity,
where SAR is normalized to output power.
The values of c ε and c σ have a simple relationship with frequency that can be described using
polynomial equations. For dipole antennas at frequencies from 4 MHz to 6 GHz, the 1 g
averaged SAR c ε and c σ are given by
−7, 854 × 10−4 f 3 + 9, 402 × 10−3 f 2 − 2, 742 × 10−2 f − 0, 202 6

cε = (9)
cσ = 9, 804 × 10−3 f 3 − 8, 661× 10−2 f 2 + 2, 981× 10−2 f + 0, 782 9 (10)
where f is the frequency in GHz. Above 6 GHz, the sensitivity is non-varying with frequency
due to the small penetration depth; the values of c ε = −0,198 and c σ = 0 shall be used.
For frequencies from 4 MHz to 6 GHz, the 10 g averaged SAR c ε and c σ are given by:
cε = 3, 456 × 10−3 f 3 − 3, 531× 10−2 f 2 + 7, 675 × 10−2 f − 0,186 0 (11)
cσ = 4, 479 × 10−3 f 3 − 1, 586 × 10−2 f 2 − 0,197 2 f + 0, 771 7 (12)

IEC/IEEE 62209-1528:2020 – 83 –
© IEC/IEEE 2020
where f is the frequency in GHz. Above 6 GHz, the sensitivity is non-varying with frequency
due to the small penetration depth; the values of c ε = −0,250 and c σ = 0 shall be used. For
specific antennas and frequencies, the procedure of O.9.5 may be applied to determine the
sensitivity coefficients, as appropriate.
7.8.3 Uncertainty of the correction formula
The mean power uncertainty of the formulas in 7.8.2, specified in Douglas et al. [28] as the
RMS error between the SAR deviation predicted by the formulas and the simulated deviation
over 440 analysed cases, is shown in Table 6 for the 1 g psSAR and 10 g psSAR. Table 6
shows how the mean power error increases as the maximum allowable value of Δε r and Δσ
increases. It has also been shown in [28] that these corrections are valid for realistic DUT
models.
Using this approach, the measurement uncertainty is lower, because this correction eliminates
the need for uncertainty items that account for the deviation of the dielectric properties from
the targets. Instead, there is an uncertainty item which accounts for the error of the correction
formula. The value of this uncertainty item is given in Table 6. For ±10 % deviation in
permittivity and conductivity, enter 1,9 % and 1,6 % in the uncertainty budget for 1 g and 10 g
psSAR, respectively. These uncertainty values should be entered into the appropriate row of
Table 9 where a normal probability distribution is assumed.
Table 6 – Root-mean-squared error SAR correction formula as a

function of the maximum change in permittivity or conductivity [28]
Max. change in RMS uncertainty RMS uncertainty

ε r or σ for SAR 1g % for SAR 10g %
±5 % 1,2 0,97
±10 % 1,9 1,6
7.9 Minimization of testing time
7.9.1 General
Subclause 7.9 gives methods for minimizing the test time to identify a subset of the highest
SAR conditions for a DUT. Full SAR testing, i.e. testing that is fully compliant with the
requirements of 7.4 of this document, shall be performed for the subset of device
configurations identified as having the highest SAR values from these tests. One of two
methods for reducing the overall test time for a DUT may be used:
a) test reduction methods in which a set of rules are applied to reduce the total number of
device configurations that require testing specified in 7.9.3;
b) fast SAR testing that reduces the measurement time for each device configuration as
specified in 7.9.2.
Test reduction methods shall not be used in conjunction with fast SAR testing.
NOTE Studies are ongoing for a future revision of this document to check whether test reduction methods and
fast SAR testing may be used in conjunction with each other, while still providing that the reported SAR is
conservative.
– 84 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.9.2 Fast SAR testing
7.9.2.1 General
Full SAR testing uses measurement systems and procedures that fully comply with all the
requirements of this document are referred to herein as "full SAR" systems (see also 3.20).
The procedures specified in 7.9.2 apply to fast SAR testing. Fast SAR testing differs from full
SAR testing in that it uses special techniques, methods, algorithms, or different measurement
equipment that do not comply with all the normative requirements of this document, but
decreases the measurement time (see also 3.17).
Fast SAR testing may be used together with full SAR testing to reduce the overall
measurement time needed to evaluate a wireless communication device. Fast SAR testing
may be used to identify the highest SAR test configurations, and is therefore used to evaluate
relative SAR levels. Full SAR testing shall be used to assess the absolute SAR levels. Full
SAR testing shall be performed on at least the highest SAR test configuration, and those
within the confidence levels given in the procedures specified in 7.9.2.2 and 7.9.2.3.
NOTE 1 These testing alternatives are not mandatory; the user of this document remains free to apply the full
SAR testing procedure specified in 7.1 through 7.5 for all testing.
Fast SAR testing does not fall in the category of test reduction methods (see 7.9.3). Fast SAR
testing generally aims at decreasing the time spent to perform a measurement, whereas test
reduction methods aim to reduce the total number of measurements to be performed.
Due to the decreased time needed to perform fast SAR testing, it can be anticipated that
multiple fast SAR testing methods may be performed using a single battery charge when
testing a device. The requirements of 7.4.3 shall be satisfied regarding the power drift
remaining within ±5 % from the average value throughout the measurement interval.
The following procedures do not address all forms of multiple transmitter SAR evaluations.
Additional full SAR testing can be necessary to meet the requirements outlined in 7.4.4.
Two classes of fast SAR testing are designated for the purposes of this document.
a) Class 1: These are fast SAR testing methods based on measurements made using
equipment that conforms with all of the normative clauses of this document but where
special techniques are applied to improve the measurement speed; for example by
measuring reduced numbers of points then using additional post-processing. Class 1 fast
SAR testing is based on scanning and post-processing requirements of the individual fast
SAR testing method used to reduce measurement time, but with less accuracy or higher
measurement uncertainty than for full SAR testing. The measurement uncertainty shall be
determined according to Clause 8 and Annex C, and shall be documented. Each individual
fast SAR testing method shall be fully validated to determine its applicability for testing
different wireless communication devices, as specified in A.4.2. The valid ranges of
frequencies, SAR distributions, and SAR levels applicable to the fast SAR testing method
shall be documented, as specified in A.4.2. The users of a fast SAR testing method shall
perform a system check to demonstrate the reliability and consistency of the method, as
specified in A.4.3.
IEC/IEEE 62209-1528:2020 – 85 –
© IEC/IEEE 2020
b) Class 2: These are fast SAR testing methods based on measurements made using
equipment that does not conform with all of the normative clauses of this document.
Scanning systems or array systems that employ field-reconstruction algorithms fall into
this class. Class 2 fast SAR testing systems shall meet the SAR measurement system
verification and validation requirements of IEC 62209-3, with the exception that the
measurement uncertainty is not limited to a threshold of 30 %, and the validation as well
as the system verification may also be reduced to the ranges of operating modes and
frequencies tested. When used in conjunction with the fast SAR measurement Procedure
A (see 7.9.2.3), Class 2 fast SAR testing can effectively identify the highest SAR test
configurations that shall subsequently be measured with full SAR testing to assess the
maximum peak-spatial SAR that would occur in the head or body (e.g. for compliance
demonstration purposes). Like with Class 1 fast SAR testing, Class 2 fast SAR testing
systems shall be calibrated and fully validated to determine their applicability for testing
different wireless communication devices for the specific communication technologies,
exposure conditions, and device configurations being assessed for SAR. The
measurement uncertainty shall be determined and documented (see Annex C).
Requirements for vector measurement-based systems are given in IEC 62209-3. It is

recommended to apply its requirements to this document.
NOTE 2 Field reconstruction techniques are specified in IEC 62209-3.
7.9.2.2 Fast SAR testing Procedure A
Procedure A may be applied to determine the maximum SAR of a DUT for all the required test
configurations, or a subset thereof. Class 1 fast SAR testing and Class 2 fast SAR testing
may use Procedure A. This procedure shall be applied separately for head and body phantom
types. For the head phantom type, the left head and right head data can be considered as one
data set. For the body phantom type, all body phantom configurations (e.g. if multiple
phantom sizes were used) can be considered as one data set.
NOTE 1 An extensive inter-laboratory study was carried out to assess the robustness of fast SAR testing
Procedure A. In this study, supported by IEC TC 106 and organized by Innovation, Science and Economic
Development (ISED) Canada, 10 laboratories performed over 7 500 measurements using full SAR testing and
Class 1 and Class 2 fast SAR testing. One outcome of the study was a revision of Procedure A to improve its
success rate. Results obtained from this study showed that both Class 1 and Class 2 methods can be used in
conjunction with fast SAR testing Procedure A with a very high level of confidence while considerably reducing the
amount of time and effort necessary to complete the compliance evaluation of a DUT.
Figure 24 is a simplified workflow depicting the steps detailed in Procedure A.
In order to substantially increase the user-friendliness of this procedure as well as to ensure

that the steps of Procedure A are consistently being followed correctly, a tool called the Fast
SAR Wizard has been developed. The Fast SAR Wizard can be downloaded from:
http://www.iec.ch/dyn/www/f?p=103:227:0::::FSP_ORG_ID,FSP_LANG_ID:1303,25
Procedure A has the following steps:
a) Use fast SAR testing to perform measurements on a chosen set of test configurations (see
3.52). The test configurations shall be grouped separately into N frequency band and
operating mode combinations, each having M i test configurations. These fast SAR testing
values are denoted SAR fast,(i,j) (i = 1 to N and j = 1 to M i ). When a Class 1 fast SAR testing
system is used, it is sufficient to include the centre channel for each test configuration.
When a Class 2 fast SAR testing is used, all required channels specified in 7.2.8 shall be
tested for each test configuration.
b) Use full SAR testing to measure the highest SAR test configuration from each of the N
groups. These N fast SAR testing values are denoted SAR max,fast,i (i = 1 to N). Select the
maximum fast SAR testing value SAR max,fast = max(SAR max,fast,i ).
– 86 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
With Class 1 fast SAR testing, it is not necessary to repeat the entire area scan during full
SAR testing when the number of SAR peaks is already known from fast SAR testing. The
full SAR testing method used in this step may use a reduced area, where all of the
requirements of 7.4.2 b) and Table 3 apply, except that the area may be smaller than the
projection of the DUT and antenna. The reduced area scan shall include at least the grid
point at each SAR peak found in step a), and the eight adjacent points surrounding each
SAR peak. The resulting area scan is valid if it also identifies the same SAR peaks and
peak locations found in step a), within positioning tolerance.
c) Use full SAR testing to measure all other test configurations (i,j), at the channel giving
highest fast SAR testing result, where the condition of Formula (13) is met. The second
paragraph of step b) applies.
SARfast,(i , j ) ≥ SARmax,fast ×  Bfast,i − Bfast,

2 
i − 1 (13)
 
B fast,i is calculated according to Formula (14).
1
Bfast,i = 2
(14)
1− (1, 64 U fast,i )
U fast,i in Formula (14) is the standard measurement uncertainty (k = 1) of fast SAR testing
results at the i-th frequency band and operating mode combination.
When using Class 2 system for fast SAR testing, U fast,i shall be set to the standard
uncertainty or 35 % / 2 = 17,5 %, whichever is greater. The 35 % figure is based on
analysis of a fast SAR testing interlaboratory comparison study completed in 2017.
NOTE Formula (13) and Formula (14) are derived for a 95 % confidence interval where
SARfast,(i, j ) ≤ SARmax,fast,i
This confidence interval is given as:
( SARmax,fast − SARfast,(i, j ) ) − 1,64U fast,i 2

Smax,fast 2
+ Sfast,(i,j ) ≤ 0 .
Solving for SAR fast,(i,j) results in Formula (13) and Formula (14), where U fast,i is the standard uncertainty and
has a Gaussian distribution. The value of 1,64 is the factor for a one-sided 95 % confidence interval of a
Gaussian distributed random variable.
d) Use full SAR testing to measure all other test configurations where the condition of
Formula (15) is met. The second paragraph of step b) and the last paragraph of step c)
apply.
SARfast,(i, j ) ≥ SARlimit ×  Bfast,i − Bfast,

2 
i −1 (15)
 
e) Use full SAR testing to measure all required channels specified in 7.2.8 across the
frequency band for the test configuration that resulted in the highest psSAR,
SAR highest,full = max(SAR full,(i,j) ), if this has not already been done. The reduced area scan
requirements outlined in step b) for Class 1 methods apply.
f) Use full SAR testing in accordance with 7.3 to measure any other test configurations that
have not been tested by fast SAR testing in step a) and are not excluded from testing by
this document or by applicable regulations.
IEC/IEEE 62209-1528:2020 – 87 –
© IEC/IEEE 2020
Figure 24 – Fast SAR Procedure A
7.9.2.3 Fast SAR testing Procedure B
Procedure B may be applied to determine the maximum SAR of a DUT for all the required test
configurations, or a subset thereof. Only Class 1 fast SAR testing may use Procedure B. This
procedure shall be applied separately for head and body phantom types. For the head
phantom type, the left head and right head data can be grouped together. For the body
phantom type, all body phantoms (e.g. if multiple phantom sizes were used) can be grouped
together.
Figure 25 shows a simplified workflow depicting the steps detailed in Procedure B.
Procedure B has the following steps:
a) Use fast SAR testing to perform measurements on a chosen set of test configurations (see
3.52). The test configurations shall be grouped separately into N frequency band and
operating mode combinations, each having M i test configurations. These fast SAR testing
values are denoted SAR fast,(i,j) (i = 1 to N and j = 1 to M i ). It is sufficient to include the
centre channel for each test configuration, but measurement of other required test
channels shall also be done for the configuration giving highest SAR for each of the N
groups.
b) Use full SAR testing to measure the highest SAR test configuration from each of the N
groups.
c) Use full SAR testing to measure all other test configurations (i,j), at the channel giving
highest fast SAR testing result, where the condition of Formula (15) is met. These values
in steps b) and c) are denoted SAR full,(i,j) .
– 88 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Full SAR testing can be limited to the channel giving highest fast SAR testing result only
when relative SAR distributions between channels are identical. Otherwise, full SAR
testing is required for all channels. In some specific cases (e.g. narrowband signals), a
visual comparison of SAR measurement plots for different channels can be applied to
determine that the SAR distributions are identical.
d) For all the full SAR testing values in steps b) and c), calculate and record the R i,j
according to Formula (16):
SARfast,(i, j ) − SARfull,(i, j )
Ri, j = (16)
SARfull,(i, j )
Use full SAR testing to measure all other test configurations, for all groups N where R i,j ≥
U fast,i and highest SAR fast,(i,j) and SAR full,(i,j) are at least 6 dB above their stated noise
thresholds. This testing is done at the channel having the highest fast SAR testing result.
Some test configurations could have only a single fast SAR testing result available, which
shall be interpreted as the highest result.
e) Use full SAR testing to measure all required channels specified in 7.2.8 across the
frequency band for the test configuration that resulted in the highest psSAR,
SAR highest,full = max[SAR full,(i,j) ], if this has not already been done.
f) Use full SAR testing in accordance with 7.3 to measure any other test configurations that
have not been tested by the fast SAR testing method in steps a) and e) and are not
excluded from testing by this document or by applicable regulations.
IEC/IEEE 62209-1528:2020 – 89 –
© IEC/IEEE 2020
Figure 25 – Fast SAR Procedure B
7.9.3 SAR test reductions
7.9.3.1 General
The number of SAR tests required to identify the configuration giving the maximum psSAR
can be large for a DUT capable of transmitting in multiple frequency bands and with different
usage configurations. This is also the case if the DUT can be used with different accessories.
The total measurement time required to test a personal wireless communication device can be
reduced by using fast measurement techniques that decrease the measurement time for each
series of tests (per 7.9.2), or by directly reducing the number of tests to be performed (in
accordance with 7.9.3). Combining test reduction techniques with fast SAR testing requires
further investigation, which will be considered in subsequent revisions of this document. SAR
test reduction methods applied during testing shall be reported as specified in 9.2 f).
Alternative approaches for reducing the number of tests are presented in 7.9.3. These
approaches are:
a) test reduction for different operating modes at the same test frequency band using the
same wireless technology;
– 90 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
b) test reduction based on characteristics of the DUT design;
c) test reduction based on peak SAR level analysis;
d) test reduction based on simultaneous multi-band transmission considerations.
NOTE 1 For example, GSM, GPRS, and EDGE (all using GMSK modulation) are considered to be the same
wireless technology, but GSM and UMTS are not considered the same wireless technology.
NOTE 2 The SAR test reduction procedures specified in 7.9.3 allow users of this document to save time required
for testing. These testing alternatives are not mandatory.
Supporting information on SAR test reduction can be found in Annex B.
7.9.3.2 Test reduction for lower power operating modes in the same frequency band
using the same wireless technology
When different operating modes (mod X, mod Y , etc.) work in the same frequency band,
procedures may be employed in some cases to reduce the number of measurements for the
lower power operating modes (mod Y , etc.) without compromising the stringency of the tests for
assessing the conservative exposure psSAR. The procedures may be applied if the following
conditions are met.
a) The same RF amplifier stage is used for mod X and mod Y .

b) The same antenna is used for mod X and mod Y .
c) No MIMO (multiple-input and multiple-output) techniques are applied for both mod X and
mod Y .
d) The same wireless technology is used for mod X and mod Y .
e) Either the same signal modulation is used and this modulation has constant amplitude, or
the time-averaged output power of mod Y is at least 2 dB below the output power of mod X .
f) The SAR probe calibration is valid for both mod X and mod Y (i.e. the probe calibration
certificate specifies the validity and uncertainty of the calibration applicable to both mod X
and mod Y ).
g) The same SAR measurement system (probe, phantom, tissue-equivalent medium, and
electronics) is used for testing all operating modes.
h) The RF carrier frequency of mod X is the same as for mod Y .
i) The channel bandwidth ratio (R m ) of mod X and mod Y satisfies Formula (17):
BWmod Y
Rm
= − 1 ≤ 0,3 (17)
BWmod X
To reduce the number of tests required, the following procedure may be applied if all of the
preceding conditions are met:
Step 1 Test all operating modes (mod X , mod Y etc.) in any one of the standard test
configurations (e.g. left-ear touch position).
Step 2 Fully test in all standard test configurations the operating mode that had the highest
SAR value in step 1.
Step 3 Fully test in all standard test configurations all other operating modes evaluated in
step 1 for which both of the following conditions apply:
• if the SAR value measured in step 1 for the operating mode under consideration is
within 15 % of the SAR value measured in step 1 of the highest SAR operating
mode (i.e. the operating mode chosen for use in step 2), and
• if the highest SAR value measured in step 2 is within 15 % of the applicable SAR
limit.
IEC/IEEE 62209-1528:2020 – 91 –
© IEC/IEEE 2020
The 15 % margin corresponds approximately to half the maximum expanded measurement
uncertainty of 30 %.
NOTE Both GSM and GPRS use GMSK, which is a constant amplitude modulation; therefore, the maximum time-
averaged output power with respect to the maximum number of time slots used in each operating mode can be
used to determine the most conservative operating mode for SAR testing. Similarly, EGPRS (which uses GMSK
and 8PSK) can be included with GSM and GPRS in this determination of the most conservative operating mode for
SAR testing, due to its innate similarities to GSM and GPRS. On the other hand, WiMAX uses OFDMA with high
peak-to-average power ratio, where the SAR measurement error is expected to vary with power.
7.9.3.3 Test reduction based on characteristics of DUT design
There is a wide range of different designs of handsets on the market. The SAR data analysis
in Clause B.2 has shown that the following test reduction and exclusion principles can be
applied for handsets used at the side of the head with internally mounted antennas and
operating frequencies between 800 MHz and 2 GHz. These criteria are based on the antenna
locations within a DUT and the variations in DUT styles. An extension above 2 GHz could be
considered if further supporting data becomes available.
The following requirements related to multiple operating modes within the same wireless
technology and frequency bands shall be satisfied to apply the 7.9.3.3 test reduction
procedure.
a) The same RF circuitry shall be used (transmitter, antenna feed-point, matching

components etc.).
b) The same antenna and antenna grounding shall be used.
c) No MIMO (multiple-input and multiple-output) techniques shall be applied.
d) The same SAR measurement system (probe, phantom, tissue-equivalent medium, and
electronics) shall be used.
For a DUT with a bottom-mounted internal antenna (i.e. an internal antenna mounted in the
bottom half of the DUT) that is fully embedded within 2,5 cm from the bottom of the device,
when the highest cheek position psSAR for a frequency band and operating mode is at least
3 dB from the SAR limit, testing in the tilt position is not required. For other antenna locations,
when SAR measured in tilt position using the highest SAR channel tested for the cheek
position is at least 3 dB from the SAR limit, testing of the other required channels for the tilt
position is not required.
To apply additional test reduction for the DUT using the same bottom-mounted internal
antenna for multiple operating modes within the same wireless technology and frequency
bands, select the communications system with the highest maximum time-averaged output
power then perform the full SAR testing procedure, including all configurations and test
positions specified in 7.3. Testing at tilt positions for other operating modes within this same
frequency band is not required regardless of the SAR level in cheek position(s), if both of the
following two conditions listed below are met.
• The cheek position has the highest psSAR value for a frequency band.
• The psSAR values for the configurations in the tilt position are at least 30 % below the
SAR limit.
The SAR measurement report shall include drawings or photos illustrating the layout as well
as the locations of the antennas in the DUT, and describing the wireless operating modes
applicable to each antenna, to support the considered test reduction and exclusion. When the
test reduction related to devices with the same bottom-mounted internal antenna for multiple
operating modes within same frequency band is to be applied, the antenna and the RF and
antenna matching circuitry shall be described for each operating mode in the SAR
measurement report.
Clause B.2 provides the rationale and supporting information for the analysis performed to
support these test reduction and exclusion principles.
– 92 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.9.3.4 Test reduction based on SAR level analysis
The purpose of this test reduction approach is to eliminate the “zoom scan” session for each
single measurement (7.4.2 d)) when the measured peak SAR from the “area scan” is below a
threshold, by assuring that the maximum psSAR value will be:
• correctly assessed and not underestimated, especially when it could generate a psSAR
close to the applicable compliance limit;
• correctly identified, even if the “zoom scan” session is not executed.
NOTE This test reduction procedure is different from the fast SAR testing procedure specified in 7.9.2. In this test
reduction protocol, the decision to apply the full SAR testing procedure is taken after looking at values measured
from the area scan (not the final psSAR). Moreover, a threshold value is necessary since the goal is to avoid the
zoom scan for some particular measurement configurations and not depend on a particular fast SAR fast SAR
testing algorithm to estimate the psSAR.
The rationale and studies supporting this test reduction approach are reported in Clause B.4.
When applying the protocol specified below, it is necessary:
– to adjust the “area scan” parameters specified in the Table 3 to use grid spacing not larger
than 10 mm;
– to perform the area scans at a fixed 4 mm maximum distance between the measured
points (geometric centre of the sensors) and the phantom inner surface.
The single band test reduction protocol (valid for each operating mode supported by the DUT)
is specified as follows.
a) Measure the local SAR as specified in 7.4.2 a).

b) Perform the area scan procedure for one of the positions specified in 7.4.2, at the channel
of the considered band having the highest rated output power, as indicated in the
complete measurement scheme specified in 7.4.2.
c) Determine the peak local SAR of the area scan; this is designated as the Absolute psSAR
(APS).
d) Perform the zoom scan, as indicated in the complete measurement scheme specified in
7.4.2.
e) Perform post-processing to determine the psSAR as in 7.4.2 e).
f) Measure the local SAR and evaluate the measurement drift as specified in 7.4.2 f).
g) For all the other positions and frequency channels to be tested, repeat the following
steps 1) through 5).
1) Measure the local SAR as specified in 7.4.2 a).
2) Perform the area scan procedure as indicated in the complete measurement scheme
specified in 7.4.2 b).
3) Evaluate the peak local SAR of the area scan; if the peak local SAR of the area scan is
greater than APS, then APS shall be assigned to the peak SAR of the area scan, after
completing step 4).
4) Perform the zoom scan required by 7.4.2 if the area scan peak SAR is not less than
1,3 W/kg (80 % of 1,6 W/kg) or if the peak SAR is greater than or equal to TH(f) × APS,
where the frequency dependent threshold value TH(f) is specified in Table 7 and APS is
the Absolute psSAR. If multiple peaks (as in 7.4.2 c)) are required to be measured,
apply this step to all the peaks.
5) Perform post-processing to determine the psSAR as in 7.4.2 e).
6) Evaluate the measurement drift as indicated in 7.4.2 f). If the drift is higher than 5 %,
the drift shall be addressed according to 7.4.2 f).
h) Step g) shall be performed sequentially for all applicable bands. The frequency dependent
threshold values used in step g) 4) are summarized in Table 7.
IEC/IEEE 62209-1528:2020 – 93 –
© IEC/IEEE 2020
i) When the 7.9.3.4 test reduction procedure is followed:
1) The block diagram on the left of Figure 19 shall be applied.
2) The block diagram on the right of Figure 19 shall be applied only for the first SAR
measurement. Subsequent SAR measurements shall be performed according to
Figure 26, which is a modification of the procedure specified in Figure 19.
Clause B.4 provides the rationale for the threshold levels used in this test reduction protocol,
based on SAR spatial gradient analysis.
Table 7 – Threshold values TH(f) used in this proposed test reduction protocol
Operating mode TH(f)

GSM900 0,75
a
UMTS VI 0,75
DCS1800 0,60
a
UMTS IX 0,60
UMTS I 0,60
a 3GPP UMTS-FDD Band VI and Band IX have uplink frequencies very close to
GSM900 and DCS1800, respectively. Even if only devices operating in GSM900,
DCS1800 and UMTS I bands were considered in Clause B.2, the thresholds in this
table shall be applied for UMTS-FDD Band VI and for UMTS-FDD Band IX.
Figure 26 – Modified chart of Figure 19

– 94 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.9.3.5 Test reduction based on simultaneous multi-band transmission
considerations
Simultaneous multi-band transmission means that the device can transmit in multiple
frequency bands at the same time, e.g. a Wideband Code Division Multiple Access (WCDMA)
transmission at 2 GHz and a Wireless Local Area Network (WLAN) transmission at 2,45 GHz.
The time-averaged output power of a secondary transmitter (i.e. the lower power transmitter,
e.g. Bluetooth, WLAN) can be much lower than that of the primary transmitter (i.e. the higher
power transmitter, e.g. WCDMA). In some cases, the secondary transmitter can be excluded
from SAR testing when used alone. However, when the primary and secondary transmitters
are used together (simultaneous transmission), the SAR limit could still be exceeded. A
means of determining the threshold power for the secondary transmitter that allows it to be
excluded from SAR testing is required.
One way of determining the threshold power level for the secondary transmitter (P available ) is
to calculate it from the measured psSAR of the primary transmitter (SAR 1 ) according to
Formula (18):
Pavailable = Pmax,m × (SARlim − SAR1) / SARlim (18)
where P max,m is the maximum threshold exclusion power level, which is calculated by
SAR lim × m, and m is an averaging mass.
For example, an exposure limit of SAR lim = 2,0 W/kg and an averaging mass of m = 10 g gives
total transmitting power of P max,m = 20 mW.
For example, an exposure limit of SAR lim = 1,6 W/kg and an averaging mass of m = 1 g gives
total transmitting power of P max,m = 1,6 mW.
Formula (18) can be easily generalized to the case where more than two transmitters are
transmitting simultaneously. If there are N simultaneous transmitters and the psSAR of the
first N − 1 transmitters are known (SAR i ), then the threshold power level for the Nth transmitter
can be found from Formula (19):
N −1
Pavailable = Pmax,m × ( SARlim − ∑ SARi ) / SARlim (19)
i =1
NOTE The applicability of power thresholds as specified in Annex B of IEC 62479:2010 [32] requires
investigation, to replace P max,m in the case where: i) the DUT is measured at the SAM ear, and ii) the radiating
parts of the DUT are closer than 5 mm to the phantom surface. The definition of new power thresholds for such
configuration is needed.
7.9.3.6 SAR measurements and test reductions for LTE devices
7.9.3.6.1 General LTE SAR testing considerations
LTE technology shows an added complexity over previously available radio schemes. In order
to configure and test LTE devices, many signal parameters are taken into account: frequency
band, channel bandwidth (e.g. from 1,4 MHz to 20 MHz), modulation (e.g. QPSK and 16-QAM
and higher-order QAM), number of resource blocks (RBs) allocated, offset of the resource
blocks within the channel bandwidth, as well as maximum power reduction (MPR). The
combinations of parameters in a given frequency band can result in hundreds of LTE modes
and SAR test configurations. In order to address this, a specific protocol is necessary for SAR
assessment of LTE devices. The main purpose of this protocol is to support demonstration of
DUT compliance with applicable limits based on a reasonable number of SAR evaluations.
NOTE 1 National regulatory agencies might have different requirements on test configurations for LTE devices.
IEC/IEEE 62209-1528:2020 – 95 –
© IEC/IEEE 2020
For a given LTE mode and device position, the psSAR is related to the maximum RF output
power. As a consequence, RF conducted power measurements can be used to quickly identify
high SAR LTE modes. SAR and RF conducted power are however not directly proportional
because:
a) RF conducted power is measured with a 50 Ω load impedance;

b) the antenna impedance of a DUT is generally not 50 Ω and varies over frequency;
c) the antenna impedance can be affected by device position and phantom coupling
conditions.
Because of this, a single SAR measurement using the LTE mode with the highest measured
maximum conducted output power in a frequency band might not be sufficient to demonstrate
compliance, unless the SAR value is significantly lower than the applicable compliance limit.
The required test conditions are established by applying the protocol in 7.9.3.6.3.
NOTE 2 The power measurement procedures are applicable to the LTE modes in a frequency band. SAR is
measured for the highest measured power condition in a frequency band for all device test positions according to
LTE conditions.
The LTE test configurations are measured according to the SAR measurement protocols in
this document. This document only specifies the procedures to identify the LTE test
configurations that will most likely result in SAR levels closest to the highest, and
conservative exposure can be obtained according to procedures in 7.9.3.6.2.
For the preceding reasons, studies on the relationship between RF conducted power and SAR
were conducted by IEC TC 106 using DUTs operating in 3GPP LTE bands 1, 4, and 17. The
results in Annex W show the following.
– QPSK modulation with 1 RB allocation generally produces the highest psSAR; MPR
(maximum power reduction) does not apply in this case.
– psSAR has good correlation with the measured RF conducted output power. The
relationship deviates from proportionality by less than 25 % (k = 2).
– For LTE modes with maximum conducted power lower than 85 % of P max , where P max is
the highest measured maximum RF conducted output power across all LTE modes in the
frequency band, it is highly unlikely that highest SAR results would be expected.
7.9.3.6.2 Description of LTE mode selection
Conducted power shall be measured for the largest channel bandwidth supported by the LTE
modes in each frequency band, using QPSK modulation with 1 RB allocation. The required
test channels shall be determined according to 7.2.8.
– When the number of required channels is one, the 1 RB shall be allocated at offset =
centre.
– When the number of required channels is three, the 1 RB shall be allocated at: i) offset =
0, ii) centre, and iii) maximum within the channel bandwidth.
– When the number of required channels is five, the 1 RB shall be allocated at: i) offset = 0,
ii) below centre, iii) centre, iv) above centre, and v) maximum within the channel
bandwidth, respectively, for channels from lowest to highest.
Other LTE modes besides “QPSK modulation with 1 RB allocation” shall also be measured for
the different channel bandwidth configurations using the modulations and RB allocations
specified by 3GPP for conformance testing. MPR settings are found in the appropriate 3GPP
documentation.
Conducted maximum output power shall be measured using the following test channel and RB
offset configurations.
a) When MPR does not apply, the configurations in Clause W.2 are measured for the low
and middle channels with RB offset = 0, and RB offset = maximum for the high channel.
– 96 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
b) When MPR applies, the configurations in Clause W.3 are measured
1) for the low channel with RB offset = maximum,
2) for middle channel with RB offset = 0 and maximum,
3) for high channel with RB offset = 0.
For “QPSK modulation with 1 RB allocation” in smaller channel bandwidth configurations,

when the same RB offset has already been measured in the highest channel bandwidth, such
RB offset configurations may be omitted. These correspond to RBs allocated next to the
channel edges; for example, offset = 0 for the low channel and maximum for the high channel,
for 5 MHz, 10 MHz, 15 MHz channel bandwidths in Band 1, may be omitted.
The conducted maximum output power measurement conditions are illustrated in Figure 27 for
5 MHz, 10 MHz, 15 MHz, and 20 MHz channel bandwidths in Band 1 (1 920 MHz to
1 980 MHz), and in Figure 28 for 5 MHz and 10 MHz channel bandwidths in Band 17
(704 MHz to 716 MHz). The orange and blue colour RBs correspond to test conditions that
are specified by 3GPP for conformance testing. The red colour RBs are the additional
configurations required for QPSK and 1 RB allocation in the largest channel bandwidth
configuration, not specified for 3GPP conformance testing. The required channels are
determined according to 7.2.8.
Conducted power testing for additional LTE modes is allowed, when the results form a
superset of the above requirements (e.g. conducted power testing according to FCC
requirements). When such a conducted power superset is reported, all results shall be taken
into account in the following SAR test procedures.
IEC/IEEE 62209-1528:2020 – 97 –
© IEC/IEEE 2020
Figure 27 – Use of conducted power for LTE mode selection,

for Band 1 (1 920 MHz to 1 980 MHz) (MPR values are in dB)
– 98 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure 28 – Use of conducted power for LTE mode

selection, for Band 17 (704 MHz to 716 MHz) (MPR values are in dB)
7.9.3.6.3 Power and SAR measurement protocol
SAR is tested at the highest conducted output power LTE mode for all device positions in
each frequency band, as determined by the power measurement procedures in 7.9.3.6.1. The
number of subsequent SAR tests is minimized according to the three alternative approaches
described in step 3.
The following measurement protocol is the same in principle as the protocol of 7.3, except
that it commences with the LTE mode with the highest conducted power, rather than the
centre of the transmit frequency band. The protocol applies if the DUT has the same
maximum output power settings regardless of the use condition (e.g. body worn versus voice
mode).
IEC/IEEE 62209-1528:2020 – 99 –
© IEC/IEEE 2020
a) Step 1 – RF conducted power measurement.
1) RF conducted power measurement shall be performed for all LTE modes specified in
7.9.3.6.2, where P(m) are the measured power for the m LTE modes. The measurement
protocol for RF conducted output power is specified in 7.9.3.6.4.
2) The LTE mode with the highest measured maximum conducted output power in each
frequency band is identified as P max .
3) The subset of LTE modes in a frequency band with measured power greater than or
equal to 85 % of P max is identified as S.
b) Step 2 – SAR measurement for the LTE mode generating maximum RF conducted power
(P max ) and device position causing maximum SAR.
SAR measurement shall be performed for all device positions for the LTE mode with
maximum RF conducted power (P max ) to determine the device position with the highest
psSAR, identified as SAR step2 .
c) Step 3 – The following three approaches can be applied to determine the highest psSAR.
1) Approach 1 (AP1) – Select the SAR measurement at device position causing maximum
SAR for step 2.
– For the test position corresponding to SAR step2 , multiply by α 95 to yield the scaled
highest psSAR, identified as AP1 SAR step3 . The factor of α 95 is the 95th percentile
value (mean plus two standard deviations) of the ratio α(m) between SAR and
conducted power (Formula (W.1)). A value of α 95 = 1,35 has been derived in the
two studies as described in Annex W. Other factors of α 95 may be used if
justification is given in the test report.
– Approach 1 is a fast method yielding conservative exposure. If maximum SAR
results obtained using this approach are above the SAR limit, Approach 2 can be
used to obtain more accurate maximum SAR results.
2) Approach 2 (AP2) – SAR measurement at device position causing maximum SAR for
LTE modes included in subset S.
– For the test position corresponding to SAR step2 , perform SAR measurements in this
test position for the other LTE modes included in the subset S. The highest psSAR
is identified as AP2 SAR step3 .
– Approach 2 is the more accurate method, but requires testing of the LTE modes in
each subset S.
3) Approach 3 (AP3) – Limited SAR measurements at device position causing maximum
SAR for LTE modes included in subset S.
– For the test position corresponding to SAR step2 , perform SAR measurements in this
test position for another LTE mode having the highest conducted power among all
the untested LTE modes included in the subset S.
– Multiply the measured psSAR by α 95 to yield the new scaled psSAR, identified as
AP3 SAR scaled .
– If AP3 SAR scaled is not below the SAR limit, repeat SAR measurement for the next
LTE mode as specified above in this step, otherwise select the highest value
among all the measured psSAR results and the last scaled value AP3 SAR scaled and
identify it as AP3 SAR step3 .
4) SAR step3 from AP1, AP2, or AP3 shall be reported as maximum SAR for the LTE
frequency band.
NOTE When assessing simultaneous multi-transmission SAR (e.g. LTE + WLAN), the maximum LTE SAR value
for each test position obtained using the preceding approaches AP1, AP2, or AP3, or combination of those, are
applied according to the simultaneous transmission SAR procedures in this document for this assessment.
– 100 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
7.9.3.6.4 RF conducted output power measurement for LTE modes
As explained in 7.9.3.6.1, RF conducted output power measurements are used to identify test
configurations that are unlikely to produce the highest SAR values for the LTE modes in a
frequency band. It is essential to define guidelines for ensuring repeatable and precise RF
conducted power measurements for LTE devices. Studies performed by IEC TC 106 [32]
showed that the results of such measurements can be affected by the presence or absence of
the device charger, and the number of data frames (10 ms for LTE) used for RMS power
averaging.
Hence, the following requirements apply to the measurement of conducted output power.
a) The device RF output port shall be connected to the measurement instrument (e.g. power
meter) and the cable shall not move during assessment of conducted power for a given
frequency band.
b) The device shall be battery powered and no charger shall be connected to the device.
Battery shall be fully charged for all power measurements.
c) The same subframe allocation shall be used for power measurements of all LTE modes in
a given frequency channel. As an example, measurements for all LTE modes in band 17
could be carried out with the following scheme: only subframe #0 is actively transmitting in
each data frame with the remaining nine subframes inactive.
d) The measured RF conducted output power shall be averaged over at least 20 frames.
Conducted power measurement results shall be documented, and measurement uncertainty

shall be taken into account. For a given frequency channel and a given LTE mode m, U[P(m)]
denotes the relative standard deviation of conducted power measurement results of mode m
obtained over at least five repeated measurements. The relative standard deviation is the
standard deviation divided by the average P ( m ) across the five measurements.
Conducted output power P(m) for a given LTE mode m at a specified frequency channel is
defined as P ( m ) . The maximum output conducted power across the m LTE modes is
{ }
designated as max m =1:M P ( m )max . U(P, k = 2) denotes the k = 2 uncertainty of conducted
{ }
power assessment, designated as max m =1:M U  P ( m )  . The notation P(m) and P max is used
for simplicity.
8 Measurement uncertainty estimation
8.1 General
The uncertainty budget is applicable to head and body SAR measurement results. The
estimate of the measurement repeatability, as verified by the system check procedure, is
obtained by omitting the Type B contributions from the uncertainty budget. Table 9 (see 8.3)
is an example uncertainty budget for 1 g or 10 g SAR measurement of a DUT or a validation
antenna.
Each input quantity is described together with an example of how the associated standard
uncertainty is evaluated. These are general examples that might not be applicable to all
measuring systems, and additional quantities not addressed in this document can be required
in specific cases.
Further information on assessing the uncertainty contributions for system manufacturers and
calibration laboratories is given in 8.4.1.1 through 8.4.1.6, 8.4.1.8, 8.4.1.9, and 8.4.2.1.
IEC/IEEE 62209-1528:2020 – 101 –
© IEC/IEEE 2020
The expanded measurement uncertainty is calculated from Formula (20):
U= k ⋅ uc ( y ) (20)
A coverage factor of k = 2 is used in this document, corresponding to a 95 % confidence

interval. The combined standard uncertainty, u c (y), is calculated from Formula (21):
N 2 2 N 2 2 N 2
uc y =
=
( )
i 1 i =∑ ( )
c ⋅  a xi qi  = ∑
ci ⋅ u xi =
i 1=
u
i 1 i
( ) ∑ ( y) (21)
where
xi is the estimated value of the i-th input quantity X i that influences the uncertainty;
y is the value of the output quantity Y that corresponds to the combined uncertainty;
a(x i ) is the uncertainty value of x i having a probability density function PDF i ;
qi is the divisor corresponding to the probability density function PDF i of x i ;
u(x i ) = a(x i ) / q i is the standard uncertainty of x i , corresponding to a normal distribution;
ci is the sensitivity coefficient associated with x i . It describes how y varies with small
changes in x i . It is calculated as the partial derivative of the model function f with
respect to x i as described in Formula (22):
ci =∂f ∂xi (22)
u i (y) = c i ⋅ u(x i ) is the standard uncertainty of y, corresponding to a normal distribution.
The divisors q i shall be applied according to the probability density function PDF i as specified
in ISO/IEC Guide 98-3. Table 8 gives divisors for common PDFs.
Table 8 – Divisors for common probability density functions (PDFs)
PDF Normal Rectangular Triangular U-shape

Divisor q k √3 √6 √2
8.2 Requirements on the uncertainty evaluation
The expanded measurement uncertainty, U, is the result of the uncertainty evaluation, specific
details of which are recorded in an uncertainty budget. Separate uncertainty budgets shall be
evaluated for SAR measurements according to the following.
a) The expanded uncertainty in the measured value of psSAR (1 g or 10 g) shall be

assessed in accordance with the guidance and explanations given in ISO/IEC Guide 98-3.
b) The frequency range and measurement conditions for which the uncertainty analysis is
valid shall be specified.
c) The reported uncertainty shall represent the maximum expanded uncertainty within this
range.
d) The expanded uncertainty in the system check and system validation measurements shall
also be evaluated. The uncertainty shall be quoted at the 95 % confidence level, and the
expanded uncertainty shall not exceed 30 % of the psSAR value in the range 0,4 W/kg to
10 W/kg. For psSAR values below 0,4 W/kg, the expanded uncertainty shall not exceed
0,12 W/kg (30 % of 0,4 W/kg).
e) Where acceptable by national requirements, if the expanded uncertainty is greater than
30 %, the reported data needs to take into account the percentage difference between the
actual uncertainty and the 30 % target value – e.g. see method of [34].
– 102 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
8.3 Description of uncertainty models
8.3.1 General
The uncertainty models in 8.3 are described in the format of Formula (23):
Y = X1 + X1 +  + X N (23)
Y = ∆SAR is the output quantity corresponding to the combined SAR assessment uncertainty.
The input quantities X i are described by the symbols in 8.4. Formula (21) (see 8.1) shall be
used to calculate combined standard uncertainty u c (y).
With reference to the guidelines for uncertainty assessment in ISO/IEC Guide 98-1 [35] and
ISO/IEC Guide 98-3, the uncertainty models for the following processes are specified using
the symbols described in 8.4.
8.3.2 SAR measurement of a DUT
The measurement uncertainty model formula (model equation) for results of DUT SAR
measurements is given by Formula (24).
2
∆SAR = CF + CFdrift + LIN + BBS + ISO + DAE + AMB + ∆ xyz + DAT + LIQ ( σ )
δ (24)
+ LIQ (Tc ) + 0, 25 EPS + 2 DIS + Dxyz + H + MOD + TAS + RFdrift + C ( ε ′, σ ) + C ( R )
Table 9 shows the uncertainty budget template for DUT psSAR measurement results, as well
as descriptions of the symbols used in Formula (24).
8.3.3 System validation and system check measurement
The measurement uncertainty model formula for results of system validation and system
check SAR measurements is given by Formula (25).
2
∆SAR = CF + CFdrift + LIN + ISO + DAE + AMB + ∆ xyz + DAT + LIQ ( σ )
δ (25)
+ LIQ (Tc ) + 0, 25 EPS + 2 DIS + RFdrift + Pin + VAL + C ( ε ′, σ )
Table 9 shows the uncertainty budget template for system validation psSAR measurement
results, as well as descriptions of the symbols used in Formula (25).
8.3.4 System check repeatability and reproducibility
The measurement uncertainty model formula for results of system check repeatability and
reproducibility SAR measurements is given by Formula (26); symbols are as described in
Table 9.
2
∆SAR
= CFdrift + ∆ xyz + LIQ ( σ ) + LIQ (Tc ) + 0, 25 EPS + 2 DIS + RFdrift + Pin + C ( ε ′, σ ) (26)
δ
8.3.5 Fast SAR testing (relative measurement)
The measurement uncertainty model formula for results of fast SAR (relative) measurements
is given by Formula (27); symbols are as described in Table 9.
IEC/IEEE 62209-1528:2020 – 103 –
© IEC/IEEE 2020
2
∆SAR= LIN + ISO + DAE + AMB + ∆ xyz + DAT + 2 DIS + H + Dxyz + MOD + RFdrift (27)
δ
Table 9 – Uncertainty budget template for evaluating the uncertainty

in the measured value of 1 g or 10 g psSAR from a DUT or validation
antenna (N = normal, R = rectangular)
Prob.
Input quantity X i Unc. Div. a u(x i ) = u(y) =
Symbol Ref. Dist. a ci νi
(source of uncertainty) a(x i ) qi a(x i )/q i c i ∙u(x i )
PDF i
Measurement system errors
N
CF Probe calibration 8.4.1.1 2 1 ∞
(k = 2)
CF drift Probe calibration drift 8.4.1.2 R √3 1 ∞
Probe linearity and detection

LIN 8.4.1.3 R √3 1 ∞
limit
BBS Broadband signal 8.4.1.4 R √3 1 ∞
ISO Probe isotropy 8.4.1.5 R √3 1 ∞
Other probe and data

DAE 8.4.1.6 N 1 1 ∞
acquisition errors
AMB RF ambient and noise 8.4.1.7 N 1 1 ∞
∆ xyz Probe positioning errors 8.4.1.8 N 1 2/ δ
DAT Data processing errors 8.4.1.9 N 1 1 ∞
Phantom and device (DUT or validation antenna) errors
Measurement of phantom
LIQ(σ) 8.4.2.1 N 1 cε, cσ ∞
conductivity(σ)
LIQ(T c ) Temperature effects (medium) 8.4.2.2 R √3 cε, cσ ∞
see
EPS Shell permittivity 8.4.2.3 R √3 ∞
8.4.2.3
Distance between the radiating

DIS element of the DUT and the 8.4.2.4 N 1 2 ∞
phantom medium
Repeatability of positioning the

D xyz DUT or source against the 8.4.2.5 N 1 1 5
phantom
H Device holder effects 8.4.2.6 N 1 1
Effect of operating mode on

MOD 8.4.2.7 R √3 1 ∞
probe sensitivity
TAS Time-average SAR 8.4.2.8 R √3 1 ∞
Variation in SAR due to drift in

RF drift 8.4.2.9 N 1 1
output of DUT
Validation antenna uncertainty

VAL 8.4.2.10 N 1 1
(validation measurement only)
Uncertainty in accepted power

P in 8.4.2.11 N 1 1
(validation measurement only)
Corrections to the SAR result (if applied)
Phantom deviation from target

C(ε′,σ) 8.4.3.1 N 1 1
(ε′,σ)
C(R) SAR scaling 8.4.3.2 R √3 1
u(∆SAR) Combined uncertainty
Expanded uncertainty and

U U= ν eff =
effective degrees of freedom
a
Other probability distributions and divisors may be used if they better represent available knowledge of the quantities
concerned.
– 104 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
8.4 Parameters contributing to uncertainty
8.4.1 Measurement system errors
8.4.1.1 Probe calibration
CF – Calibration uncertainty for the sensitivity of the SAR probe in the phantom, obtained
from the calibration certificate. A protocol for the evaluation of the SAR probe calibration
coefficients and the uncertainty estimation is given in Annex E.
8.4.1.2 Probe calibration drift
CF drift – Drift in CF during the calibration interval, estimated from the probe calibration history,
knowledge of similar probes, or manufacturer’s specifications.
8.4.1.3 Probe linearity and detection limit
LIN – Change in probe sensitivity as a function of applied SAR for CW signals between the
measured SAR and the level at which the probe was calibrated. The probe linearity and
detection limit errors, measured according to Clause O.1 over the range of at least 0,12 W/kg
to 100 W/kg, are given in the probe calibration certificate or manufacturer’s specifications.
8.4.1.4 Broadband signal
BBS – Uncertainty due to the variable response of the probe as a function of frequency. This
term is negligible for measurement of narrowband signals. Otherwise, it shall be calculated as
specified in Clause O.2.
8.4.1.5 Probe isotropy
ISO – Change in sensitivity of the SAR probe due its rotation and inclination relative to the
field vector being measured. It is usually given in the calibration certificate, or from
manufacturer’s specifications as axial isotropy (change in sensitivity as the probe is rotated
about its axis) and hemispherical isotropy (change in sensitivity as the probe is rotated about
its axis and inclined with respect of the field vector). The isotropy of the probe shall be
measured according to the protocol specified in Clause E.4. During the measurements, the
probe axis is maintained within a specified angle to the normal of the phantom surface, in
order to limit the isotopy errors (Table 3 and Table 4). An empirical approximation for the total
isotropy error for this case is given by Formula (28):
2 2
=ISO 0,5 µaxial + 0,5 µhemispherical (28)
and a rectangular probability distribution shall be used.
8.4.1.6 Other probe and data acquisition errors
DAE – Other probe and data acquisition errors, including spatial resolution (fast SAR testing),
sensor offset distance, probe integration-time and response-time (Clause O.6), boundary
effects and corrections (Clause O.3), and data acquisition errors (Clause O.4), which are
given in manufacturer’s specifications.
8.4.1.7 RF ambient noise and reflections
AMB – The effects of ambient RF signal and system noise levels as well as reflections are
assessed by performing a SAR measurement without the DUT transmitting. This document
specifies that the results shall be less than or equal to 0,012 W/kg (i.e. 3 % of 0,4 W/kg). It is
not necessary to check RF ambient noise prior to each SAR test, provided the laboratory can
demonstrate that there are no new RF sources. The procedure to evaluate the contributions of
the ambient noise and reflections to the uncertainty is given in Clause R.1.
IEC/IEEE 62209-1528:2020 – 105 –
© IEC/IEEE 2020
8.4.1.8 Probe positioning errors
∆ xyz – Error in the position of the probe tip relative to the surface of phantom shell over the
volume used for the calculation of 1 g or 10 g SAR. The uncertainty is obtained from
manufacturer’s specifications, or by measuring the probe tip position and separation from the
phantom shell at several points within the scan volume and comparing these to the requested
positions. In the case of the flat phantom, this assessment shall include the effect of sag of
the bottom of the tank. The positional tolerance shall be ≤ 0,2 mm. Because the SAR decays
with distance from the phantom surface as e −2x/ δ , the sensitivity coefficient c i is approximately
2/ δ , where δ is the penetration depth in the phantom in mm. Details are given in Clause O.7.
Since N ≥ 96 measurement positions are used to determine the 1 g or 10 g SAR, the Type A
contributions associated with the scanning system repeatability will be reduced by √N.
8.4.1.9 Data processing errors
DAT – Error resulting from calculation and data processing algorithms used in the
measurement software for the 1 g or 10 g averaged SAR calculated from a limited number of
spatial samples. The errors include those due to extrapolation and interpolation. The error
magnitude is assessed by the manufacturer or software supplier using a set of test functions.
Details are given in Clause O.8.
8.4.2 Phantom and device (DUT or validation antenna) errors
8.4.2.1 Measurement of phantom conductivity
LIQ(σ) – Error in the measurement of phantom conductivity using the dielectric properties
measurement techniques specified in Annex H, normally obtained from the system vendor.
The procedure is specified and an example uncertainty table is given in Clause O.9.
8.4.2.2 Temperature effects
LIQ(T c ) – Error due to a temperature variation ∆T of the phantom between the dielectric
properties measurement and the SAR measurement.
For each phantom recipe, the temperature coefficient T c (°C) of ε′ and σ at the frequency of
interest should be estimated from measurements with the tissue-equivalent medium at
different temperatures from 18 °C to 25 °C, and measured to an accuracy of 0,1 °C.
The error due to the temperature variation of ∆T is then calculated from Formula (29):
{
LIQ (Tc ) =∆T × 0, 25 Tc ( ε ′ ) + 0, 75 Tc ( σ ) } (29)
where the weighting factors are obtained from consideration of the correction formulas of
O.9.6. The value of ∆T shall not exceed 2 °C.
8.4.2.3 Shell permittivity
EPS – Tolerance in the real relative permittivity of the shell relative to ε′ = 4 is calculated from
Formula (30):
∆ε ′ ε ′ − 3, 5
EPS
= = (30)
ε′ 3, 5
– 106 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The actual relative permittivity shall lie in the range 2 to 5 for head phantoms, and 3 to 5 for
body phantoms. Within this range, and for frequencies greater than 3 GHz, a 25 % change in
ε′ yields an approximately 5 % change in 1 g or 10 g averaged SAR, so a sensitivity
coefficient c i = 0,25 shall be used. For frequencies up to 10 GHz, the change in psSAR
increases to 10 %. For frequencies of 3 GHz and below, the error due to the shell permittivity
is negligible. For the uncertainty of the psSAR due to the shell permittivity, the following
sensitivity coefficient c i shall be applied as described in Formula (31):
0 f ≤ 3 GHz

=ci 0, 25 3 GHz < f ≤ 6 GHz (31)
0, 5 6 GHz < f ≤ 10 GHz

Further information can be found in [36].
8.4.2.4 Distance between DUT and tissue-equivalent medium
DIS – Uncertainty in the separation distance between the radiating element within the DUT
and the tissue-equivalent medium. The separation d of the current source to the phantom is
the sum of the spacer thickness, the shell thickness, the sag, and the 5 mm default distance
from the radiating element to the surface of the DUT. This distance can be changed to reflect
the actual distance if available.
The sag in the bottom of the flat phantom when filled with the required depth of tissue-
equivalent medium (sag) can be assessed by placing a straight edge such as a metal ruler
under the shell. For the head phantom, the sag is usually insignificant and can be assumed to
be zero.
The resulting distance error, DIS, is then calculated from Formula (32):
∆d 1
( ∆spacer ) + ( ∆ thickness )2 + ( sag )2
2
DIS
= = (32)
d d
where ∆ thickness is obtained from the manufacturer’s specifications for the tolerance of the
shell thickness, and shall be less than 0,2 mm. Because the SAR in the phantom decays as
approximately d 2 , a sensitivity coefficient of 2 shall be applied.
8.4.2.5 Positioning repeatability
D xyz – Error in SAR measurements due to the deviation of the DUT position from the nominal
position. It is assessed by performing at least five repeated 1 g or 10 g SAR measurements
while repositioning the DUT in the holder after each measurement. This error may be
assessed for a group of DUTs having very similar shape (form factors), dimensions, and SAR
distributions, provided that at least six different devices are tested and each is repositioned
five times with half the measurements being in the cheek position and half being in the tilt
position. Where a database of such tests is maintained for a particular device holder, it is
recommended that this is updated yearly in order to account for DUT design changes. Details
are given in Clause R.2.
8.4.2.6 Device holder effects
H – Change in the 1 g or 10 g SAR caused by the device holder. This depends on the DUT.
The error shall be assessed by performing 1 g or 10 g SAR measurements with the DUT in
contact with the phantom: a) with the DUT in the device holder, and b) with the DUT
supported by a polystyrene block. A device holder uncertainty can be attributed to a generic
device type provided that the variability between a sample of at least six of these devices is
assessed, and an additional Type A uncertainty term is added to account for this variability.
Details are given in Clause R.2.
IEC/IEEE 62209-1528:2020 – 107 –
© IEC/IEEE 2020
8.4.2.7 Signal modulation
MOD – Probe measurement error due to the operating mode of the DUT. This occurs when
the probe detectors do not yield the RMS value of the modulated signal emitted by the DUT.
Where the uncertainty for a particular modulation scheme is not provided by the probe
manufacturer, this shall be assessed in accordance with the procedure given in Clause R.3.
8.4.2.8 Time-average SAR
TAS – The uncertainty related to time-period averaged SAR assessment is due to the
sampling rate of the measurement. A fast sampling rate will allow the SAR measurement
system to accurately capture fast changes in the signal amplitude or duty cycle over the
averaging time, allowing accurate calculation of the TX factor (details are given in
Clause R.4).
8.4.2.9 Variation of SAR due to drift
RF drift – Change in the SAR at a fixed position within the phantom due to drift in the RF output
of the DUT during the SAR measurement sequence (7.2.4). Where the drift is greater than
5 %, additional stability measurements of the DUT shall be made (Clause R.5), and
corrections may be applied to the SAR measurement data to compensate for this variation. If
such corrections are applied, RF drift may be omitted from the uncertainty analysis.
8.4.2.10 Validation antenna uncertainty
VAL – Error arising from deviation of the physical validation antenna from that used to derive
the reference values, including the accepted power, power losses and dimensions (Annex S).
8.4.2.11 Uncertainty in accepted power
P in – Error in the power accepted by the validation antenna.
8.4.3 Corrections to the SAR result (if applied)
8.4.3.1 Phantom deviation from target dielectric properties
C(ε′,σ) – Correction given in 7.8 is applied to the 1 g or 10 g SAR to correct for deviations of
the phantom target properties from those specified in this document. The mean tolerance of
this correction, specified as the square root of the mean of the squared differences between
the SAR deviation predicted by the formulas and the simulated deviations, is shown in
Table 10 for the 1 g psSAR and the 10 g psSAR. Table 10 shows how the uncertainty
increases as the maximum allowable values of ∆ε′ and ∆σ increase. It has also been shown
that these corrections are valid for realistic DUT models [28], [29].
Table 10 – Uncertainty of Formula (8) (see 7.8.2) as a function of

the maximum change in permittivity or conductivity
Max. tolerance Standard uncertainty Standard uncertainty

for ε or σ for 1 g SAR, % for 10 g SAR, %
±5 % 1,2 0,97
±10 % 1,9 1,6
8.4.3.2 SAR scaling
C(R) – Error in the power ratio R used to determine the SAR from a DUT based on
measurements made with the device transmitting a different mode and or power (7.2.3.2). If
the signal envelopes of the two operating modes are the same, this error corresponds to the
linearity of the power measurement device used. Where the envelopes are different, the effect
on the sensitivity of the power meter shall also be included. The error should not exceed 5 %.
– 108 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The uncertainty may be evaluated by measuring the SAR from the DUT at the location of the
psSAR in both modes/power levels and comparing the ratio obtained with R. In this case a
rectangular probability distribution should be used. Details are specified in Clause R.6.
9 Measurement report
9.1 General
All test results shall be recorded in a measurement report and shall include all the information
necessary for the interpretation of the DUT configurations tested, calibration performed and
all information required by the method and instrumentation used.
A measurement report, compliant with ISO/IEC 17025 [17] and including as a minimum the
items listed below, shall be used demonstrate compliance with the requirements of this
document.
9.2 Items to be recorded in the measurement report
All of the information needed for performing repeatable tests, calculations, and measurements
giving results within the required calibration and uncertainty limits, shall be recorded. The
measurement report shall include the following items.
a) General introduction
1) Identification of the test laboratory.
2) Identification of the DUT including hardware and software revision numbers, serial
number, e.g. IMEI (international mobile equipment identity).
3) Compliance requirements, e.g. test standards, guidelines, recommendations, etc.
4) Exposure limits applicable, e.g. ICNIRP, IEEE/ICES, etc.
5) A list of any accreditations provided by national or international bodies to perform
testing in accordance with the standards listed in a) 3). This shall include the date of
expiry.
b) Measurement system (for SAR measurement, and fast SAR testing if used)
1) Measurement system main components descriptions, including positioner, tissue-
equivalent medium, readout electronics, device holder, phantom(s), and any other
relevant components.
For the probe(s) used, include dimensions, isotropy, spatial resolutions, dynamic
range, and linearity.
2) Calibration data for relevant components, including fast SAR testing system critical
components, e.g. probe calibration certificates, shall be provided in the measurement
report.
3) Description of the interpolation and extrapolation algorithms used in the area scans
and/or zoom scans shall be documented in the measurement report.
4) Tissue-equivalent media and materials used and characteristics.
For each tissue-equivalent medium or material used, include:
i) dielectric properties for each frequency band;
ii) deviation from target value;
iii) tissue-equivalent medium temperature;
iv) composition summary of the tissue-equivalent media.
5) Results of system check:
i) measurement results for each frequency band;
ii) deviation from the SAR target value;
iii) radiating source description.
IEC/IEEE 62209-1528:2020 – 109 –
© IEC/IEEE 2020
c) Uncertainty estimation (system check for SAR measurement, system validation, and fast
SAR testing if used)
1) To include measurement uncertainty values from Clause 8 and Annex C.
2) Any other relevant items.
d) Device and test details
1) Description of the form factor of the DUT and a brief description of its intended
function.
2) Description of the positions and orientations tested, including photos, and rationale for
any test reductions. If any offset in the DUT position (in accordance with 6.5), then this
shall also be documented.
3) Description of the available and tested antenna(s) and accessories, including batteries.
4) Description of the available and tested operating modes, power levels, and frequency
bands; and rationale for any test reductions.
5) Testing environmental condition, e.g. temperature.
6) Results of all tests performed (psSAR value for each test, and graphical representation
of the coarse scans with respect to the device for the maximum SAR value of each
operating mode) and details on scaling of the results.
e) The measurement report shall include the following information relevant to the validation
of the SAR measurement method.
1) Description of the validation procedure(s).
2) Results of the computations, measurements, and/or other assessments performed by
the method developer (e.g. system manufacturer) in order to validate the SAR
measurement method.
3) Additional analyses or conditions imposed by the method developer and applied by the
user to satisfy the SAR measurement method if applicable.
4) Radiating source description and SAR distribution for each frequency band.
5) Operating frequencies, operating modes, device operating configurations, DUT
position on the phantom, and SAR distributions for each frequency band specific to the
method.
6) SAR results measurement uncertainty.
7) In the case of fast SAR testing, details of any modifications to post-processing
algorithms from those implemented for full SAR testing.
f) SAR test reduction reporting
When test reduction procedures are applied during the SAR measurements of a DUT, the
measurement report shall include additional information related to the following test
reduction alternatives.
1) Test reduction for different operating modes at the same frequency (see 7.9.3.2). The
measurement report shall provide a detailed description on how the conditions of
7.9.3.2 have been met.
2) Test reduction based on characteristics of the DUT design (see 7.9.3.3). The
measurement report shall include drawings or photographs illustrating the layout and
locations of antennas in a DUT, and describing the wireless operating modes
applicable to each antenna, to support the test reduction and test exclusion
considered.
3) Test reduction based on peak SAR level analysis (see 7.9.3.4). The measurement
report shall include a systematic description on how the test reduction protocol 7.9.3.4
was applied for the measurements of the DUT.
4) Test reduction based on simultaneous multi-band transmission considerations (see
7.9.3.5). The measurement report shall include the measured time-averaged output
power, and how it meets the threshold power level available, in accordance with
Formula (18) and Formula (19).
– 110 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
g) Measurement report summary (for all fast SAR testing and full SAR testing data)
1) Frequency bands and configurations:
i) list of all operating modes tested;
ii) list of all test configurations assessed.
2) Tabulated SAR values over the testing positions, bands, operating modes, and device
configurations.
3) Tabular and graphical results for the highest fast SAR testing value for each operating
mode.
4) Results of all full SAR tests performed as a result of 7.4 and 7.5, which include the
psSAR value for each required test, and graphical representation of the scans with
respect to the device.
5) The values related to the conditions of Formula (13) through Formula (19) shall be
documented in the measurement report.
6) A systematic rationale based on 7.9.2 and 7.9.3 for excluding full SAR testing shall be
provided.
7) Reference to exposure limits and a statement of compliance, or otherwise a list of the
exposure limits and a compliance statement.
h) Proximity sensors reporting
For a DUT requiring SAR testing with active proximity sensor(s), the measurement report
shall include all reporting requirements of 7.7. In addition, the measurement report shall
include the following.
1) Description of the proximity sensor(s), location(s) in the DUT, the tissue-equivalent
media used in the triggering procedures, and any other relevant information.
2) Description of the test set-up for the conducted power measurements (or radiated
power measurements, if applicable).
3) Test set-up photos.
4) Tabulated results of sensor triggering distance, sensor coverage, and sensor
sensitivity (capacitance/mm as specified by the sensor manufacturer). The triggering
points shall also be documented graphically, with the antenna and sensor clearly
identified, along with all relevant dimensions.
5) The test set-up that establishes the 1 mm movement related to the triggering distance
shall be explained in the measurement report (see 7.7.4).
6) The measured output power or sensor state determined by the test software, within
±5 mm of the triggering points, or until the DUT is touching the phantom, for
movements to and from the phantom, shall be tabulated in the report (see 7.7.4). The
output power (or sensor state in case of test software usage) versus the DUT to
phantom spacing shall also be plotted.
7) The default power level for sensor failure and malfunctioning, including all compliance
concerns, shall be fully documented in the report.
8) If the triggering procedures applied varied from the procedures outlined in 7.7, the
measurement report shall include the product-specific triggering procedures. It shall
also include a description of the proximity sensor triggering conditions in the normal
operating mode of the DUT, location of each proximity sensor, applicable transmission
mode, and frequency range of the antenna(s) covered by each proximity sensor, and
any other information that is required for regulatory approval or deemed relevant for
testing the device.
9) If additional SAR tests were required because the measured peak SAR location for the
antenna was not within the triggering coverage area, the test configurations for the
additional testing shall be clearly described and justified in the measurement report.
i) Time-averaged SAR
1) Description/specifications of the SAR measurement system ability to measure and
capture the time-averaging algorithm results, including the minimum response time.
IEC/IEEE 62209-1528:2020 – 111 –
© IEC/IEEE 2020
2) Description of the time-averaging implementation, including the accuracy of the
averaging algorithm(s), and minimum response time necessary to properly assess
the TX factor.
3) Detailed description of the applicable test mode software, including version number,
required to obtain worst case TX factor, if required.
4) Signal characteristics that are controlled/modified by the time-averaging algorithm
(e.g. output power, duty factor, modulation, bit rate, etc.).
5) Applicable frequency bands and operating modes.
6) Applicable exposure conditions, test positions, and associated test separation
distances.
7) Maximum psSAR evaluation for all applicable modes of operation.
8) TX factor assessment, including output power/SAR measurement plots used to
determine the TX factor.
9) Simultaneous transmitter SAR considerations.
10) Repeated TX factor assessment for the worst-case test condition. If more than one
time-averaging algorithm implementation is being used by the device for different
chipsets (e.g. one for WLAN and one for WWAN), the TX factor assessment shall be
repeated for each chipset for the worst-case tested configuration.
11) Mechanism(s) that are in place to prevent the power control algorithm(s) from being
altered or accessed by the user.
12) Time-averaged SAR results measurement uncertainty evaluation.
13) Results of the computations, and/or measurements performed in order to validate the
measurement system’s capability to perform TX factor assessment.
– 112 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex A
(normative)
SAR measurement system verification
A.1 Overview
Procedures are specified for two levels of SAR measurement system verification:
– system check;
– system validation.
System check provides a fast and reliable test method that can be applied routinely to verify
SAR system measurement accuracy. The objective is to ascertain that the SAR system and
tissue-equivalent media are acceptable for testing at the operating frequencies of the DUT.
The system check test requires a flat phantom (see A.2.2) and a radiating antenna (see
A.2.3).
System validation provides a means of system-level validation against the specifications of

the SAR system and its components. The test set-up consists of a flat phantom and a system
validation antenna (Annex D). This test is performed annually (e.g. after probe calibration),
before measurements related to interlaboratory comparison (Annex T), and every time
modifications have been made to the system, such as new probes, software changes, or
adding different readout electronics or probes.
Since a flat phantom is used, both system check and system validation do not address
measurement uncertainty relating to the anthropomorphic SAM phantom (when applicable)
and test device positioning variability.
NOTE For measurements made using the anthropomorphic SAM phantom, interlaboratory comparisons (see
Annex T) enable SAR measurement reproducibility to be quantified using a reference DUT. The measurements
address both data scattering due to the SAM phantom and device positioning uncertainty, which are not included in
system check and system validation. The tests can also be used to establish the expected measurement accuracy
and uncertainty across various laboratories.
A.2 System check
A.2.1 Purpose
The purpose of the system check is to verify that the measurement system operates within its
specifications at the device test frequencies. The system check verifies the measurement
accuracy and repeatability of a SAR system before compliance testing; the system check is
not a validation of all system specifications. The latter is not required for testing a device, but
is mandatory before the system is deployed. The system check detects possible short-term
drift and unacceptable measurement errors or uncertainties in the system, such as:
a) incorrect tissue-equivalent medium parameters, or changes in the medium parameters;

b) test system component failures;
c) test system component drift;
d) operator errors in measurement set-up or software parameters;
e) other possible adverse conditions in the system configuration, e.g. RF interference.
IEC/IEEE 62209-1528:2020 – 113 –
© IEC/IEEE 2020
The system check is a complete 1 g or 10 g psSAR measurement in a simplified set-up with a
system check antenna (see A.2.3). The system components, software settings, and other
procedures in the system check shall be the same as those used for the compliance tests.
The system check shall be performed using the same tissue-equivalent medium and probe
calibration frequency point as in the compliance tests, and within the valid frequency range of
the probe calibration, tissue-equivalent medium dielectric properties, and antenna return loss
required for the SAR measurements. The system check shall be performed at a chosen fixed
frequency that is within the validity of the calibration conversion factors, and within ±10 % or
±100 MHz of the compliance test mid-band frequency, whichever is greater. Above 6 GHz, the
system check test frequency shall be within ±1 GHz of the compliance test mid-band
frequency. The environmental requirements for system check tests are specified in 6.1.
The system check shall be performed within 24 hours before the DUT SAR evaluation, and on
the same SAR measurement system that is used for the SAR evaluation of the DUT.
The system check acceptance criteria are provided in A.2.6.
A.2.2 Phantom set-up
A flat phantom shall be used with the recommended tissue-equivalent medium for system
check and system validation. The specifications for the flat phantom are given in Clause G.3.
For dipole antennas, the feed-point shall be positioned below the centre of the flat phantom
(dimensions given in 6.2.4), and the dipole arms shall be aligned with the major or long axis
of the phantom (see D.2.1 for dipole specifications). For waveguides, the longer side of the
waveguide shall be aligned with the major axis of the phantom. For confined loop antennas
(CLA), the device shall be centred below the middle of the flat phantom. For meander dipole
antennas, the feed-point shall be positioned below the centre of the flat phantom. For VPIFAs,
the feed-point shall be positioned below the centre of the flat phantom, and the ground plane
(see Figure D.5) shall be aligned with the major or long axis of the phantom.
A.2.3 System check antenna
The phantom shall be irradiated using an antenna for the required frequency, e.g. a half-wave
dipole, CLA, VPIFA, meander dipole, patch antenna, or open-ended waveguide. The system
validation antennas (see Annex D) are typically, but not necessarily, used for the system
check. A system check antenna shall have good positioning repeatability, mechanical stability,
and impedance matching. In the following positioning instructions, a half-wave dipole is used
as an example to illustrate system check antenna positioning requirements. Similar
instructions shall be applied for other antennas.
A half-wave dipole shall be positioned below the bottom of the phantom and centred with its
axis parallel to the longest dimension of the phantom, within ±2°. The distance between the
inner surface of the phantom and the dipole feed-point, s, is specified for each test frequency.
For the reference dipoles specified in Clause D.2, a dielectric spacer is used to provide the
distance s and spacer mechanical tolerance, as given by the following:
a) s = 15 mm ± 0,2 mm, for 150 MHz ≤ f ≤ 1 000 MHz

b) s = 10 mm ± 0,2 mm, for 1 000 MHz < f ≤ 6 000 MHz
c) s = 5 mm ± 0,1 mm, for 6 000 MHz < f ≤ 10 000 MHz
– 114 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
A spacer with low loss (loss tangent < 0,05) and low permittivity (relative permittivity < 5) shall
be used to establish the correct distance between the top surface of the dipole and the bottom
surface of the phantom. Below 3 GHz, the spacer shall not change the measured 1 g and 10 g
psSAR by more than 1 %, compared to the no-spacer condition. Above 3 GHz, the spacer
could affect the SAR target value of a dipole, and shall be accounted for through additional
experimental validation (see NOTE 3 in D.2.2, Table D.2). The dipole shall have a return loss
better than 20 dB at the system check test frequency, which shall be measured annually
during system validation using a network analyser, to ensure SAR measurement uncertainty
due to power reflection remains low.
A.2.4 System check antenna input power measurement
The uncertainty of the power delivered to the antenna shall be as low as possible. This
document requires the use of a test set-up with directional couplers and power meters during
the system check. The recommended set-up is shown in Figure A.1 (which uses a half-wave
dipole as an example of a system check antenna). To limit the power drift during the
measurement, automatic input power level control may be implemented, as specified in
Annex V.
NOTE This set-up uses a dipole antenna as an example to illustrate system check antenna and measurement
configurations; same or equivalent set-up applies to other antennas.
Figure A.1 – Test set-up for the system check
The input return loss of the system check antenna shall be checked using a network analyser
(e.g. during annual performance characterization intervals) to ensure that the return loss is
greater than 20 dB. If a different antenna or a waveguide is used where a greater mismatch is
inherent to its design (e.g. waveguide specified in Clause D.3), a lower return loss is
acceptable only if it has been fully characterized to be stable. The specified return loss shall
be determined at the frequency at which the system check is performed.
IEC/IEEE 62209-1528:2020 – 115 –
© IEC/IEEE 2020
The component and instrumentation requirements are as follows.
a) The signal generator and amplifier power outputs shall be stable to within 2 % (after
warm-up) of the desired output value. The forward power supplied to the system check
antenna shall produce a psSAR of at least 0,4 W/kg. The 1 g or 10 g psSAR range is
0,4 W/kg to 10 W/kg. If the signal generator can deliver 15 dBm or more, an amplifier is
generally not necessary if it is connected to the system check antenna with a low loss
cable. Some high power amplifiers shall not be operated at a level far below their
maximum output power, e.g. a 100 W power amplifier operated at 250 mW output power
can be too noisy. An attenuator between the signal generator and amplifier is
recommended to protect the amplifier input.
b) The low pass filter inserted after the amplifier reduces the effect of harmonics and noise
from the amplifier. For most amplifiers in normal operating ranges, the filter is not
necessary.
c) The attenuator after the amplifier improves the matching and the accuracy of the power
sensor (consult the power meter manual).
d) The directional coupler (recommended −20 dB coupling coefficient) is used to monitor the
forward power for making adjustments to the signal generator output to maintain constant
forward power at PM2. The coupler shall have a return loss greater than 25 dB at the input
and output ports. A dual directional coupler shall be used when both forward and reflected
power need to be measured (e.g. when waveguides are used).
e) The power meter PM2 shall have low drift and a resolution of 0,01 dB; otherwise, the
absolute accuracy has negligible impact on the power setting to the system check antenna
(absolute calibration is not required).
f) The power meter PM1 and attenuator Att1 shall be high-quality components. These shall
be calibrated, preferably together. The attenuator (10 dB) improves the accuracy of the
power reading (some high-power heads come with a built-in calibrated attenuator). The
exact attenuation of the attenuator at the test frequency shall be known – for some
attenuators this can vary by more than ±0,2 dB from the specified value across the
frequency band of operation.
g) A fixed power level setting shall be used for PM1 and PM2 to avoid linearity and range
switching uncertainties in the power measurements. If power level is adjusted, the same
power level setting shall be used for PM1 and PM2.
h) The system check antenna shall be connected directly to the cable at location X. If the
power meter has a different connector type, high-quality adapters shall be used.
i) The insertion loss of cables, especially the cable that connects the directional coupler to
the antenna, shall be checked periodically to ensure the insertion loss is stable across
frequencies. It should be considered that a cable that works well at one frequency
(e.g. 900 MHz) might not perform equivalently at a different frequency (e.g. 5 GHz). During
system check measurements, all cable movements shall be avoided as this could cause
changes to cable loss characteristics and introduce SAR errors.
A.2.5 System check procedure
System check is a complete 1 g and/or 10 g psSAR measurement performed in accordance

with the requirements in A.2.1 to A.2.4.
System check antennas may have two pairs of target values:
a) Numerical targets SAR 1g,num and SAR 10g,num – If the antennas of Annex D are used, the
numerical targets provided in Annex D shall be used. Otherwise, the numerical targets
shall be determined in accordance with the requirements ofClause D.1.
b) Experimental targets SAR 1g,sys and SAR 10g,sys – These are measured SAR values of
specific system check antennas. They are determined using a system check test set-up
(see A.2.4) with calibrated equipment. The experimental targets shall be determined by
the calibration laboratory within two weeks of the antenna calibration. SAR 1g,sys and
SAR 10g,sys shall be provided to the user. The values of SAR 1g,sys and SAR 10g,sys shall not
deviate from the numerical targets SAR 1g,num and SAR 10g,num by more than ±10 %.
– 116 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
A.2.6 System check acceptance criteria
The measured 1 g and/or 10 g psSAR shall be normalized to 1 W by the input power of the
system check antenna (forward power for dipoles, meander dipole, VPIFAs, and CLA, and net
power for waveguides). The normalized SAR values, SAR 1g,meas and SAR 10g,meas , are
compared against the targets. The following requirements apply.
a) The measured psSAR values shall not deviate from the numerical targets by more than
the expanded measurement uncertainty for the system check antennas, 2u s , or ±15 %,
whichever is less, as described in Formula (A.1) and Formula (A.2).
SAR1g,meas − SAR1g,num SAR1g,num ≤ 2 ⋅ us,1g (A.1)
b) The measured psSAR values shall not deviate from the experimental targets by more than
±10 %, as described in Formula (A.3) and Formula (A.4).
SAR1g,meas − SAR1g,sys SAR1g,sys ≤ 10 % (A.3)
A.3 System validation
A.3.1 Purpose
The system validation procedure tests the SAR system using system validation antennas to
verify the measurement accuracy and performance of probe, readout electronics, and system
software. It is a validation of the system with respect to all performance specifications. This
set-up utilizes a flat phantom, and a system validation antenna as specified in Annex D. Thus,
this validation process does not include data scattering and device positioning uncertainty.
System validation shall be performed at least annually, when a new system is put into
operation, or whenever modifications have been made to the system, such as software
updates, using different readout electronics or probes, and after probe calibration. System
validation shall be done with a calibrated probe.
The objective of Clause A.3 is to provide a methodology for SAR measurement system
validation. Since SAR measurement equipment, calibration techniques, phantoms, and tissue-
equivalent media can vary among laboratories, a validation methodology is needed to
ascertain that uniform results are obtained within reasonable measurement uncertainties.
Numerically calculated reference SAR values for the system validation antennas specified in
A.3.3 are listed in Table D.2, Table D.4, Table D.5, Table D.7, and Table D.10. These
numerical values have also been validated experimentally. The environmental requirements
for system validation tests are as specified in 6.1.
A.3.2 Phantom set-up
The flat phantom set-up specified for the system check (see A.2.2 and Figure A.1) is also
used for the system validation tests. The system validation shall be performed using tissue-
equivalent media having dielectric properties as specified in Table 2 (see 6.2.2).
A.3.3 System validation antennas
System validation antennas are specified in Annex D.

IEC/IEEE 62209-1528:2020 – 117 –
© IEC/IEEE 2020
A.3.4 Input power measurement
The input power measurement set-up specified for the system check (A.2.4) is also used for
system validation tests.
A.3.5 System validation procedure
System validation is used to verify the measurement accuracy of a complete SAR system,
including the software algorithms. Test device positioning and phantom shape uncertainties
are not considered during system validation. The forward power shall be limited so that the
measured SAR values are within the dynamic range of the probe, to avoid probe damage.
The system validation procedure consists of up to seven steps, i.e. the following steps a) to
g). Step a) shall be done for each combination of the probe, readout electronics, and
measurement and post-processing system version, every time that system validation is
required. Tests b) to f) shall be done every time that system components have been modified
(e.g. new software release, new readout electronics, new probe or calibrations). Step g) may
be omitted if sufficient information is provided by the system manufacturer to the user that the
uncertainty budget is valid for capacitively-coupled sources. The system validation procedure
is as follows.
a) SAR evaluation: A complete 1 g and/or 10 g psSAR measurement is performed using a

system validation antenna as specified in Annex D. The input power of the system
validation antenna is adjusted to produce a 1 g and/or 10 g psSAR value within the range
of 0,4 W/kg to 10 W/kg. The 1 g and/or 10 g psSAR is measured at the frequencies in
Table D.2, Table D.4, Table D.5, Table D.7 and Table D.10. The psSAR results are
normalized as specified in Table D.2 (dipole antennas), Table D.4 (waveguides),
Table D.5 (CLAs), Table D.7 (meander dipole), or Table D.10 (VPIFAs) and compared with
the normalized numerical SAR values The differences between the measured values and
the numerical target values shall be less than the expanded uncertainty for the system
validation using the procedures of Table 9 but not more than 10 %.
b) Extrapolation routine: Local SAR values are measured along a vertical axis directly above
the centre of the system validation antenna (i.e. dipole, meander antenna or VPIFA feed-
point, centreline of waveguide, or centre of the CLA) using the same test grid-point
spacing as used for the zoom scan of the DUT SAR evaluations. The measured values are
extrapolated to the phantom surface and compared to the appropriate numerical target
value given in Table D.2 (column 5), Table D.4 (column 5 with d = 0), Table D.5 (column
5), Table D.7 (column 5) and Table D.10 (column 5). If the dipole antenna, meander dipole
or VPIFA is used, this measurement is repeated along another vertical axis with a 2 cm
transverse offset (y-direction of Figure A.1) from the feed-point. SAR values are
extrapolated to the phantom surface and compared with the normalized numerical values
given in column 6 of Table D.2, column 6 of Table D.5 and column 6 of Table D.10. The
difference between the extrapolated values and the numerical target values given in
Table D.2 (or values for the specific antenna used) shall be less than the expanded
uncertainty for the system validation using the procedures of Table 9 or 15 %, whichever
is less. Note that step b) may be performed at the same time as step a) if an enlarged
zoom scan is used in step a) to provide the extrapolated SAR values.
c) Probe linearity for CW equivalent signals: The measurements in step a) are repeated
using different input power levels. The power levels are selected for each frequency to
produce 1 g or 10 g psSAR values of approximately 10 W/kg, 2 W/kg, 0,4 W/kg and 0,12
W/kg. The measured psSAR values are normalized as specified in the appropriate tables
(see step a)) and compared with the normalized measured results in step a). The
difference between these values shall be less than the expanded uncertainty for the
linearity component using the procedures of Table 9 and 8.4.1.3 or 10 %, whichever is
less.
– 118 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
d) Probe linearity for periodic pulsed signals: This step shall be conducted after step c)
requirements are satisfied. The measurements in step a) are repeated with pulse-
modulated signals having a duty factor of 0,1 and pulse repetition rate of 10 Hz. The
power is adjusted to produce a 1 g or 10 g psSAR of approximately 8 W/kg with the
periodic pulse modulated signal or a peak SAR of approximately 80 W/kg. The measured
SAR values are scaled to a duty factor of 1, then normalized as specified in the
appropriate tables (see step a) and compared with the normalized measured results in
step a). The difference between these values shall be less than the expanded uncertainty
for system validation using the procedures of Table 9 or 10 %, whichever is less.
e) Probe linearity for non-pulsed signals (e.g. CDMA and OFDM signals): The measurements
in step a) are repeated with the specific operating modes that will be tested by the
laboratory. This test may be done at just one frequency per operating mode. The power is
adjusted to produce a 1 g psSAR of approximately 8 W/kg. If the peak-to-average power
ratio of the operating mode is high enough such that the peak local SAR exceeds the
dynamic range of the system, the input power should be reduced to produce a 1 g psSAR
lower than 8 W/kg but still as high as possible. The measured SAR values are scaled to a
duty factor of 1, then normalized as specified in the appropriate tables (see step a) and
compared with the normalized measured results in step a). The difference between these
values shall be less than the expanded uncertainty for system validation using the
procedures of Table 9 or 10 %, whichever is less.
NOTE 1 A more detailed procedure to validate the system for all operating modes is planned for a future
revision of this document through an amendment, revision, or accompanying Technical Report. The procedure
will include steps to ensure that the bandwidth of the validation source is sufficient to cover the channel
bandwidth and any carrier aggregation modes.
f) Probe axial isotropy: The geometric centre of the sensors of the probe is placed directly
above the centre of the system validation antenna at a measurement distance of
approximately one probe tip diameter from the inner phantom surface. The probe is
rotated around its axis ±180° in steps no larger than 15°. The maximum and minimum SAR
readings are recorded. The difference between these values shall be less than the
expanded uncertainty for the axial isotropy component using the procedures of Table 9
and 8.4.1.5 or 5 %, whichever is less.
g) SAR evaluation for E-field component normal to the phantom surface: A complete 1 g
and/or 10 g psSAR measurement is performed using the VPIFAs specified in Clause D.5.
The zoom scan for frequencies above 3 GHz specified in 7.4.2 should be used, regardless
of the operating frequency of the VPIFA. The results are normalized to 1 W forward power
and compared against the numerical reference values in Clause D.5. The purpose of this
validation is to test the accuracy of the measurement system using antennas with a
dominant normal E-field component. This step tests the system implementation and the
probe design for suitability for capacitively coupled fields and can be performed by the
manufacturer of the measurement system and need not be repeated by the test laboratory
as long the appropriate documentation from the manufacturer is available. The differences
between the measured values and the numerical target values shall be less than the
expanded uncertainty for the system validation but not more than 10 %.
NOTE 2 The system validation procedure is neither an alternative to probe calibration nor to the uncertainty
estimation of Clause 8. The probe and readout electronics are calibrated regularly according to the procedures
given in Annex E. Probe hemispherical isotropy is not considered in the protocol for system validation.
For frequencies below 300 MHz, appropriate validation antennas specified in Annex D shall
be used to perform system validation procedure.
IEC/IEEE 62209-1528:2020 – 119 –
© IEC/IEEE 2020
A.4 Fast SAR testing system validation and system check
A.4.1 General
Class 1 and Class 2 fast SAR testing methods specified in 7.9.2 are based on assumptions or
approximations that are typical of each fast SAR method and that deviate from the standard
measurement procedure. For a given SAR measurement system, since fast SAR testing is
based on simplifications and approximations specific to each approach, system check and
system validation methods required for full SAR testing have been considered insufficient for
dealing with 7.9.2 fast SAR methods. New system check and system validation methods are
hence introduced.
Subclause A.4.2 describes the system validation that shall be done by the method developer.
It defines a set of minimal requirements or principles to be followed by the fast SAR method
developer to validate that the method has been implemented correctly. These requirements
are also minimum requirements applicable to fast SAR methods specified in Annex C. The
developer shall clearly document the computations, measurements and other assessments
performed to validate the fast SAR method for users to determine how to apply the specific
method for SAR evaluation. Any additional analyses or conditions the user shall satisfy to
apply the fast SAR method shall be clearly specified.
Subclause A.4.3 describes the fast SAR method system check the user shall perform. For a
given SAR measurement system, a single system check antenna may be used for both full
SAR testing and fast SAR testing using the procedures in Clause A.2 and A.4.3.
A.4.2 Fast SAR testing system validation
Fast SAR testing system validation is a way to quantify the validity and reliability of the
method. Since the validation and the measurement uncertainty are both determined by the
developer of the fast SAR testing method, the system validation serves also as to confirm the
uncertainty budget. The following principles are applied for system validation of a Class 2 fast
SAR method.
a) Multiple radiators with SAR distributions suitable for validating the method shall be used.
The radiators may be wireless DUTs or other antennas or waveguides that comply with
the requirements of Clause A.4. Multiple samples (e.g. one sample for each band) can be
used for fast SAR testing system validation. The developer shall demonstrate that the fast
SAR testing method is not biased by the selected radiators or SAR distributions in
validating or calibrating the method to such radiators.
b) The system validation shall be performed over the full range of operating frequencies,
operating modes, power levels (e.g. 10 g psSAR between 0,1 W/kg and 10 W/kg), device
operating configurations, exposure conditions, and SAR distributions that are specified
with the method and used for device evaluation.
c) The radiators chosen shall cover the range of SAR distributions (e.g. distributions with a
SAR which is larger than −3 dB of a local maximum within a radius ranging from 5 mm to
50 mm; dominant polarization both normal and parallel to the phantom surface; locations
within the measurement area specified for the fast SAR testing method; and number of
SAR peaks), supported by the fast SAR testing method. The fast SAR testing method
developer shall specify the radiators used to validate the particular fast SAR testing
method and provide the necessary validation details to ensure users can apply the method
correctly.
d) The radiators shall be chosen so that the measurement uncertainty during system
validation is not strongly dependent on or dominated by the radiator (e.g. positioning,
uncertainty, or the validation antenna), and phantom configurations.
e) A minimum of 10 configurations shall be tested, per frequency band, so that there is high
confidence in the system validation. This could be achieved, for example, by using two
radiators at five different distances or orientations.
– 120 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
System validation protocols and procedures, and validation antennas, for use with vector
measurement-based systems are given in IEC 62209-3. The validation guidance in
IEC 62209-3 shall be applied for the purposes of testing in accordance with Annex A.
SAR uncertainty for fast SAR testing system validation is determined by comparing the psSAR
determined with the fast SAR testing method, SAR fast,i with the target SAR of full SAR testing,
SAR full,i , for the i-th test configuration (radiator, frequency, SAR distribution). The standard
deviation of the difference between the two methods is determined according to
Formula (A.5), Formula (A.6), and Formula (A.7):
N
1 2
SARtol
= %
N ∑ ( ∆i − µ ) (A.5)
i =1
 SARfast,i − SARfull,i 
∆i % =100 ×   (A.6)
 SARfull,i 
 
N
1
=µ
N ∑ ∆i (A.7)
i =1
The system validation shall be applied separately for each frequency band and operating
mode supported by the fast SAR testing method, and as used for device evaluation.
The standard deviation of Formula (A.6) shall be within the standard uncertainty of the fast
SAR testing method (Table 9 and Annex C). Otherwise, the developer of the fast SAR testing
method shall either revise the combined uncertainty of the fast SAR testing method in
accordance with the system validation results, or consider whether the fast SAR testing
method is applicable.
The fast SAR testing system validation shall be performed when modifications to the fast SAR
testing or post-processing algorithms are made. A summary of the developer’s system
validation methods and results shall be described in the SAR report, to support the test
results.
A.4.3 Fast SAR testing system check
A.4.3.1 General
SAR system check is an easy way for laboratories to verify the reliability and consistency of
fast SAR testing implemented in a SAR system. The fast SAR testing system check
requirements are similar to the system check of full SAR testing, as specified in Clause A.2.
The following principles apply for SAR system check.
a) It is applied separately for each frequency band in which components of the system (e.g.
probe, tissue-equivalent medium) are different.
b) It shall be performed within 24 hours before fast SAR testing of the DUT.
c) It shall be performed on the same measurement system that is used for the fast SAR
testing of the DUT.
d) It is performed using a single system check antenna (either the dipole antennas or
waveguides and the set-up specified in Clause A.2, or another well characterized
radiator). The fast SAR testing reference value for the radiator shall be determined for the
fast SAR testing system check set-up, and a measurement uncertainty budget for the
system check shall be provided.
IEC/IEEE 62209-1528:2020 – 121 –
© IEC/IEEE 2020
A.4.3.2 Acceptance criteria for fast SAR testing system check
The fast SAR testing system check measurement result is compared to the fast SAR testing
reference value for the same radiator and frequency band. The acceptance criteria are the
same as those for the full SAR system check.
a) Measured values, SAR 1g,meas and SAR 10g,meas , are compared to the numerical targets
SAR 1g,num and/or SAR 10g,num . The difference is within two standard deviations of the fast
SAR method system standard uncertainty u s as described in Formula (A.8) and
Formula (A.9) :
b) All measured values are within ±10 % of the relative targets SAR 1g,sys and SAR 10g,sys for
that fast SAR testing method measurement system using the system check antenna, and
same system or system type as described in Formula (A.10) and Formula (A.11):
Otherwise, the fast SAR testing system shall not be used for compliance testing purposes.
– 122 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex B
(informative)
SAR test reduction supporting information
B.1 General
Annex B provides useful supporting information related to the test reduction procedures
described in 7.9.3.
Clause B.2 is related to the test reduction procedure based on characteristics of DUT design,
while Clause B.4 deals with the test reduction procedure based on SAR level analysis. Finally,
Clause B.5 describes other possible statistical approaches to identify high SAR test
configurations. Clause B.5 provides general information rather than detailed procedures. For
this reason, the approaches specified in Clause B.5 generally require considerable user input
to determine which test configurations could be effectively excluded. In conclusion, it is
strongly recommended not to use the statistical approaches specified in Clause B.5 for
compliance testing purposes. In principle, they can however be used for early investigation
purposes (e.g. pre-compliance during research and development stages).
B.2 Test reduction based on characteristics of DUT design
B.2.1 General
When the first editions of IEEE Std 1528 and IEC 62209 were developed, there were many
DUTs with extendable antennas. It was considered that the tilt position could induce a higher
SAR in the head in comparison to the touch or cheek position, because tilting the DUT could
decrease the separation distance between the extendable antenna and the head. However,
the vast majority of the latest DUT designs contain internal antennas. When the antenna is
located at the bottom of the DUT, in contrast the tilt position will increase the separation
distance between the antenna and the head. In addition, the contributions to the SAR in the
head from RF currents that flow on the surface of the DUT will be decreased when the phone
is in the tilt position, again because of the increased separation.
B.2.2 Statistical analysis overview
B.2.2.1 Aims and hypothesis
After years of testing, experience has shown that for internal antennas the SAR value in the
tilt position is generally lower than the SAR value at cheek in the same test configuration. The
aim of the supporting investigation was to conduct a probability analysis of the relationship
between the SAR measured in the tilt position versus the cheek position for the same test
configurations to ascertain if, and under what conditions, testing in the tilt position could be
eliminated.
The hypothesis for this study was that there is a very low probability that the SAR value in the
tilt position could exceed the SAR in the cheek position by more than a factor of two, and
therefore, the SAR in the tilt position could not exceed the applicable SAR limit, and the
device is expected to be compliant with the SAR regulation if the SAR in the cheek position is
3 dB below the applicable SAR limit.
IEC/IEEE 62209-1528:2020 – 123 –
© IEC/IEEE 2020
B.2.2.2 DUTs under analysis
The analysis was undertaken with 139 DUTs from two manufacturers (which were
predominantly year 2009 models). The analysis included DUTs with various antenna
locations, device types and multiple operating frequency bands. In most cases, both 1 g and
10 g psSAR values were measured, and therefore the total number of SAR measurements
used for the analysis was actually larger than the number of DUTs (139). Table B.1 shows the
basic characteristics of the DUTs used for the study.
Table B.1 – The number of DUTs used for the statistical study
Manufacturer Manufacturer
DUT type Total
A B
Total 112 27 139
Top antenna 30 14 44
Bottom antenna 82 13 95
Slider, slide open 29 4 33
Slider, slide closed 58 8 66
Clamshell 23 4 27
Bar 60 19 79
B.2.3 Analysis results
B.2.3.1 Overall analysis results
Subclause B.2.3 shows the result of the statistical analysis with 270 test results including 1 g
and 10 g psSAR values of all test devices regardless of DUT types, antenna locations, and
manufacturers. Figure B.1 shows the distribution of ‘Tilt/Cheek’ values for the test devices.
The mean and standard deviation values are 0,52 and 0,19, respectively.
Table B.2 shows the probability that the SAR in the tilt position over the SAR in the cheek
position is higher than certain values (x) analysed by using normal distribution and cumulative
probability analysis, based on mean and standard deviation identified with sample data shown
in Figure B.1. While the actual distribution was close to a gamma distribution, the analysis has
been undertaken assuming a normal distribution, because this gives an upper bound for the
probability function and simplifies the calculation significantly.
The analysis result shows that the probability that the SAR at tilt is higher than twice the SAR
at cheek is 4,330 × 10 −15 , which can be considered as zero probability. Even the probability
that the SAR at tilt is higher than the SAR at cheek is 6,057 × 10 −3 , which is still very low
probability.
– 124 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure B.1 – Distribution of Tilt/Cheek
Table B.2 – Statistical analysis results of

P(Tilt/Cheek > x) for various x values
x P(Tilt/Cheek > x)
1,0 6,057 × 10 −3
1,1 1,208 × 10 −3
1,2 1,866 × 10 −4
1,3 2,221 × 10 −5
2,0 4,330 × 10 −15
B.2.3.2 Comparison between 10 g and 1 g psSAR
Table B.3 is the analysis result showing the differences between 1 g psSAR and 10 g average
SAR. It shows even for 1 g average SAR, the probability that the SAR at tilt is higher than
twice the SAR at cheek is 6,428 × 10 −14 .

P(Tilt/Cheek > x) for 1 g and 10 g psSAR
P(Tilt/Cheek > x)
x
10 g psSAR 1 g psSAR
1,0 4,307 × 10 −3 4,307 × 10 −3
1,1 7,122 × 10 −4 7,122 × 10 −4
1,2 8,759 × 10 −5 8,759 × 10 −5
1,3 7,978 × 10 −6 7,978 × 10 −6
2,0 1,110 × 10 −16 1,110 × 10 −16

IEC/IEEE 62209-1528:2020 – 125 –
© IEC/IEEE 2020
B.2.3.3 Comparison between DUTs with bottom antenna and top antenna
Table B.4 is the analysis result showing the differences between DUTs with bottom antennas
and with top antennas. It shows the probability that the SAR at tilt is higher than twice the
SAR at cheek is 5,217 × 10 −9 even for devices with top antennas, which have been
considered before to introduce higher SAR value at tilt than at cheek for DUTs with
extendable antennas.

P(Tilt/Cheek > x) for various antenna locations
P(Tilt/Cheek > x)
x
Bottom antenna Top antenna
1,0 1,718 × 10 −3 3,725 × 10 −2
1,1 2,095 × 10 −4 1,472 × 10 −2
1,2 1,815 × 10 −5 5,062 × 10 −3
1,3 1,112 × 10 −6 1,511 × 10 −3
2,0 0,0 a 5,217 × 10 −9
a This number is below the numerical precision of the

computer used to perform statistical analysis.
B.2.3.4 Comparison among frequency bands
Table B.5 is the analysis result showing the differences among operating frequency bands. It
shows the probability that the SAR at tilt is higher than twice the SAR at cheek is 3,492 × 10 −7
even for higher frequency bands.

P(Tilt/Cheek > x) for various frequency bands
P(Tilt/Cheek > x)
x
850 MHz 900 MHz 1 800 MHz 1 900 MHz
1,0 7,638 × 10 −3 4,578 × 10 −3 6,536 × 10 −2 1,920 × 10 −2
1,1 1,526 × 10 −3 6,350 × 10 −4 3,170 × 10 −2 7,076 × 10 −3
1,2 2,334 × 10 −4 6,172 × 10 −5 1,385 × 10 −2 2,283 × 10 −3
1,3 2,721 × 10 −5 4,180 × 10 −6 5,442 × 10 −3 6,435 × 10 −4
2,0 3,331 × 10 −15 0,000 × 10 0 3,492 × 10 −7 1,839 × 10 −9
B.2.3.5 Comparison among device types
Table B.6 is the analysis result showing the differences among device types. It shows the
probability that the SAR at tilt is higher than twice the SAR at cheek is very low, regardless of
device types.
– 126 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
P(Tilt/Cheek > x) for various device types
P(Tilt/Cheek > x)
x
Bar Clamshell Slide (open) Slide (closed)
1,0 6,853 × 10 −3 2,992 × 10 −8 5,441 × 10 −4 1,771 × 10 −3
1,1 1,334 × 10 −3 3,769 × 10 −10 2,138 × 10 −5 1,739 × 10 −4
1,2 1,982 × 10 −4 2,801 × 10 −12 4,385 × 10 −7 1,131 × 10 −5
1,3 2,240 × 10 −5 1,221 × 10 −14 4,658 × 10 −9 4,844 × 10 −7
2,0 1,998 × 10 −15 0,000 × 10 0 0,000 × 10 0 0,000 × 10 0
B.2.4 Conclusions
As a result of statistical analysis with the latest DUT designs and models (i.e. as of 2009), the
probability that the SAR at tilt is higher than the SAR at cheek is very low, and very close to
zero, which means the SAR at tilt is most likely never higher than the applicable SAR limit if
the SAR at cheek is 3 dB lower than the SAR limit. This approach will ensure that the
handsets are within the respective SAR limits, but does not ensure that the maximum SAR will
be measured. Therefore if the main goal is to show compliance with the respective SAR limit,
the requirement for the testing of DUTs in the tilt position can be excluded under the
conditions reported in 7.9.3.3.
B.2.5 Expansion to multi-transmission antennas
The analysis for this study has used hand-held devices with a single transmitting antenna.
However the conclusion can also be applied to hand-held devices with multiple transmitting
antennas. SAR measurement method applicable to DUTs with multiple transmitting antennas
is based on summation of SAR distributions from each transmitting antenna. Therefore, it can
be concluded that the SAR at tilt even with multiple transmitting antennas is not higher than
twice of SAR at cheek with the same antennas, where the SAR at tilt or cheek is calculated by
summing SAR distributions from multiple transmitting antennas.
B.3 Test reduction based on analysis of SAR results on other signal

modulations
B.3.1 General
When the first editions of IEEE Std 1528 and IEC 62209 were developed, DUTs typically used
only one operating mode per frequency band. The standardized test procedure required each
operating mode for a given frequency band to be fully tested. Current generation DUTs now
have several operating modes in use for a given frequency band, e.g. GSM/GPRS, CDMA,
WCDMA (3G), LTE (4G) and 5G NR. Therefore, the amount of testing for each frequency
band has increased considerably. It is expected that this trend will continue for future DUTs,
unless test reduction procedures can be applied.
Generally, signal modulation will have little effect on the SAR distribution if all signals are
transmitted using the same antenna, antenna feed-point, and matching circuit (see full list of
requirements in at the end of B.3.1). This is also true concerning the currents on the body of
the transmitting device, because the same antenna excites the same currents regardless of
the signal modulation used.
It is usually enough to fully measure one signal modulation, mod X , in all

positions/configurations, then the other signal modulations mod Y , mod Z , etc., in the
position/configuration giving highest psSAR result for the signal modulation mod X. The
preceding consideration applies if the following parameters are the same for all the
transmitted signal modulations:
IEC/IEEE 62209-1528:2020 – 127 –
© IEC/IEEE 2020
• frequency band;
• frequencies of the required test channels;
• antenna radiating element;
• antenna feed-point;
• antenna grounding;
• antenna matching circuitry (if any);
• antenna tuner and tuner settings (if any).
If the design of the phone is not sufficiently known, i.e. it is not known if the preceding
conditions are met, then test reductions cannot be applied and full SAR testing should be
performed for all signal modulations.
B.3.2 Analysis results
Studies described in B.3.2 and [37] provide the rationale for performing the SAR
measurement for other signal modulations mod Y , mod Z , etc., at the same
position/configuration yielding the maximum psSAR level for signal modulation mod X . For
example, if the left cheek position yields the highest psSAR value for GSM transmission, the
left cheek position will typically generate the highest psSAR value for other modulations such
as WCDMA and LTE operating within the same frequency band.
Minor deviations from this rule can be explained by the measurement uncertainties associated
with testing. For instance, a shift in the maximum SAR value from the left cheek position to
the right cheek position when the SAR distribution is highly symmetrical along the phone’s
vertical axis will be dependent on the positioning accuracy. Deviations are often seen in cases
where the SAR levels for both cheek and tilt positions are very low (less than 0,4 W/kg). This
is due to decreasing sensitivity of the measurement system, which introduces noise to the
measurement results.
To decrease the possibility of not obtaining the maximum psSAR value of a DUT, a psSAR
value threshold was included in the test reduction procedure (7.9.3.4) of this document.
Namely, if the maximum psSAR values were obtained at the cheek positions and the psSAR
values for the tilt positions are at least 30 % below the applicable SAR limit for the signal
modulation mod X, measurements for the tilt positions are not required for other signal
modulations within the same frequency band.
The analysed data consist of a total of 181 DUT and frequency band combinations. The DUT
models with multiple operating modes within the same frequency band were selected from
different manufacturers. The data consisted of a total of 115 DUTs and frequency band
combinations operating below 1 GHz, and 66 DUT and frequency band combinations
operating above 1 GHz, altogether representing a total of 181 cases.
The data were divided into two categories, because handsets operating below 1 GHz have
higher electromagnetic fields from RF currents flowing along the chassis of the device, while
for handsets operating above 1 GHz the currents are concentrated close to the antenna
element and chassis. Based on current antenna design trends in multimode handsets, the
devices were assumed to have bottom antennas (as specified in 7.9.3.3 of this document)
fulfilling all the conditions listed in B.3.1. Frequency bands that had at least two different
operating modes were analysed. Each frequency band generally had two or three different
operating modes. Results for the 850 MHz, 900 MHz, 1 700 MHz, 1 800 MHz, 1 900 MHz, and
1 950 MHz frequency bands were analysed.
The 1 g or 10 g psSAR results for all test positions (left cheek, left tilt, right cheek, and right
tilt) were collected. The averaging volume was the same for each phone and frequency band
combination.
– 128 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
For each frequency band and operating mode, all psSAR values were divided by the
maximum psSAR value found for the GSM mode. The GSM mode was chosen as the
baseline, because it was available in all the analysed test cases. This means that if any
relative SAR values for other operating modes (e.g. WCDMA or LTE) were greater than 1, the
maximum SAR for that operating mode was measured in a different test configuration than for
GSM.
In Figure B.2, the x-axis represents all results with the maximum absolute SAR value of each
measurement, and the y-axis represents the SAR value relative to maximum SAR value for
the GSM mode. The results are divided by cheek and tilt position below 1 GHz and above
1 GHz. Figure B.2 shows that the majority of the results are below 1, i.e. the maximum SAR
was found in same position for example for LTE as it was found for GSM.
Among the 181 test cases, the maximum psSAR was found in a different position than for
GSM in one of the other operating modes for only 19 cases. In all 19 cases, the maximum
psSAR values were less than 0,4 W/kg for both GSM and the other operating modes. Of these
19 cases, only three cases occurred for tilt positions, with a ratio less than 1,2. In the
remaining 162 cases where the psSAR values were greater than 0,4 W/kg, the maximum SAR
test configuration was correctly identified based on GSM tests.
Figure B.2 – SAR relative to SAR in position with maximum SAR in GSM mode
B.4 Test reduction based on SAR level analysis
B.4.1 General
In the SAR measurement procedure specified in 7.4, the results from only a single zoom scan
are used to obtain psSAR to be compared with basic restrictions. Consequently, a
considerable number of zoom scans could be avoided, saving measurement time, if there is a
criterion able to forecast that the psSAR for a test configuration will be lower than the highest
value of psSAR among all test configurations. By extending this concept to multi-band and
multi configuration devices, the measurement time can be further reduced by defining a SAR
measurement session not limited by the single band operational mode, but extended to all the
operational bands and configurations supported by the device and included in the scope of
the present test reduction method. The basic idea is to avoid performing the zoom scan
session for each single test (i.e. 7.4.2 d)) if the maximum measured SAR value of the area
scan is below a threshold. The threshold is chosen so that the highest psSAR value will be:
IEC/IEEE 62209-1528:2020 – 129 –
© IEC/IEEE 2020
• correctly assessed and not underestimated, especially when it could generate a psSAR
close to the basic restriction limit;
• correctly identified, even if not executing the zoom scan session.
In order to specify the value of the threshold satisfying the first requirement, the psSAR from
the zoom scan measurement has been correlated to the maximum measured SAR value found
in the corresponding area scan [38]. It was observed that if the maximum measured SAR
value in the area scan is lower than a threshold, then the basic limit for the psSAR is never
exceeded; in particular there was a good correlation between the maximum measured SAR of
the area scan and the 1 g psSAR. In [38], choosing a threshold of 1,3 W/kg (80 % of basic 1 g
limit of 1,6 W/kg) gives a high probability that the limit for the psSAR is never exceeded. The
same threshold can be used also for the evaluation of the 10 g psSAR.
The statistical analysis described in B.4.2 has been performed to address the question of
correct identification of the highest peak spatial average SAR [39].
B.4.2 Statistical analysis
The statistical analysis presented here considers DUTs operating in the GSM 900 / DCS 1800
and UMTS I bands. The area scan grid spacing parameters used in this analysis were both (x
and y) set to 10 mm, so they fit the requirements of this document, and area scans were
performed at a specified fixed distance (4 mm spacing between sensors and SAM surface in
this case) from the SAM internal surface.
Because the maximum SAR value could not be exactly identified from the area scan, due to
the fact that it could be located (in x and y) inside a square region having 10 mm size, it is
recommended to first experimentally evaluate the spatial gradient in order to estimate the
magnitude of the peak SAR value that could be missed in applying only the area scan. To do
this, 420 SAR distributions in the GSM 900 band, 420 distributions in the DCS 1800 band, and
300 distributions in the UMTS I band have been considered. The distributions are relevant to
mobile phones marketed between 2007 and 2010. All handset form factors, including
clamshell and slider, have been considered in this study. The first step was to identify the
“iso-level” as the set of points having a fixed SAR value, and then compute the minimum
distance D min between the “iso-level” at a fixed SAR level and the position of the interpolated
maximum SAR (SAR max , the distance between the two points marked with an X in Figure B.3).
Figure B.4 shows the histogram related to the random variable D min in the case of GSM 900
band and the 0,6 × SAR max level. The red curve shows a normal probability density function
(PDF) fitted to data using a maximum likelihood estimation procedure.
Even if in some cases data fail the normality hypothesis test (e.g. Jarque-Bera normality test)
and data can be more efficiently fitted by using other PDF (e.g. hyperbolic PDF), for the
purpose of this analysis only normal fits have been considered in this work. Considering
*
normal PDF, the value Dmin represents the value for which the probability to have D min less
*
than Dmin is 5 %.
Table B.7 summarizes all the analysis performed, for each frequency band considered in this
study. In this table the iso-level is expressed as a percentage, so a value of 90 % means the
iso-level at 0,9 × SAR max .
– 130 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure B.3 – Two points identifying the minimum distance between the
position of the interpolated maximum SAR and the points at 0,6 × SAR max
Figure B.4 – Histogram for D min in the case of GSM 900 and iso-level at 0,6 × SAR max
*
Table B.7 – Distance Dmin for various “iso-level” values
GSM 900 DCS 1800 UMTS I

* * *
Iso-level (%) Dmin Iso-level (%) Dmin Iso-level (%) Dmin
90 4,23 90 3,04 90 1,82

80 6,74 80 4,82 80 4,63
75 7,86 70 6,19 70 6,10
70 8,90 65 6,92 65 6,81
60 10,96 60 7,54 60 7,48
50 13,15 50 9,06 50 8,88
IEC/IEEE 62209-1528:2020 – 131 –
© IEC/IEEE 2020
To summarize the results of this analysis, considering for example the GSM 900 band, it can
be concluded that is improbable to find SAR values having the value 0,75 × SAR max if a
distance of 7,86 mm from the SAR max point is considered. In particular, the values shown in
*
bold in Table B.7 have Dmin values higher than 7 mm, which is the value in the middle of the
10 mm chosen grids. In other words, still considering GSM 900 as an example, it is
improbable to miss a SAR value inside a 10 mm area scan grid that is higher than 0,75 times
the maximum SAR value measured during the area scan.
To conclude the statistical analysis, the correlation between the measured maximum
measured SAR value of the area scan and the psSAR values found after zoom scans has
been investigated. To do this, a database of 100 handsets operating in the GSM 900 and
DCS 1800 bands, and 50 DUTs operating in the UMTS I band, have been considered.
For each handset the maximum measured SAR value of the area scan, the 1 g psSAR and the
10 g psSAR, as measured after the complete measurement procedure were collected.
Figure B.5 shows the histogram, as well as the normal PDF fit for the random variable
Factor 1g,1800 , designated as the ratio between the psSAR for 1 g mass and the maximum
measured SAR value of the area scan in case of DCS 1800 band.
In this particular case, it can be shown that the 95 % probability is achieved to find
Factor 1g,1800 in the interval [0,859, 1,015]. In this particular case, it can be concluded that if
there are two area scans with two different measured peak SAR values (not the interpolated
one), and they differ from a factor of 0,859/1,015 = 0,84, then it is improbable that the
distribution having the higher maximum SAR value measured during the area scan will have
the lower 1 g psSAR. In this particular case, 0,84 can be assumed as an experimental
threshold: if two different area scans are considered and the measured peak SAR values
measured for each one differ by a factor lower than the threshold, then there is a 95 %
probability that the area scan with the higher SAR will yield the higher psSAR.
Similar conclusions can be obtained analysing the other bands, as well as the 10 g psSAR.
Table B.8 summarizes the overall results.
In conclusion, the same thresholds found in Table B.7 (75 % for GSM 900, 60 % for
DCS 1800 and UMTS I) are sufficient also to determine (with a 95 % probability value) the
fact that the psSAR will be correctly identified even without performing a zoom scan.
– 132 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure B.5 – Histogram for random variable Factor 1g,1800
Table B.8 – Experimental thresholds to have a 95 % probability that the

maximum measured SAR value from the area scan will also have a psSAR
GSM 900 DCS 1800 UMTS I

Threshold 1 g Threshold 10 g Threshold 1 g Threshold 10 g Threshold 1 g Threshold 10 g
0,86 0,76 0,84 0,63 0,74 0,62
B.4.3 Test reduction applicability example
Two examples of application of the procedures introduced in B.4.2 are described in the
following paragraphs. In the first, the test reduction method is applied to a single operational
mode (i.e. GSM 900), while in the second a dual mode (GSM 900/DCS 1800 and UMTS I)
DUT is considered.
Table B.9 and Table B.10 show the reported SAR values, which are the maximum measured
SAR values from the area scan (GSM 900 band).
Because this example is related to GSM 900, a threshold of 0,75 should apply. For the first
position (left cheek, middle uplink frequency) the zoom scan is needed because the maximum
SAR value is 1,3 W/kg (80 % of 1,6 W/kg), while for the left tilt 0,6 W/kg is lower than
0,75 × 1,3, so the zoom scan is not needed. Then in the right cheek position the zoom scan is
needed, and the new absolute peak SAR (APS) becomes 1,5 W/kg. This new APS precludes
the need to perform the zoom scan for right tilt (middle frequency) and right cheek (lower
frequency), while it is not high enough to exclude the last zoom scan (right cheek, higher
frequency).
For the second example, GSM 900 band is tested before DCS 1800 and UMTS I, but in
general the order in which the bands are tested may be different.
IEC/IEEE 62209-1528:2020 – 133 –
© IEC/IEEE 2020
In this second example, the procedure applied in GSM 900 is repeated considering all the
other tests and previous APS found. For each band the user should choose an appropriate
threshold (B.4.2). For the first measurement in a different band, the user will use, for the
comparison with the maximum measured peak SAR value from the area scan, the previous
APS found and the threshold related to the new band. As an example, the first area scan
measurement of DCS 1800 has a peak maximum SAR value of 0,9 W/kg that is equal to 60 %
(the threshold of DCS 1800) of 1,5 W/kg (the previous APS), so in this case zoom scan is
needed.
As can be seen in this example, nine measurements (50 %) do not require a zoom scan to be
performed (i.e. bold rows in Table B.10). The number of zoom scans that can be precluded
depends on maximum measured peak SAR values from the area scans.
Table B.9 – SAR values from the area scan (GSM 900 band): Example 1
a
Frequency Position Maximum SAR value [W/kg]
Middle Cheek left 1,3
Middle Tilt left 0,6
Middle Cheek right 1,5
Middle Tilt right 0,8
Lower Cheek right 1,0
Higher Cheek right 1,3
a For bold rows, according to the proposed test reduction
procedure, a zoom scan is not required.
Table B.10 – SAR values from the area scan (GSM 900 band): Example 2
a
Frequency Position Maximum SAR value [W/kg]
Middle (GSM 900) Cheek left 1,3
Middle (GSM 900) Tilt left 0,6
Middle (GSM 900) Cheek right 1,5
Middle (GSM 900) Tilt right 0,8
Lower (GSM 900) Cheek right 1,0
Higher (GSM 900) Cheek right 1,3
Middle (DCS 1800) Cheek left 0,9
Middle (DCS 1800) Tilt left 0,6
Middle (DCS 1800) Cheek right 0,8
Middle (DCS 1800) Tilt right 0,7
Lower (DCS 1800) Cheek right 1,1
Higher (DCS 1800) Cheek right 0,8
Middle (UMTS I) Cheek left 1,1
Middle (UMTS I) Tilt left 0,7
Middle (UMTS I) Cheek right 1,4
Middle (UMTS I) Tilt right 0,6
Lower (UMTS I) Cheek right 1,2
Higher (UMTS I) Cheek right 1,4
a For bold rows, according to the proposed test reduction procedure, a
zoom scan is not required.
– 134 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
B.5 Other statistical approaches to search for the high SAR test
configurations
B.5.1 General
A device could operate in different transmission modes and could be usable with several
antenna options, battery options and other accessories, and the number of possible
combinations can be very large. Methods are therefore needed to streamline the
measurement process, so that the highest SAR test configurations can be quickly identified.
For example, a device with two antenna configurations (antenna extended and retracted) and
four battery types testing all possible combinations would result in a considerable number of
tests. It is unnecessary to test all possible combinations; statistical techniques can be used to
show trends from a smaller set of data and determine which device-accessory combinations
result in higher SAR values.
Three traditional methods of experimentation are described in B.5.3 and B.5.4. These
methods have drawbacks compared to the design-of-experiments (DOE) approach specified
in B.5.2.
B.5.2 Test reductions based on a DOE
Analysis of SAR data, e.g. statistical analysis based on a DOE approach, may be used to
develop scientific or engineering rationales for the test reduction of certain SAR tests. For
example, if devices are available with optional faceplates with paint coatings of varying metal
content, statistical analysis of SAR data may be used to justify excluding the testing of
faceplates with less than a certain amount of metal content. The application of the test
reduction should be limited to products that are sufficiently similar to the original product for
which the test reduction was determined.
Using a design of experiments (DOE) is the preferred statistical method of achieving test
reductions. A DOE is a structured, organized method for analysing the influence of factors
and the interactions between factors on the output of a process. The DOE approach is
extensively covered in the literature [40].
B.5.3 One factor at a time (OFAT) search
With this method, the experimenter starts with a baseline test configuration, and then
successively varies one factor at a time while holding all other factors constant. For example,
this could be achieved by first varying antenna configurations, then battery types, then carry
accessory types, then audio accessory types. At the end of each step, the factor giving the
highest SAR is selected for the next steps. The main drawback of this approach is that it does
not consider any interactions between the different accessory types (e.g. the interaction of the
battery and the antenna on SAR, which is not explained by the influences of each factor
independently). If interactions exist, the OFAT approach might not find the optimum (i.e.
highest SAR) solution.
B.5.4 Analysis of unstructured data
A common source of unstructured data is historical data. This data typically was collected
without any specific objective in mind, or it could have been collected for different purposes
than the current experimental objective. This data may be useful in spotting trends, but it
might be very difficult to have high confidence in the findings.
For this reason, any findings from the analysis of unstructured data should be verified (e.g.
using a DOE).
IEC/IEEE 62209-1528:2020 – 135 –
© IEC/IEEE 2020
Annex C
(informative)
Measurement uncertainty of results obtained from

specific fast SAR testing methods
C.1 General
Among the existing and known fast SAR testing methods, many of them focus on reducing
probe movements, especially by using scanning techniques or a probe array, and employing
field reconstruction techniques to determine the spatial distribution of the SAR and the
psSAR. Probes of various types may be used (e.g. probes measuring only two tangential field
components, or thermal probes). Some of the systems also use specific probe technologies
and electronics allowing measurement of the phase of the electric field in a given plane.
Different types of dedicated post-processing algorithms can then be used to reconstruct the
volumetric SAR distribution inside the phantom.
NOTE Field reconstruction techniques are described in IEC 62209-3.
The dielectric properties of the tissue-equivalent medium used for fast SAR testing should
comply with the requirements of Annex G. The tissue-equivalent medium may be liquid, gel,
or solid. If the tissue-equivalent medium of the fast SAR testing system is accessible and
replaceable by the user, the dielectric properties should be characterized as specified in
7.2.1. If the tissue-equivalent medium is inaccessible or not replaceable by the user, the
system manufacturer should demonstrate and document compliance of the dielectric
properties with the requirements of Annex G, including any effects of spatial variations of the
dielectric properties (C.2.11) or aging of the material.
The fast SAR testing methods that are specified in Annex C use the SAR measurement
procedure specified in 7.9.2. They also use the system validation and system check specified
in Clause A.4. The full range of sensor locations should also be considered for the type of fast
SAR testing systems specified in Annex C. These ranges should be provided with the system
validation results. Annex C is informative because the system validation requirements have
not yet been fully determined for the type of fast SAR system that is specified herein.
Because this class of fast SAR testing methods uses specific hardware and software that do
not conform to those used for full SAR testing required by this document, the information in
Clause 8 might not be sufficient to determine the measurement uncertainty. Additional
uncertainty components that are specific to the measurement system and the method under
consideration can be needed. Clause C.2 gives general guidance for determining the
measurement uncertainty of results obtained using this class of fast SAR testing methods.
C.2 Measurement uncertainty evaluation – contributing parameters
C.2.1 General
General guidance is provided to determine the relative measurement uncertainty of a Class 2

fast SAR testing method. For Class 1 fast SAR testing systems, the uncertainty analysis is the
same as that specified in Clause 8, except that additional sampling error uncertainty is added
as specified in C.2.7.
Uncertainty assessment methods for SAR results obtained using vector measurement-based
systems are given in IEC 62209-3. The uncertainty assessment guidance in IEC 62209-3 shall
be applied for the purposes of this document.
– 136 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
For a specific fast SAR testing method, rigorous uncertainty analysis should be applied for all
applicable uncertainty terms, including any terms that are not specified in Clause C.2. The
relative uncertainty is required, so uncertainty contributions that represent fixed offsets for a
set of test configurations are not considered in the uncertainty budget. This includes fixed
offsets due to system calibration, dielectric properties that do not change significantly during
the tests (tissue-equivalent medium temperature variations between tested configurations
fulfil the requirements in 6.1), uncertainties in phantom construction that are fixed for a given
system, and other offsets. Uncertainty terms that result in variations in the measured values
from their nominal values for a given set of measurements should be considered in the
relative uncertainty budget. This includes noise in the electronics, variations due to device
positioning and measurement errors, and other items.
For Class 2 fast SAR testing, two uncertainty budgets are presented in Annex C.
a) Table C.1 (see Clause C.3) is the uncertainty budget to be used when fast SAR testing is
applied for several test configurations within the same frequency band and modulation
(e.g. GSM in the 824 MHz to 849 MHz frequency band). Fixed offsets that apply when
relative assessments are conducted are not included in the uncertainty. Also, fixed offsets
that are specific to frequency and operating mode (e.g. offsets due to calibration
uncertainty and modulation) are not included.
b) Table C.2 (see Clause C.3) is the uncertainty budget for the system check for Class 2 fast
SAR testing. This uncertainty budget is very similar to the uncertainty budget of Clause 8.
All uncertainty contributions should be computed in a rigorous way, using appropriate

statistical techniques. A description of the procedures used in the evaluations of uncertainty
contributions, if different from those reported in this document, should be included in the
measurement report.
C.2.2 Probe calibration and system calibration drift
If the fast SAR testing method uses one probe, then the same method specified in 8.4.1.1
should be used, where appropriate. If the Class 2 fast SAR testing method uses several
probes, then the calibration should be applied over the number of sensors used by the
system, where appropriate. The system calibration uncertainty should be documented and
appropriate to the calibration method applied.
The probe calibration drift over time is an estimate for the change in calibration values of the
system over the calibration interval within a specific calibration laboratory, and is based on
the calibration history of the system. It is determined by Type A evaluation (8.4.1.2). For
example, the calibration laboratory may have a specific system that it uses for regular re-
calibration. In this case, the calibration drift should be specified as the standard deviation of
the re-calibrations, relative to the mean value. It is recommended that the system calibration
measurements that are used to determine this uncertainty are made over a period that is
longer than the calibration period of the other equipment (e.g. power meters) used during
calibration. This ensures that the calibration drift uncertainty includes variations due to the re-
calibration of this equipment. Where a calibration history is not available, the manufacturer’s
estimate for the stability of the system over the calibration interval should be used. If the
calibration drift uncertainty is not available, the full calibration uncertainty should be used
instead.
C.2.3 Isotropy
Probe isotropy uncertainty specified in 8.4.1.5 is applicable to probes that can be positioned
in a range of angles with respect to the phantom shell according to the specifications of
Table 3 and Table 4 (see 7.4.2).
IEC/IEEE 62209-1528:2020 – 137 –
© IEC/IEEE 2020
If the probes of the fast SAR testing system are positioned in a similar way as for Class 1 fast
SAR testing, then the probe isotropy uncertainty procedure specified in 8.4.1.5 should be
used, where appropriate. If the probe angles with respect to the phantom shell do not comply
with the probe angles specified in Table 3 and Table 4, then additional uncertainty evaluation
can be needed. If the probe angles are limited to a narrower range, the isotropy uncertainty
can be performed over that range. The isotropy assessment should be applicable to the
measurement system used. If algorithms are applied to compensate for the deviation from
isotropic response, then the SAR uncertainty shall be determined with the same evaluation
hardware and software as is used for performing the SAR measurements.
For Class 2 fast SAR testing, technology-specific considerations shall be taken into account.
The manufacturer shall evaluate this contribution during the calibration. If algorithms are
applied to compensate for the deviation from isotropic response, then the SAR uncertainty
shall be determined with the same evaluation hardware and software as is used for
performing the SAR measurements. To evaluate the isotropy, the SAR measurement system
shall be exposed to reference fields containing waves with varying angles of incidences with
respect to the line normal to the phantom surface. The SAR gradients in the direction normal
to the phantom shell shall be low enough that probe positioning tolerance does not
significantly influence the isotropy assessment. The isotropy shall be characterized for at
least one frequency in each frequency band of interest to the communication system of the
DUT. The assessed tolerance valid at the sensor location needs to be translated to the
uncertainty for the psSAR and filled into Table C.1. The translation is a function of the actual
implemented 3D reconstruction and can only be performed by the manufacturer. All steps
shall be documented.
C.2.4 Probe positioning
There is some uncertainty of the exact location of the sensor(s) used by the fast SAR testing
method. If the fast SAR testing system uses one or more moveable sensors, then the
uncertainty is due to uncertainties in the positioning of the sensors. If the fast SAR testing
system uses one or more fixed sensors, the uncertainty is due to mechanical uncertainties in
the sensor location during manufacturing.
The uncertainty is evaluated using the set of reference distributions in O.8.2. The procedure
is as follows.
a) For each reference distribution, the field is sampled at points and in polarizations which
correspond to the measurement grid applied by the physical system.
b) From these original sampled data, the psSAR is computed using the same processing as
used for measurements with this particular system. In this process, step a) gives the
reference SAR, SAR ref,i , for each field distribution case i (several i distributions are
assumed per frequency).
c) The known tolerance on probe positioning is used to generate measurement scan grids G j
(j = 1 to J) affected by mechanical errors. The manufacturer shall make a model to
represent such mechanical errors. For example, each tolerance may have a rectangular or
triangular distribution if the manufacturer defines maximum acceptable deviations from
nominal locations. In the case of moving probes, the repeatability uncertainty of the probe
positioning shall be included in the mechanical error. J sampled distributions of each
reference vector field are then created by using the G j sampling grids.
d) For each reference distribution i and each grid j, the reconstruction algorithms of the
measurement system are applied to compute the psSAR, SAR i,j . The software shall
assume ideal positions of the points in the grid. Consequently, the algorithms are used as
if sampled field values were obtained from a probe or probe-array with perfectly known
position, whereas the probe or probe-array is actually sampling the field on a grid which
deviates from the targeted locations.
e) For each (i, j) pair, the relative difference ∆SAR i,j = (SAR i,j − SAR ref,i ) / SAR ref,i is computed.
– 138 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
U i is then calculated for each reference field configuration i as the 95th percentile of the
absolute value of ∆SAR i,j for all J probe or probe-array grid locations. The uncertainty is then
estimated as the RMS value of U i across the applicable frequency range for all I exposure
conditions. This value is calculated on I × J cases with I × J − 1 degrees of freedom. A normal
distribution (k = 2) is assumed.
The assessed tolerance valid at the sensor location needs to be translated to the uncertainty
for the psSAR and filled into Table C.1. The translation is a function of the actual
C.2.5 Mutual sensor coupling
For systems using more than one probe, coupling between the probe sensors might exist, due
to imperfect isolation of the sensors. This uncertainty term is caused by port-to-port coupling,
where the readout at a given sensor port is influenced by the voltage and/or current at
another sensor port. Mutual sensor coupling may be evaluated and compensated during
calibration. However, this coupling is in principle dependent on the polarization, distribution,
and angle of incidence of the impinging field, and is therefore difficult to remove completely
during calibration.
The assessed tolerance valid at the sensor location needs to be translated to the uncertainty
for the psSAR and filled into Table C.1. The translation is a function of the actual
In order to isolate this uncertainty contribution from other components:
a) the probe or probe-array shall be immersed in tissue-equivalent medium complying with

the requirements of this document;
b) the reference antennas used shall have a polarization purity of more than 10 dB.
Polarization purity is the ratio of the field strength of the dominant field component to the
field strength of the two orthogonal field components. It is specified at the location of the
psSAR. It is assessed from numerical simulations of the reference antenna;
c) the reference antennas shall be far enough from the phantom so that the interaction with
the probe or probe-array can be neglected;
d) the field levels shall be sufficiently above the noise floor.
In the case of probe arrays with probes that are calibrated independently prior to assembly, or
in the case of probe arrays that are calibrated together in the final operating condition without
compensation for mutual coupling, the mutual sensor coupling uncertainty shall be evaluated
by computing the difference between the single point SAR measured using one probe in
isolation and the SAR measured by the same probe surrounded by the other probes. Several
source antennas with different polarizations and angles of incidence are required. The
antennas are moved to the required positions so that each sensor is actually exposed to a
field strength compatible with the dynamic range of the measurement system. For each test
condition, the calibration coefficients and post-processing are applied on the measured data
to evaluate the vector field distribution over the measurement surface. It shall be documented
how the error propagates and how the uncertainty is coupled to other uncertainty terms.
For probe-arrays where the probes are calibrated together in the final operating condition and
the mutual coupling between the sensors is characterized as a scattering matrix then
compensated (e.g. during the probe-array calibration), mutual coupling will be reduced.
However, the remaining mutual sensor coupling error shall be assessed. Also, the uncertainty
of the compensation is taken into account as an additional contribution to the measurement
system calibration uncertainty. It shall be documented how the error propagates and how the
uncertainty is coupled to other uncertainty terms.
IEC/IEEE 62209-1528:2020 – 139 –
© IEC/IEEE 2020
C.2.6 Scattering within the probe array
Scattering of the electromagnetic fields in the phantom due to the presence of the probe array
causes distortion of the impinging field distribution captured by the sensors. These reflections
are strongly dependent on the distribution, polarization, and angle of incidence of the incident
fields; therefore, they cannot be fully calibrated out. This uncertainty term is correlated with
the uncertainty due to mutual coupling between the sensors.
It is also possible to assess together the isotropy (C.2.3), mutual sensor coupling (C.2.5), and
scattering within the array (C.2.6) uncertainty components.
C.2.7 Sampling error
The measured SAR distribution is estimated from the direct sampling of the field at a set of
discrete points. If the sampling resolution is coarse compared to the gradients of the induced
SAR distribution, the psSAR can be inaccurate. The uncertainty is evaluated using the set of
reference distributions in O.8.2.
The evaluation of this uncertainty may be combined with the evaluation of the uncertainty of
reconstruction algorithms.
C.2.8 Array boundaries
In addition to SAR errors due to the spatial resolution and the locations of the sensors, there
are errors at the boundary of the measurement area due to truncation of the measured field
distribution and extrapolation of the measured field distribution beyond the measurement
area.
This uncertainty may be evaluated together with the uncertainty of reconstruction algorithms.
The uncertainty is evaluated using the reference distributions of O.8.2. The reference
distributions are applied close to the boundary of the measurement area. If the measurement
system provides warnings or error messages to the user about the measurement error at
locations near the boundary, the evaluation shall be performed inside of these locations.
Compare the resulting 1 g and 10 g psSAR with the target values.
C.2.9 Probe or probe array coupling with the DUT
In addition to the coupling specified in C.2.5, there can be coupling between the sensors and
the DUT that also contribute to measurement uncertainty. To assess the uncertainty, several
radiating devices and test positions should be assessed, as appropriate for the measurement
system, using both fast SAR testing and full SAR testing.
C.2.10 Measurement system immunity / secondary reception
As with full SAR testing systems, fast SAR testing systems should be designed in such a way
as to mitigate any noise or offsets in the measurements due to signal pickup that is not
measured at the sensor(s). Incident electric fields on high impedance transmission lines can
be picked up by the lines, causing interference (secondary reception) with the signal picked
up by the sensor(s). For full SAR testing, this effect is minimized by using high impedance
lines along the probe axis. For fast SAR testing systems using different sensor arrangements,
the immunity to these secondary signals should be measured. This should be accomplished
by performing a single point measurement of one probe of the fast SAR testing system using
the system validation set-up specified in Annex A, then comparing the measured value
against the value using a probe used for full SAR testing. To ensure that potential coupling
between the sensors and the DUT (C.2.5 and C.2.9) is not included, the radiating source
should be kept away from the fast SAR testing system by a minimum of one quarter of a
wavelength in air.
– 140 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
NOTE It is sufficient to test this uncertainty term using the densest probe spacing of the probe array if the probe
array has variable spacing.
C.2.11 Deviations in phantom shape
This uncertainty term takes into account measurement variations due to the phantom shape, if
the phantom shape deviates from the standardized phantom specified in Annex G. This term
should be determined from the difference between the SAR in the phantom shape used by the
fast SAR testing method and the SAR in the standardized phantom. It is recommended to
perform the measurements in both phantoms using full SAR testing, so as not to include the
sensor coupling with the DUT (C.2.9) in the uncertainty analysis. Different sources and source
positions with respect to the phantom should be applied, as appropriate, to evaluate the
influence of the phantom shape on different SAR distributions.
C.2.12 Spatial variation in dielectric properties
The dielectric properties of the tissue-equivalent medium can vary spatially. The spatial
variation in the dielectric properties can be assessed on a production sample of the material
by making dielectric properties measurements at multiple regions (at least 10, including
different locations in x-, y- and z-directions in the volume surrounding the sensor locations) in
a dielectric sample across the frequency range of interest. If the tissue-equivalent medium is
inaccessible to the user, this measurement should be performed by the system manufacturer.
The methods of Annex H can be used to measure the dielectric properties, if appropriate. The
dielectric properties measurement methods should be appropriate to the material being
tested. The uncertainty term is calculated from the highest percentage difference in each of
the permittivity and conductivity from the average value, and a rectangular distribution is
assumed.
C.2.13 Reconstruction
The method to evaluate the uncertainty contribution due to reconstruction algorithms,

excluding psSAR averaging, is as follows.
a) Sample reference analytical field distributions at grid points and using polarizations
corresponding to the measurement scan performed by the actual system.
b) Process the sampled data by utilizing system reconstruction and peak spatial-averaging
algorithms.
c) For each distribution, compute the relative difference between the psSAR obtained in
step b) and the reference SAR targets for the provided analytical functions.
NOTE Analytical field distributions are described in IEC 62209-3.
C.3 Uncertainty budget
The measurement uncertainty budget specified in Table C.1 is the uncertainty of the
measured SAR value for one test configuration compared to the measured SAR of the other
test configurations that have been measured using the same fast SAR testing system at the
same frequency band and modulation.
The relative uncertainty budget for the system check performed for a fast SAR testing method
is shown in Table C.2. This uncertainty budget is similar to the repeatability uncertainty
budget for system check. The uncertainty terms that can be neglected have weighting
coefficients of c i = 0.
IEC/IEEE 62209-1528:2020 – 141 –
© IEC/IEEE 2020
Table C.1 – Measurement uncertainty budget for relative

SAR measurements using Class 2 fast SAR testing,
for tests performed within one frequency band and modulation
Prob.
Input quantity X i Unc. Div. u(x i ) = c i (1 g, u(y) = ν i or
Ref. Dist.
(source of uncertainty) a(x i ) qi a(x i )/q i 10 g) c i ∙u(x i ) ν eff
PDF i
Measurement system
System calibration C.2.2 N 1 0 ∞
Isotropy C.2.3 R √3 1 ∞
Boundary effect 8.4.1.6 R √3 1 ∞
Linearity 8.4.1.3 R √3 1 ∞
System detection limits 8.4.1.3 R √3 1 ∞
Modulation response 8.4.2.7 R √3 1 ∞
Readout electronics 8.4.1.6 N 1 0 ∞
Response time 8.4.1.6 R √3 0 ∞
Integration time 8.4.1.6 R √3 1 ∞
RF ambient conditions – noise 8.4.1.7 R √3 1 ∞
RF ambient conditions – reflections 8.4.1.7 R √3 0 ∞
Probe positioning C.2.4 R √3 1 ∞
Mutual sensor coupling C.2.5 R √3 1 ∞
Scattering within the array C.2.6 R √3 1 ∞
Sampling error C.2.7 R √3 1 ∞
Array boundaries C.2.8 R √3 1 ∞
Probe or probe array coupling with

C.2.9 R √3 1 ∞
DUT
Measurement system immunity C.2.10 R √3 1 ∞
Post-processing of measurement
8.4.1.9 R √3 1 N−1
data
Spatial variation in dielectric

C.2.12 R √3 1 ∞
properties
Reconstruction C.2.13 R √3 1 ∞
Test sample related
Test sample positioning 8.4.2.5 N 1 1 ∞
Device holder uncertainty 8.4.2.6 N 1 1 ∞
Output power variation – SAR drift

8.4.2.9 R √3 1 ∞
measurement
SAR scaling 8.4.3.2 R √3 0 ∞
Phantom and tissue parameters
Deviations in phantom shape C.2.11 R √3 1 ∞
Uncertainty in SAR correction for

deviations in permittivity and 8.4.3.1 N 1 0 ∞
conductivity
Medium conductivity measurement 8.4.2.1 N 1 0 M
Medium permittivity measurement 8.4.2.1 N 1 0 M
Medium conductivity – temperature

8.4.2.2 R √3 0 ∞
uncertainty
Medium permittivity – temperature

8.4.2.2 R √3 0 ∞
uncertainty
Combined standard uncertainty,

RSS ν eff
u(ΔSAR)
Expanded uncertainty, U k=2

– 142 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table C.2 – Measurement uncertainty budget for
system check using Class 2 fast SAR testing
Prob.
Input quantity X i Unc. Div. u(x i ) = c i (1 g; u(y) = ν i or
Ref. Dist.
(source of uncertainty) a(x i ) qi a(x i )/q i 10 g) c i ∙u(x i ) ν eff
PDF i
Measurement system
System calibration drift C.2.2 N 1 1 ∞
Isotropy C.2.3 R √3 0 ∞
Boundary effect 8.4.1.6 R √3 0 ∞
Linearity 8.4.1.3 R √3 0 ∞
System detection limits 8.4.1.3 R √3 0 ∞
Modulation response 8.4.2.7 R √3 0 ∞
Readout electronics 8.4.1.6 N 1 0 ∞
Response time 8.4.1.6 R √3 0 ∞
Integration time 8.4.1.6 R √3 0 ∞
RF ambient conditions – noise 8.4.1.7 R √3 0 ∞
RF ambient conditions – reflections 8.4.1.7 R √3 0 ∞
Probe positioning C.2.4 R √3 1 ∞
Mutual sensor coupling C.2.5 R √3 1 ∞
Scattering within the array C.2.9 R √3 0 ∞
Sampling error C.2.7 R √3 1 ∞
Array boundaries C.2.8 R √3 0 ∞
Probe or probe array coupling with C.2.6 R √3 ∞

0
DUT
Measurement system immunity C.2.10 R √3 1 ∞
Post-processing of measurement
8.4.1.9 R √3 0 N−1
data
Spatial variation in dielectric ∞

C.2.12 R √3 0
properties
Reconstruction C.2.13 R √3 1 ∞
Test sample related
Test sample positioning 8.4.2.5 N 1 0 ∞
Device holder uncertainty 8.4.2.6 N 1 0 ∞
Output power variation – SAR drift 8.4.2.9 R √3 ∞

0
measurement
SAR scaling 8.4.3.2 R √3 0 ∞
Phantom and tissue parameters
Deviations in phantom shape C.2.11 R √3 1 ∞
Uncertainty in SAR correction for 8.4.3.1 N 1 1;0,84 ∞

deviations in permittivity and
conductivity
Medium conductivity measurement 8.4.2.1 N 1 0,78;0,71 M
Medium permittivity measurement 8.4.2.1 N 1 0,23;0,26 M
Medium conductivity – temperature 8.4.2.2 R √3 0,78;0,71 ∞

uncertainty
Medium permittivity – temperature 8.4.2.2 R √3 0,23;0,26 ∞

uncertainty
Combined standard uncertainty RSS ν eff
Expanded uncertainty, U k=2

IEC/IEEE 62209-1528:2020 – 143 –
© IEC/IEEE 2020
Annex D
(normative)
SAR system validation antennas
D.1 General antenna requirements
Annex D describes antennas that are used to validate SAR measurement systems that
conform to the normative requirements of this document. Antennas other than those specified
in the subclauses of Annex D may be considered for system validation. The following are
minimum requirements for any system validation antenna.
a) Justification shall be provided for the inclusion of an antenna and how its use strengthens
the validation protocol.
b) The test configurations (e.g. frequencies, locations on the phantom, distance, rotation
angle) shall be provided.
c) Detailed drawings, including all dimensions and tolerances, shall be provided. CAD files of
the antennas shall be made available with enough detail of the construction and
parameters so that the numerical targets can be independently verified.
d) Numerical target SAR values shall be determined from numerical simulations of the CAD
file of the antenna. The numerical simulations shall conform to the requirements of
IEC/IEEE 62704-1 [41].
e) Numerical target SAR values shall be available for all specified test configurations on the
phantoms.
f) Physical antennas shall be measured using a SAR measurement system that has an
accredited calibration (e.g. ISO/IEC 17025 [17]) to validate the numeric target values and
generate a certificate of traceability.
NOTE Regulatory acceptance of alternative antenna designs might involve additional requirements other than
those described above.
Various antennas that meet the preceding requirements are specified in Clauses D.2 to D.5.
D.2 Standard dipole antenna
D.2.1 Mechanical description
The standard dipole antennas of Figure D.1 with mechanical dimensions given in Table D.1
will produce the SAR values given in Table D.2 when the system validation test of A.3.5 is
followed. If dipole antennas are used that have different dimensions than those given in
Table D.1, or if dipole antennas are used at frequencies other than those listed in Table D.1,
the reference SAR values for those antennas shall be documented and independently verified
using procedures that are consistent with the methodologies specified in this document.
The reference dipole shall be positioned below the flat phantom and centred with its axis
parallel to the longest side of the phantom. A low loss (loss tangent < 0,05) and low
permittivity (relative permittivity < 5) spacer should be used to establish the correct distance
between the top surface of the reference dipole and the bottom surface of the phantom. Below
3 GHz, the spacer shall not change the measured 1 g and 10 g psSAR values by more than
1 %, compared to the no-spacer condition. Above 3 GHz, the spacer could affect the
measured SAR and introduce deviations from the numerical target values in Table D.2;
therefore, experimentally validated SAR targets with specific spacers shall be used
(see Table D.2). The reference dipole shall have a return loss better than 20 dB at the test
frequency to reduce power reflection and SAR measurement uncertainty.
– 144 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
For the reference dipoles, a dielectric spacer is used to provide the distance s and spacer
mechanical tolerance as given by the following:
s = 15 mm ± 0,2 mm, for 300 MHz ≤ f ≤ 1 000 MHz

s = 10 mm ± 0,2 mm, for 1 000 MHz < f ≤ 6 000 MHz
s = 5 mm ± 0,1 mm, for 6 000 MHz < f ≤ 10 000 MHz
The computation of the reference values in Table D.2 for frequencies above 5 000 MHz
requires specific consideration of the build structure of the small dipoles (both internal and
external) and numerical values are therefore specific to the dipole antenna design. The
dielectric spacer used for the dipole shall also be modelled as it can affect the numerically
determined SAR value.
The reference dipole shall have a return loss better than 20 dB at the test frequency to reduce
the uncertainty in the power measurement. To meet this requirement, it is acceptable to fine-
tune the reference dipoles by using low-loss dielectric or metal tuning elements at the ends of
the dipole (Figure A.1).
The reference dipole arms shall be parallel to the flat surface of the phantom within a
tolerance of ±2° or less (Figure D.1). This can be assured by carefully positioning the empty
phantom and the reference dipole to horizontal level using a spirit level.
The numerical target values above 3 GHz cannot be universally given as for below 3 GHz,
due to the greater effect from the spacer, phantom bottom, and mechanical uncertainties.
Thus, the numerical target values may be different from one dipole to another. It is important
that for each dipole used for system validation a fully documented analysis is provided based
on both numerical simulations and experimental validation.
IEC/IEEE 62209-1528:2020 – 145 –
© IEC/IEEE 2020
Key
L length of the dipole

d1 diameter of the dipole
d2 diameter of the coaxial cable at the balun
h length of the balun choke section
PTFE polytetrafluoroethylene
Figure D.1 – Mechanical details of the standard dipoles

– 146 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table D.1 – Mechanical dimensions of the reference dipoles
Frequency L h d1
MHz mm mm mm
300 420,0 250,0 6,35
450 290,0 166,7 6,35
750 176,0 100,0 6,35
835 161,0 89,8 3,6
900 149,0 83,3 3,6
1 450 89,1 51,7 3,6
1 500 86,2 50,0 3,6
1 640 79,0 45,7 3,6
1 750 75,2 42,9 3,6
1 800 72,0 41,7 3,6
1 900 68,0 39,5 3,6
1 950 66,3 38,5 3,6
2 000 64,5 37,5 3,6
2 100 61,0 35,7 3,6
2 300 55,5 32,6 3,6
2 450 51,5 30,4 3,6
2 600 48,5 28,8 3,6
3 000 41,5 25,0 3,6
3 500 37,0 26,4 3,6
3 700 34,7 26,4 3,6
a a
5 000 to 6 000 20,6 40,3 3,6
6 000 to 8 000 14,5 35,1 2,2
8 000 to 10 000 12,2 25,0 2,2
NOTE The L, h, and d 1 dimensions shall be within

±2 % tolerance.
a These dimensions are applicable for a coaxial
diameter at the balun of d 2 = 2,1 mm
(Figure D.1).
D.2.2 Numerical target SAR values
Table D.2 shows the validated numerical target SAR values for system validation (A.3.5),
using the standard dipoles specified in D.2.1. The numerical target SAR values in Table D.2
were calculated using the finite-difference time-domain (FDTD) numerical-computation
method using the flat phantom requirements in Clause G.3 and Table G.3, and also validated
against measured results with equivalent test set-ups for 300 MHz to 10 GHz. The local SAR
values in columns 5 and 6 of Table D.2 were experimentally verified for each test frequency
using fourth-order polynomial extrapolation. The values above 3 GHz are dependent on the
dipole spacer used and detailed construction of the dipoles; therefore, individual dipole
targets may vary from the values in Table D.2 by as much as ±10 %. The reason is that the
dipole length is short with respect to its diameter and the spacer dimensions, i.e. the
numerical target values are not generic and need to be determined for a particular
configuration. The dielectric properties used for the tissue-equivalent medium are specified in
Table 2 (see 6.2.2), and the dimensions of the reference dipoles are shown in Table D.1.
Different reference SAR value may apply for dipoles with mechanical dimensions that depart
from those of the reference dipoles given in D.2.1.
IEC/IEEE 62209-1528:2020 – 147 –
© IEC/IEEE 2020
Table D.2 – Numerical target SAR values (W/kg) for standard dipole and flat phantom
Local SAR at
Local SAR at
Phantom shell surface (y =2 cm
Frequency 1 g SAR 10 g SAR surface (above
thickness offset from
feed-point)
feed-point)
MHz mm W/kg W/kg W/kg W/kg
300 2,0 2,85 1,94 4,14 2,00
450 2,0 4,58 3,06 6,75 2,98
750 2,0 8,49 5,55 12,6 4,59
835 2,0 9,56 6,22 14,1 4,90
900 2,0 10,9 6,99 16,4 5,40
1 450 2,0 29,0 16,0 50,2 6,50
1 500 2,0 30,5 16,8 52,8 6,53
1 640 2,0 34,2 18,4 60,4 6,69
1 750 2,0 36,4 19,3 64,9 6,53
1 800 2,0 38,4 20,1 69,5 6,80
1 900 2,0 39,7 20,5 72,1 6,60
1 950 2,0 40,5 20,9 72,7 6,60
2 000 2,0 41,1 21,1 74,6 6,50
2 100 2,0 43,6 21,9 79,9 6,58
2 300 2,0 48,7 23,3 92,8 7,18
2 450 2,0 52,4 24,0 104 7,70
2 585 2,0 55,9 24,4 119 7,9
2 600 2,0 55,3 24,6 113 8.29
3 000 2,0 63,8 25,7 140 9,50
3 500 2,0 67,1 25,0 169 12,1
3 700 2,0 67,4 24,2 178 12,7
5 000 2,0 77,9 22,1 305 15,1
5 200 2,0 76,5 21,6 310 15,9
5 500 2,0 83,3 23,4 349 18,1
5 800 2,0 78,0 21,9 341 20,3
7 000 2,0 275,0 47,0 2 161,0 13,0
9 000 2,0 243,0 40,0 2 676,0 12,0
NOTE 1 All SAR values are normalized to 1 W forward power.
NOTE 2 The 1 g and 10 g target SAR values are only valid for the system validation specified in
Clause A.1, using dipoles with the dimensions specified in D.2.1.
NOTE 3 The SAR values above 3 GHz are sensitive to the dipole and phantom geometry (dipole
length, spacer length and permittivity, and phantom shell permittivity) and can vary by as much as
±10 %. See Onishi and Uebayashi [36] for details.
– 148 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
D.3 Standard waveguide
The standard waveguide of Figure D.2 with mechanical dimensions given in Table D.3
(corresponding to WR159 or UK WG-13 with an IEC-UDR58 flange according to IEC 60154-2
[42]) will produce the SAR values given in Table D.4 when the system validation test of A.3.5
is followed. The waveguide feed shall be placed at least one guide wavelength away from the
matching layer, to ensure that higher order modes have attenuated. The waveguide
transmission loss shall be characterized by measurement with a network analyser. If
waveguides are used that have different parameters than those given in Table D.3, or if
waveguides are used at frequencies other than those listed in Table D.3, the reference SAR
values for those waveguides shall be documented and independently verified (e.g. by
comparison of numerical simulations with measurements).
Figure D.2 – Standard waveguide (dimensions are according to Table D.3)
Table D.3 – Mechanical dimensions of the standard waveguide
Frequency Phantom shell L W Lf Wf t εr

thickness
MHz mm mm mm mm mm mm
5 200 2 40,39 20,19 81,03 61,98 5,3 6

5 800 2 40,39 20,19 81,03 61,98 4,3 6
NOTE L and W are the inner length and width of the waveguide, L f and W f are the
length and width of the waveguide flange, and t and ε r are the thickness and relative
permittivity of the matching layer. The tolerance for L and W is ±0,13 mm. The
matching layer is a lossless dielectric slab that fills the cross-sectional L · W area of
the waveguide. The waveguide and matching layer are in direct contact with the
phantom shell. The minimum height of the waveguide (from the feed to the flange) is
one free-space wavelength. This arrangement gives an input return loss of at least
8 dB. The uncertainty of the permittivity and thickness of the dielectric slab are
included in the return loss. Therefore, these do not need to be specified
independently.
IEC/IEEE 62209-1528:2020 – 149 –
© IEC/IEEE 2020
Table D.4 shows the numerically validated SAR values for system validation using the
standard waveguides specified in D.3.1. The numerically validated target SAR values are
dependent on the relative permittivity of the phantom shell; therefore, target values are given
for phantom shell relative permittivity values of 3, 4, and 5. Linear interpolation shall be
applied for phantom shell relative permittivity values between these values. The numerical
target SAR values of Table D.4 are for 1 W measured net power into the waveguide, and were
calculated using the finite-difference time-domain (FDTD) method with volume averaging
requirements specified in IEEE Std C95.3-2002 [43] and verified by measurements [44]. The
power delivered to the phantom is equal to the forward power into the waveguide minus the
reflected power and waveguide transmission losses (including losses in the adapter and walls
of the waveguide). The transmission loss of the adapter can be determined by measuring S 11
at its coaxial port with three different known waveguide terminations connected to the
waveguide port, for example a short plus two different offset shorts (short-circuited
transmission lines at different offset lengths). The waveguide transmission loss shall be
measured at least annually using a calibrated network analyser.
The waveguide used in the simulations was modelled as a perfect electric conductor with a
matching dielectric window having dimensions as specified in D.3.1. The phantom used in the
simulations has a length of 216 mm, a width of 152 mm, a depth of 80 mm and a shell
thickness of 2 mm. The dielectric properties of the tissue-equivalent medium are as specified
in Table 2 (see 6.2.2).
Table D.4 – Numerical target SAR values for waveguides
Point SAR as a function

Phantom shell
of distance, d [mm] into
Frequency relative 1 g SAR 10 g SAR
the phantom along its
permittivity
centreline
MHz W/kg W/kg
3 165 53,7 667 exp(−2d/6,2)
5 200 4 180 56,5 733 exp(−2d/6,2)
5 194 59,1 796 exp(−2d/6,2)
3 165 49,3 804 exp(−2d/5,5)
5 800 4 184 52,5 907 exp(−2d/5,5)
5 200 55,2 982 exp(−2d/5,5)
In case the net power produces measured SAR values that are above the dynamic range of the probe,
lower powers shall be used so as not to introduce additional measurement uncertainty or damage the
probe.
NOTE 1 The waveguide flange is placed in direct contact with the outer shell surface of the phantom.
NOTE 2 All SAR values are normalized to 1 W net power (i.e. the power delivered to the phantom).
NOTE 3 The 1 g and 10 g reference SAR values are only valid for the system validation specified in
Clause A.1, using waveguides having dimensions and construction as specified in D.3.1.
D.4 System validation antennas for below 150 MHz
D.4.1 General
At frequencies of 150 MHz and below, the physical dimensions of standard dipoles and
waveguides become too large to be practical. Therefore other antennas may be used that
have been calibrated for 1 g and 10 g SAR.
– 150 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
D.4.2 Confined loop antenna
D.4.2.1 Mechanical description
An example for a confined loop antenna (CLA) is a resonant loop as shown in Figure D.3. The
requirements for such antennas are:
a) resonant at the target frequency: f target ± 1 MHz;

b) return loss at resonant frequency: > 10 dB;
c) stability of 1g and 10 g psSAR (at maximal specified input power): < 2 % (> 60 minutes);
d) interlaboratory reproducibility of 1 g and 10 g psSAR: < 10 %;
e) distortion (by changing the reflectivity of the floor or rotation of the structure below the
phantom for typical validation configurations): < 3 %.
The return loss requirement is 10 dB compared to the 20 dB requirement for dipole antennas
because the devices are considerably smaller than the wavelength. In general, antennas
other than the described dipoles and waveguides have higher losses, and their efficiency can
change at high input power due to thermal effects. Furthermore, high quality-factor structures
are sensitive to small changes by the reflective environment. Therefore, these antennas shall
be designed to minimize the influence of the environment.
Key
C capacitance
NOTE Dimensions: a1 = 220 mm, a2 = 180 mm, a3 = 25 mm, a4 = 110 mm, a5 = 45 mm.
Figure D.3 – Drawing of the CLA that corresponds to a resonant loop integrated
in a metallic structure to isolate the resonant structure from the environment
IEC/IEEE 62209-1528:2020 – 151 –
© IEC/IEEE 2020
D.4.2.2 Numerical target SAR values
The target values for the CLA have been assessed by FDTD modelling of the antenna under
the elliptical phantom (600 mm × 400 mm) with a 2 mm thick lossless dielectric shell (ε r = 3,7).
The antenna is fed by a stub of a coaxial cable that is terminated by a voltage source with an
inner resistance of 50 Ω. The maximum mesh step is limited to 3 mm in the phantom and to
15 mm in free space. For the lower frequencies and larger meshes, the maximum mesh step
is relaxed; it is always kept below λ/10. The absorbing boundary conditions terminate the
computational domain at about 200 mm from the antenna. At frequencies of 64 MHz and
below, this distance is increased to approximately λ/4. In all simulations, the antennas are
excited with a broadband pulse. Due to their high quality factor (Q), the simulations are run for
50 periods of the operational frequencies. In particular, at the lower end of the investigated
frequency range, the input impedance becomes highly reactive for small deviations from the
resonance frequency, which prejudices the evaluation of the power budget of the simulations.
Because the observed deviations from the nominal resonance frequency are small (< 5 %),
the SAR evaluations are carried out at the actual resonance of the numerical model. To obtain
reliable target values, the psSAR values are normalized to the peak of the H-fields in the
plane at 5 mm distance from the medium bottom inside the phantom. Table D.5 lists the 1 g
and 10 g psSAR results normalized to the H-field. The H-field normalization provides target
SAR values that are independent of the internal antenna losses; the latter can be large and
vary between implementations. For the same CLA type, a calibration laboratory can convert
the primary H-field value to the secondary input power value that is more convenient to be
used than the primary H-field values and without loss of repeatability, provided the CLA will
not be opened and modified.
The assessment of the numerical uncertainty is carried out at 150 MHz, which is regarded as
representative for most of the configurations. The uncertainty contribution of the power budget
is evaluated for the worst case simulation at 30 MHz. For the evaluation of the grid resolution
uncertainty, the mesh was further refined to a maximum step of 1,5 mm in the phantom and
7 mm in free space. In the bounding box of the antenna, the mesh step maximum was set to
1 mm. The expanded uncertainty of the numerical parameters has been assessed as 0,5 dB
(k = 2).
Table D.5 – Numerical target SAR values for CLAs
Local SAR at Local SAR at

surface surface
f 1 g psSAR 10 g psSAR Uncertainty
(above feed- (20 mm
point) offset)
MHz W/kg/(A/m) 2 W/kg/(A/m) 2 W/kg/(A/m) 2 W/kg/(A/m) 2 dB, k = 2
CLA-6 6 0,000 84 0,000 54 0,001 33 0,000 390 0,5
CLA-13 13 0,004 04 0,002 59 0,006 36 0,001 96 0,5
CLA-30 30 0,022 3 0,014 3 0,035 5 0,010 3 0,5
CLA-64 64 0,106 0,068 1 0,168 0,047 4 0,5
CLA-128 128 0,607 0,417 0,904 0,032 6 0,5
CLA-150 150 0,842 0,586 1,24 0,489 0,5
CLA-220 220 1,20 0,818 1,82 0,686 0,5
NOTE 1 The housing of the CLA is placed in direct contact with the outer shell surface of the flat
phantom.
NOTE 2 All SAR values are normalized to a peak H-field of 1 A/m measured in a plane at a distance of
5 mm from the medium bottom inside the phantom.
– 152 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
D.4.3 Meander dipole antenna
D.4.3.1 Mechanical description
Subclause D.4.3 describes a modified 150 MHz antenna, called the meander dipole, which
can be installed under the standard elliptical phantom [45], [46]. Mechanical dimensions of the
meander dipole are given in Figure D.4 and Table D.6. For the meander dipoles, a dielectric
spacer is used to provide the distance of 15 mm ± 0,2 mm [47].
Table D.6 – Mechanical dimensions of the reference meander dipole
Frequency L1 L2 L3 h d1 d2
MHz mm mm mm mm mm mm
150 197,35 129,35 197,0 165,0 6,35 6,35
Key
L1 partial length of the dipole
h length of the balun choke section
d1 diameter of the dipole arms
d2 diameter of the stub
PTFE polytetrafluoroethylene
Figure D.4 – Mechanical details of the meander dipoles for 150 MHz
D.4.3.2 Numerical target SAR values
The target values for the meander dipole antenna are shown in Table D.7 [45].
IEC/IEEE 62209-1528:2020 – 153 –
© IEC/IEEE 2020
Table D.7 – Numerical target SAR value (W/kg) for meander dipole
Local SAR at
Local SAR at surface
1g 10 g surface Uncertainty
Frequency (y = 2 cm offset
SAR SAR (above feed- (dB, k = 2)
from feed-point)
point)
MHz W/kg W/kg W/kg W/kg
150 1,71 1,23 2,16 1,22 0,2
NOTE 1 All SAR values are normalized to 1 W forward power.
NOTE 2 The meander dipole is placed parallel to the surface of the flat phantom at a
distance of 15 mm using a dielectric spacer.
D.5 Orthogonal E-field source – VPIFA
These single band antennas are designed to be placed at a wide range of locations on the
curved surface of the SAM head or on a flat phantom. The design is chosen to provide a
predominantly E-field coupling, as seen in a range of antenna structures where antenna
metallization extends beyond the ground plane. The structure is that of a PIFA antenna. The
main resonant metallic ribbon is orthogonal to the surface of the phantom. It is short-circuited
on one side, and open-ended on the side that is close to the phantom, where a capacitive
coupling is generated. The distance between the conductor and the phantom, and its position,
are fixed by using a plastic separator, which is part of the antenna. The device should be held
orthogonal to the surface. The SAR pattern is localized, with a dominant orthogonal field
component. Four frequency versions for VPIFA-750, VPIFA-835, VPIFA-1950, and
VPIFA-3700 are specified. At higher frequencies, the normal components dominate for
d ≥ 2 mm. Figure D.5 a) shows the basic structure of the antenna, seen from the PCB1 side.
Figure D.5 b) shows the antenna as seen from the PCB2 side; the flat mask is removed in the
figure to easily see the integrated spacer. The critical dimensions are shown in Figure D.5 c).
Two PCB substrates are used for etching the metallic ribbon and the ground plane,
respectively, and each has a thickness of 0,508 mm. Both substrates are mechanically
separated by an FR4 frame. One end of this frame is used to fix the antenna to a holder. The
antenna is designed to be matched to a 50 Ω coaxial connection. The dimensions for each
frequency and the material properties are provided in Table D.8 and Table D.9, respectively.
In order to compensate for the manufacturing errors, the antenna includes a tuning structure
and a matching screw. For the tuning, a dielectric bar is adjusted to tune the resonant
frequency, and for matching a brass screw integrated in the short circuit can be adjusted to
change the distance between the feed-point and the short-circuit, and therefore, improve the
matching. The screw implementation requires a 0,3 mm prism with a hole in order to
mechanically position and fix the screw. Additionally, a lock nut is used to secure the screw
position. After the antenna is fine-tuned, both elements should be fixed. The SMA connector
requires a bending section to place the connection cable parallel to the antenna axis. To
reduce the twisting force (torque), which can break the connector, it is recommended to first
connect the bend to the cable, then connect the other side of the bend to the connector.
Figure D.6 shows the positioning mask used with the VPIFAs (see also Figure D.5 a)). The
positioning tolerance using these masks is dominated by the mask uncertainty, and the
corresponding SAR uncertainty is already considered in the uncertainty of the target values.
– 154 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table D.8 – Dimensions for VPIFA antennas at different frequencies
Ribbon Frame Matching

Length Frame Tuner
& bent thickness, l f screw
Freq. of bent width, width d b-f d b-s l gr g
width / radiator diameter /
width l b Wf Wt
W thickness, l t length
(MHz) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm)
4,00 ± 0,05
10,00 ± 1,50 ± 4,00 ± 0,05 5,00 ± 10,00 58,00 5,50 ± 59,90 2,50 ±
750 / 12,00 ±
0,05 0,10 /5,35 ± 0,05 0,05 ± 0,10 ± 0,10 0,10 ± 0,10 0,05
0,10
4,00 ± 0,05
10,00 ± 1,50 ± 4,00 ± 0,05 5,00 ± 3,50 ± 51,70 5,55 ± 53,70 2,50 ±
835 / 12,00 ±
0,05 0,10 /5,35 ± 0,05 0,05 0,10 ± 0,10 0,10 ± 0,10 0,05
0,10
4,00 ± 0,05
10,00 ± 1,50 ± 4,00 ± 0,05 5,00 ± 3,00 ± 77,70 4,55 ± 78,45 2,50 ±
1 950 / 12,00 ±
0,05 0,10 /5,35 ± 0,05 0,05 0,10 ± 0,10 0,10 ± 0,10 0,05
0,10
4,00 ± 0,05
3,00 ± 1,50 ± 4,00 ± 0,05 5,00 ± 2,00 ± 36,7 ± 5,77 ± 38,75 2,50 ±
3 700 / 12,00 ±
0,05 0,10 /5,35 ± 0,05 0,05 0,10 0,10 0,10 ± 0,10 0,05
0,10
a) General schematic view from the metallic ribbon side
b) General schematic view from the ground plane side

IEC/IEEE 62209-1528:2020 – 155 –
© IEC/IEEE 2020
c) Critical dimensions from front side d) Critical dimensions from rear side
e) Detailed view of the bent segment
Figure D.5 – VPIFA validation antenna
The recommended tolerances for the 3D printed flat mask are ±0,10 mm.
Figure D.6 – Mask for positioning VPIFAs
Table D.9 – Electric properties for the dielectric layers for VPIFA antennas
Layer 750 MHz 835 MHz 1 950 MHz 3 700 MHz

εr σ [S/m] εr σ [S/m] εr σ [S/m] εr σ [S/m]
Frame (FR4) 4,38 3,49 × 10 −3 4,38 3,84 × 10 −3 4,37 8,45 × 10 −3 4,34 15,7 × 10 −3
Mask (3D print) 2,22 1,74 × 10 −3 2,22 1,83 × 10 −3 2,22 3,09 × 10 −3 2,22 5,06 × 10 −3
PCB_1 (RO4350B) 3,75 463 × 10 −6 3,75 517 × 10 −6 3,75 1,24 × 10 −3 3,74 2,46 × 10 −3
PCB_2 (RO4350B) 3,75 463 × 10 −6 3,75 517 × 10 −6 3,75 1,24 × 10 −3 3,74 2,46 × 10 −3
Spacer (PET) 2,97 7,52 × 10 −3 2,96 8,05 × 10 −3 2,90 14,5 × 10 −3 2,84 22,6 × 10 −3
Tuner (FR4) 4,38 3,49 × 10 −3 4,38 3,84 × 10 −3 4,37 8,45 × 10 −3 4,34 15,7 × 10 −3
– 156 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table D.10 provides the target values for the VPIFA antenna when the specified spacer is
placed directly against the flat phantom at the centre location.
Table D.10 – Numerical target SAR values for VPIFAs on the flat phantom
Local SAR at
Local SAR at
surface
surface
Antenna Frequency SAR 1g SAR 10g (y = 2 cm u c (k = 1)
(above feed-
offset from
point)
feed-point)
[MHz] [W/kg] [W/kg] [W/kg] [W/kg] dB
VPIFA-750 750 12,94 3,86 66,06 0,81 0,3
VPIFA-835 835 13,22 3,82 57,02 0,729 0,3
VPIFA-1950 1 950 8,68 3,58 31,84 1,005 0,3
VPIFA-3700 3 700 9,87 4,14 27,22 0,54 0,3
NOTE All SAR values are normalized to 1 W forward power.

IEC/IEEE 62209-1528:2020 – 157 –
© IEC/IEEE 2020
Annex E
(normative)
Calibration and characterization of dosimetric (SAR) probes
E.1 Introductory remarks
Currently available SAR probes are based on Schottky diode detectors. The measured signal
at the output of each sensor is a voltage proportional to E or E 2 , depending on the magnitude
of the electric field incident on the diode.
Most isotropic probes consist of three small dipole sensors with diode detectors at their centre
gaps. The directivity patterns of such sensors are orthogonal, and the total E-field magnitude
is proportional to the root-sum-square (RSS) of the three orthogonal field components. In the
square-law region of the diode response, the sensor output voltage is proportional to the
mean square of the corresponding field component. Beyond that range the output voltage is
compressed, and therefore requires linearization within the dynamic range. Differences due to
manufacturing tolerances will introduce somewhat different sensitivities for each sensor,
which shall be accounted for during calibration and compensated during measurement.
Probe calibration for SAR testing in tissue-equivalent medium will produce either a SAR
conversion factor or an E-field conversion factor. Because SAR is proportional to tissue-
equivalent medium conductivity, a direct calibration in terms of SAR would be valid only for
media with the exact same conductivity. The E-field sensitivity depends more on the complex
permittivity and is less sensitive to the conductivity. Calibration in terms of E-field rather than
SAR has a broader range of validity, and is therefore preferred for routine SAR testing in
which the tissue-equivalent medium properties vary slightly over time.
Probe calibration is usually done with either one-step or two-step methods. In the two-step
method, the total field is given by Formula (E.1):
3 3
fi (Vi )
E
=
2
∑Ei
=
2
∑ ηi ψi
(E.1)
=i 1 =i 1
Here, E i (i = 1, 2, 3) are the components resulting from the projection of the E-field vector on
the three orthogonal sensors, f i (V i ) is a linearizing function of the rectified sensor signal V i , η i
[in μV/(V/m) 2 ] is the sensitivity of dipole sensor i in air, and ψ i is the ratio of sensor response
in air to the response in the dielectric media (sometimes referred to as the conversion factor).
This two-step method has also been called a three-step method (Meier et al. [48]), where the
third step involves the linearizing function f i (V i ).
The linearization function f i (V i ) is not only a function of the RMS value of V i but also
dependent on the signal envelope. Nadakuduti et al. [20] has shown that CW and periodic
pulse-modulated signals with constant amplitude (e.g. GSM) can be more easily corrected for
linearity than communication protocols employing complex modulations with stochastic signal
envelopes. Acceptable linearization of complex modulations can be obtained over a wide
dynamic range with a linearity uncertainty of less than 0,4 dB.
In the one-step methods specified in E.3.3, the total field is given by Formula (E.2):
3 3
fi (Vi )
E
=
2
∑Ei
=
2
∑ γi
(E.2)
=i 1 =i 1
– 158 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Here the η i factors are included in the total in-tissue sensitivity γ i . Another one-step procedure
based on a standard antenna method is described in Person et al. [49].
Probe calibrations are valid only when the sensors are sufficiently far away (at least one
probe-tip diameter) from any media boundaries. When the probe is close to such boundaries
and compensation is not applied, the sensitivity can change. The boundary effect, as well as
the isotropic response of the probe, shall be assessed using separate tests, as described in
8.4.1.5, 8.4.1.6, Clause E.4, and Clause E.6.
E.2 Linearity
Probe linearity is introduced and explained in 8.4.1.3; the associated uncertainty can be
determined in accordance with the methods set forth in 8.4.1.3 for CW signals, and 8.4.2.7 for
modulated signals. Clause E.2 describes a general methodology for assessing probe linearity
of any signal.
The determination of the linearization function f i (V i ) shall be performed and applied prior to
the assessment of the sensitivity (see Clause E.1). This is best performed using an amplitude
or power scan with a well-defined incident field having the corresponding signal
characteristics. Any of the following set-ups can be used, provided the dynamic range of
interest is assessed in steps of 3 dB or less. A true RMS power meter, or an appropriately
calibrated field probe, should be used as reference. The linearization parameters are
determined for the i-th sensor by assessing and removing the linearization error between the
measured response E inc,meas and the incident field E inc , as shown in Formula (E.3):
Einc,meas
errorlinerarization = 20lg
Einc
(E.3)
An approach is described by Nadakaduti et al. [20] to apply curve fitting to remove the
linearization error for signals having complex signal envelopes.
E.3 Assessment of the sensitivity of the dipole sensors
E.3.1 General
The sensitivity factors can be determined by applying one-step or two-step calibration

procedures. Alternatively, transfer calibration can be applied.
E.3.2 Two-step calibration procedures
E.3.2.1 General
The total field shall be evaluated according to Formula (E.2). The separation of the probe
sensitivity into two factors (η i and γ i ) allows the use of certain standardized free-space probe
calibration procedures, and provides additional validation of the probe performance and
calibration set-up.
E.3.2.2 Sensitivity in air (first step)
The most accurate set-ups used for the generation of well-defined fields to simulate free-
space conditions for use in probe calibration are waveguides. The reasons are as follows:
• waveguide set-ups require moderate power and less space than far-field calibration set-
ups;
• generation of the most accurate fields traceable to power readings is possible;
IEC/IEEE 62209-1528:2020 – 159 –
© IEC/IEEE 2020
• the uncertainty produced by the field disturbance due to the probe insertion is negligible
for small near-field probes when the waveguide dimensions are considerably larger than
probe dimensions;
• the set-ups allow easy access for orienting the probe axis normal or parallel to the field
polarization inside the set-up;
• in addition, cross-validation of the general field strengths is possible by using a set of
waveguides with overlapping frequency ranges.
As long as the resistive feed-line does not load the dipole-diode sensor and the probe is small
compared to the wavelength, the sensitivity in free-space is independent of frequency. This
gives an additional validation of the calibration set-ups, and checks for possible field
perturbations due to the probe. Effects due to probe insertion are typically negligible if high-
quality waveguide couplers and matched sources are used. An additional uncertainty source
in the waveguide set-ups is due to reflections from the terminating load, which can result in a
standing wave pattern within the set-up. Reflections can be kept below 1 % if high-quality
waveguide loads are used. Furthermore, the uncertainty can be compensated by performing
supplementary measurements with a λ/4-shifted load and averaging of the two readings.
At lower frequencies (e.g. below 750 MHz), TEM cells can be employed rather than
waveguides. However, the field inside a TEM cell is less well-defined, i.e. there is rather large
deviation from the predicted homogeneous field distribution [20]. The probe is generally
inserted through small holes in the walls of the TEM cell and positioned at the centre (above
or below the septum) where the field is mostly homogeneous over the probe dimensions.
Each sensor is evaluated with respect to the field component parallel to the sensor.
E.3.2.3 Sensitivity in media (second step)
E.3.2.3.1 General
The sensitivity in tissue-equivalent medium is determined by generating locally-known field

strength values inside the media. Two methods can be used:
a) transfer calibration with temperature probe;

b) calibration with analytical fields.
E.3.2.3.2 Transfer calibration with temperature probes
In lossy media, SAR is related both to the electric field (E) and the rate of temperature rise
(dT/dt) in the medium having specific heat capacity c h . Hence, based on Formula (E.4)
E2 dT
SAR σ= ch
= , (E.4)
ρ dt t = 0
the electric field in lossy media can be measured indirectly by measuring the rate of
temperature rise in the medium. Non-perturbing temperature probes (optical probes or
thermistor probes with resistive lines) with small sensors (< 2 mm) and fast response time
(< 1 s) are available and can be easily calibrated with high precision [50]. The set-up and the
exciting source have no influence on the calibration; only the relative positioning uncertainties
of the temperature probe and the E-field probe to be calibrated shall be considered. However,
several problems limit the available accuracy of probe calibrations with temperature probes.
a) The rate of temperature rise is not directly measurable, but shall be evaluated from
temperature measurements that are measured over a short time duration. Special
precaution is necessary to avoid measurement uncertainties caused by temperature
gradients due to energy-equalizing effects or convection currents in the medium. Such
effects cannot be completely avoided. With a careful set-up, these uncertainties can be
kept small.
– 160 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
b) The measured volume around the temperature probe is not well defined. It is difficult to
calculate the energy transfer into the probe from a surrounding gradient temperature field
into the probe (typically, temperature probes are calibrated in medium with homogeneous
temperatures). There is no traceable standard for temperature increase measurements.
c) The calibration depends on the assessment of the mass density, the specific heat
capacity, and the electrical conductivity of the medium. While mass density and specific
heat capacity can be measured accurately with standardized procedures (on the order
of ±2 % for c h ; much better for ρ), there is no standard for the measurement of the
electrical conductivity. Depending on the method and medium, the uncertainty can be
±5 %.
d) Sufficient temperature rise is required to produce measurable temperature rises; therefore
calibration is often performed at a higher power level than the E-field methods. The non-
linearities in the system (e.g. power measurements, different field components, etc.) shall
be compensated.
Considering these problems, the calibration accuracy of E-field probes using the temperature-
rise technique in a carefully designed set-up is about ±10 % (combined standard
uncertainty) [48]. A set-up using a combination of waveguide and temperature-rise techniques
was presented in [51]. The estimated combined standard uncertainty of this set-up is ±5 %
when the same medium is used for both the calibration and for actual measurements, and
±7 % to ±9 % when not, which is in good agreement with the estimates given in [48].
E.3.2.3.3 Uncertainty analysis for transfer calibration with temperature probes
When performing an uncertainty analysis of the transfer calibration using the temperature-rise
technique, at least the parameters listed in Table E.1 shall be considered.
Table E.1 – Uncertainty analysis for transfer calibration using temperature probes
Standard
Input quantity X i Probability Uncertainty uncertainty
Divisor ci ui ν i or ν eff
(source of uncertainty) distribution ±%
±%
Positioning of E-field probe R √3 1 ∞
Positioning of temperature probe R √3 1 ∞
E-field probe linearity R √3 1 ∞
Temperature probe drift and noise R √3 1 ∞
Temperature probe linearity R √3 1 ∞
Medium conductivity R √3 1 ∞
Specific heat of medium R √3 1 ∞
Medium density R √3 1 ∞
Temperature probe accuracy R √3 1 ∞
Combined standard uncertainty RSS
NOTE c i is the sensitivity coefficient.
The component tolerances of Table E.1 shall be determined as follows:
a) The positioning tolerances of the temperature and E-field probes are evaluated according
to O.7.1 using the actual penetration depth determined from the tissue-equivalent medium
dielectric properties measured at the calibration frequency. Since SAR variations at the
peak location are much smaller in directions parallel to the phantom surface than in
directions normal to the phantom surface, the procedures of O.7.1 are applicable when
both temperature and E-field probe movements are limited to parallel directions only.
IEC/IEEE 62209-1528:2020 – 161 –
© IEC/IEEE 2020
b) Linearity uncertainty of the field probe is assessed according to Clause O.1, and for
calibration shall not exceed 0,1 dB at the calibration field strength.
c) Temperature probe drifts and noise are assessed by temperature measurements at 1 s
intervals for 1 h, with an integration time of less than 0,5 s in a constant temperature
condition. The temperature tolerance is computed as 100 × [(T max − T min )/∆T min ], where
∆T min is the minimum temperature rise for the different power levels used for the
calibration.
d) The temperature probe linearity tolerance should be determined using the following
procedures. The accuracy and linearity of the temperature probe readings are compared
against a traceable temperature reference at 10 temperature steps in a range greater than
or equal to that used during the calibration. The tolerance is computed as 100 × [(T max −
T min )/∆T min ], where ∆T min is the larger of the minimum temperature rise for the different
power levels used for the calibration or the tolerance for the temperature reference.
e) Medium conductivity measurement tolerance during temperature calibration is determined
using the same procedures as in O.9.3.
f) The specific heat tolerance of tissue-equivalent media shall be determined using
calorimeter procedures ([51], [52]).
g) Medium density measurement tolerance shall be computed according to the RSS of the
volume and weight tolerances obtained using standard measurement methods for volume
and weight.
h) The temperature probe shall have a step response time of 1 s or less, which is specified
as the time required by the measurement equipment, temperature probe, and readout
electronics to reach 90 % of the expected final temperature value, after a step variation of
5 °C or more has been applied to the temperature probe.
E.3.2.3.4 Calibration with analytical fields (waveguides)
In this method, a test set-up is used in which the field can be calculated analytically from
measurements of other physical magnitudes (e.g. input power). This corresponds to the
standard field method for probe calibration in air; however, there is no standard specified for
fields in lossy media.
When using calculated fields in lossy media for probe calibration, several points shall be
considered in the estimation of the uncertainty.
• The net RF power dissipated in the waveguide shall be measured accurately. This
requirement implies precise measurements of two out of the following three quantities:
incident power, reflected power, reflection coefficient at the waveguide input port.
• The accuracy of the calculated field strength will depend on the assessment of the
dielectric properties of the medium.
• Due to the small wavelength in media with high permittivity, higher order modes can be
excited. The field distribution in the set-up shall be carefully checked for conformity with
the theoretical field distribution.
Waveguides can be utilized to generate an analytically known field inside tissue-equivalent

media, e.g. the set-up presented in [53]. In this set-up (Figure E.1), the upper part of a
vertically-oriented open-ended waveguide is filled with liquid. A dielectric slab at a
distance > λ (λ refers to the wavelength in the air section of the waveguide) from the feeding
coupler provides an impedance match (> 10 dB return loss) between air and liquid. The
symmetry of the construction and high losses in the liquid ensure that the field distribution
inside the tissue-equivalent medium follows the TE 10 waveguide mode pattern, although
higher-order modes theoretically could exist. In [53] the absence of higher modes was
carefully validated by measuring the electric field distribution in the volume of the liquid, and
this was found to be within ±2 % of the theoretical TE 10 mode pattern.
– 162 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Key
x, y, z axes of Cartesian coordinate system
3δ liquid depth (> 3 times the penetration depth)
a waveguide cross-section width
b waveguide cross-section height
Pf incident power
Pr reflected power
Figure E.1 – Experimental set-up for assessment of the sensitivity
(conversion factor) using a vertically-oriented rectangular waveguide
Inside the liquid, the field propagates almost like a TEM wave, because of the low cut-off
frequency. The liquid depth (> 3 times the penetration depth) was chosen so that the
reflections at the upper surface of the liquid are negligible. Formula (E.5) shows the
relationship between the SAR at the cross-sectional centre (x = y = 0) of the lossy waveguide
and the longitudinal distance (z) from the dielectric slab:
4( Pf − Pr ) −2 z / δ
SAR( z ) = e (E.5)
ρabδ
where
ab is the cross-sectional area of the waveguide;

Pf is the forward power inside the air-filled section of the waveguide;
Pr is the reflected power inside the air-filled section of the waveguide;
z is the distance above the upper surface of the dielectric slab;
ρ is the liquid density;
δ is the penetration depth inside the lossy liquid.
NOTE For the purposes of this document, the density ρ is assumed to be 1 000 kg/m 3 .
The penetration depth δ, which is the reciprocal of the waveguide-mode attenuation

coefficient α, is determined from a scan along the z-axis and compared with the theoretical
value determined from Formula (E.6) using the measured dielectric properties of the lossy
liquid.
IEC/IEEE 62209-1528:2020 – 163 –
© IEC/IEEE 2020
{ }
−1
2
α −1 =
δ= ℜ ( π a ) + jωμ 0 ( σ + jωε0 εr′ ) (E.6)
Table E.2 provides design guidelines for calibration waveguides with a return loss of at least
10 dB at the most important frequencies used for personal wireless communications. Values
for the penetration depth for these specific fixtures and tissue-equivalent medium dielectric
properties are also listed in Table E.2.
Table E.2 – Guidelines for designing calibration waveguides
Waveguide
Dielectric slab a
Frequency dimension
MHz a Thickness
εr′
mm mm
300 584,2 5,5 106,0
450 457,2 6,0 66,1
900 247,6 5,6 34,8
1 450 129,5 4,7 24,8
1 800 to 2 000 109,2 4,8 19,4
2 450 109,2 5,7 12,6
3 000 86,4 5,7 10,3
3 500 58,2 4,9 9,76
3 900 58,2 4,9 9,76

4 200 58,2 4,9 9,76
4 600 47,5 3,5 9,50
4 900 47,5 3,5 9,50
5 400 47,5 5,6 5,73
6 000 40,4 5,4 5,25
7 000 35,0 3,1 20,0
9 000 23,0 3,1 16,0

a Permittivity and thickness of the dielectric slab may vary from the
values shown to accommodate commercially available materials.
If the dielectric slab permittivity varies from the indicated value by
more than 2 %, it is recommended to newly optimize the slab
thickness for the best matching (return loss typically greater than
10 dB).
NOTE 1 The waveguides are filled with tissue-equivalent medium

having dielectric properties as given in Table 2.
NOTE 2 By convention, the length of the cross-section short edge is

one half that of the long edge, i.e. b = a/2.
NOTE 3 The waveguide dimensions are in accordance with the EIA

RS-261-B: 1979 [54].
NOTE 4 These dimensions are also dependent on the frequency

bandwidths of interest.
– 164 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
This technique provides excellent accuracy, with a combined standard uncertainty of < ±3,6 %
depending on the frequency and media. The calibration itself is reduced to power
measurements traceable to a standard calibration procedure. The practical limitation given by
the waveguide size to the frequency range between 750 MHz and 6 000 MHz is not severe in
the context of compliance testing, because most of the operational frequencies for mobile
communications systems are covered within this range. For frequencies below 750 MHz,
transfer calibration with temperature probes remains the most practical way to achieve
calibration with the lowest uncertainties.
E.3.2.3.5 Uncertainty analysis for calibration with analytical fields (waveguides)
When performing an uncertainty analysis of the probe calibration with analytical fields in a
waveguide, at least the parameters included in Table E.3 shall be considered.
Table E.3 – Uncertainty analysis of the probe calibration in waveguide
Probability Uncertainty u(y) =

Input quantity X i Divisor u(x i ) = c i (1 g, ν i or
distribution a(x i ) c i ∙u(x i )
(source of uncertainty) qi a(x i )/q i 10 g) ν eff
PDF i ±% ±%
Incident or forward power R √3 1 ∞
Reflected power R √3 1 ∞
Liquid conductivity ∞
R √3 1
measurement
Liquid permittivity ∞
R √3 1
measurement
Liquid conductivity deviation R √3 1 ∞
Liquid permittivity deviation R √3 1 ∞
Frequency deviation R √3 1 ∞
Field homogeneity R √3 1 ∞
Field-probe positioning R √3 1 ∞
Field-probe linearity R √3 1 ∞
Combined standard uncertainty RSS
E.3.3 One-step calibration procedure – reference antenna method
E.3.3.1 General
The procedure of E.3.2 can be combined into a one-step procedure using reference antennas.
The total field shall be evaluated according to Formula (E.2). Reference antennas are small
antennas designed for operating within the appropriate tissue-equivalent medium. Examples
of antennas developed at 900 MHz and 1 800 MHz have been described in [49], [55]. At least
two identical antennas are necessary to evaluate the gain of the main lobe for use in probe
calibration.
The one-step methods in E.3.3 and also E.3.4 assess the isotropy with respect to probe
rotation and polarization of the incident field. The method in E.3.4 additionally considers the
angle of incidence.
Typical antennas used include simple directive patch antennas with a gain of 5 dBi or higher
(see Figure E.2 for set-up). Because side lobes caused by multiple reflections at the tank wall
could interfere with the calibration, it shall be verified that these are sufficiently attenuated by
the tissue-equivalent medium before they reach the probe tip.
IEC/IEEE 62209-1528:2020 – 165 –
© IEC/IEEE 2020
Figure E.2 – Illustration of the antenna gain evaluation set-up
E.3.3.2 Reference-antenna gain measurement
The reference antenna gain evaluation is performed according to the following protocol:
a) Position the antennas in the tissue-equivalent medium such that their main-beam axes are
aligned and at a well-defined distance d, where d ≥ 2 D 2 λliquid , D is the largest dimension
of the reference antenna, and λ liquid is the wavelength in the medium. The antennas shall
be at least 10 cm from any walls of the container.
b) Measure the reflection coefficients ρ 1 and ρ 2 at the input ports of each antenna.
c) Measure the transmission coefficient S 21 between the two antennas at the same ports.
d) The gain is described in Formula (E.7):
 4πd  1
S21(d ) × eαd 
G= × (E.7)
 λ liquid 
  (1− ρ1
2
)( 1− ρ2
2
)
where the attenuation coefficient is described in Formula (E.8):
1  ε ′′  
2πf
( )
1/ 4
α= × εr′2 + εr′′2 × sin  arctan  r   (E.8)
c  2  εr′  
E.3.3.3 Uncertainty analysis for reference-antenna gain measurement
When performing an uncertainty analysis of the reference antenna measured gain, at least the
parameters included in Table E.4 shall be considered. Supplemental methods for the
calculation of the antenna gain can be found in E.3.3.6.
– 166 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table E.4 – Uncertainty template for evaluation of reference antenna gain

Input quantity X i Divisor u(x i ) = ν i or
distribution a(x i ) ci c i ∙u(x i )
(source of uncertainty) qi a(x i )/q i ν eff
PDF i ±% ±%
Incident power R √3 1 ∞
Reflection coefficients R √3 1 ∞
Distance R √3 1 ∞
Medium permittivity R √3 1 ∞
Combined standard RSS √3
uncertainty
E.3.3.4 Reference antenna method probe calibration procedure
The following protocol shall be used for evaluating the sensitivity coefficients of the probe.
a) Position one antenna in the tissue-equivalent medium. The antenna shall be at a minimum
distance of 10 cm from the walls of the container.
b) Connect a power source to the input port of the reference antenna. The theoretical electric
field E th at a distance d a from the antenna is given by Formula (E.9):
=
 P 1− Γ 2 Ge-2αd a
 in
Eth  ×
( 120π ) 2

(E.9)


2
4πd a ℜ εr { } 


where
da is the distance from the antenna to the geometric centre of the sensors where
d ≥ 2 D 2 λliquid ;
D is the largest dimension of the reference antenna;

λ liquid is the wavelength in the medium;
G is the antenna gain;
P in is the input power;
Γ is the reflection coefficient of the antenna.
It is recommended to connect a bidirectional coupler to control the input power. Tune the
input power so that E th ~ 30 V/m.
c) Position the probe in the tissue-equivalent medium so that the centre of the detectors is at
a distance d from the antenna.
d) Orient the probe in order to align the direction of the sensor with the polarization of the
reference antenna.
e) Measure the voltage V 1,meas at the transmitting antenna port.
2
f) The sensitivity coefficient K 1 for this antenna is K1 = V1,meas / Eth
g) Repeat steps d) through f) for the other two sensors to evaluate K 2 and K 3 .
h) Using the sensitivity coefficients of the probe, SAR is determined by the Formula (E.10):
σ  V1,meas V2,meas V3,meas 

SAR=  + +  (E.10)
ρ  K1 K2 K3 
where ρ is the density (1 000 kg/m 3 ).

IEC/IEEE 62209-1528:2020 – 167 –
© IEC/IEEE 2020
E.3.3.5 Uncertainty analysis for reference-antenna method probe calibration
When performing an uncertainty analysis of the calibration with reference antennas, at least
the parameters included in Table E.5 shall be considered.
Table E.5 – Uncertainty template for calibration using reference antenna

PDF i ±% ±%
Incident power R √3 1 ∞
Reflection coefficients R √3 1 ∞
Antenna gain N 1 or k 1 ∞
Medium permittivity R √3 1 ∞
Probe positioning R √3 1 ∞
Combined standard RSS
uncertainty
E.3.3.6 Extended formulas for the antenna gain in the near-field measurement
The antenna gain G in Formula (E.7) and the theoretical electric field strength E th are based
on the far-field measurement in the tissue-equivalent medium. However, it is often difficult to
measure S 21 in the far-field region due to large attenuation of the field in the tissue-equivalent
medium. In such case, an extension of the Friis transmission formula in the lossy medium can
be introduced to define the near-field gain of the reference antenna G near (d), and then E th in
the near-field region can be estimated [56]. The near-field gain G near (d) is a function of the
distance from the antenna d, and can be expressed by the power series of the inverse of d.
Then the procedure to determine the reference antenna gain G in E.3.3.1 a) is modified as
follows.
a) Position the antennas in the medium such that their main-beam axes are aligned and at a
well-defined distance d, and λ liquid is the wavelength in the medium. The antennas shall be
at least 10 cm from any walls of the container.
b) Measure the reflection coefficients ρ 1 and ρ 2 at the input ports of each antenna.
c) Measure the transmission coefficient S 21 (d) between the two antennas at the same ports
and the same distance by changing d.
d) The near-field gain G near (d) can be expressed as in Formula (E.11):
S (d ) exp(αd ) ⋅ (2 βd ) a a 
Gnear (d ) = 21 =G × exp  1 + 2 + ...  (E.11)
2 2 d 2
(1− ρ1 ) ⋅ (1− ρ2 )  d 
where
d is the distance from the antenna;
S 21 (d) is the transmission coefficient between the two antennas;
ρ1 is the reflection coefficient of the antenna 1;
ρ2 is the reflection coefficient of the antenna 2;
α is given by Formula (E.8);
β is 2π/λ liquid ;
G is the antenna gain in dB representation given in Formula (E.12):
– 168 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
 1 1 
GneardB ( d ) = GdB 1+ +  (E.12)
 A × d B×d2 
where
G dB is the antenna gain.
The constants A, B or a 1 , a 2 ,… can be determined by curve-fitting. Therefore the choice of
proper curve fitting for the range of d is important. Attenuation and phase constants α and
β in the liquid are obtained from the dielectric properties of the medium.
e) Then, E th (d) in the near-field region is given by Formula (E.13):
2
Eth ( d ) =
(
30 Pin 1− Γ
2
)G
near ( d )e
-2αd
(E.13)
ℜ { εr }d 2
where
α is given by Formula (E.8);
εr is the complex relative permittivity;
G near (d) is the near-field gain;
Γ is the reflection coefficient of the antenna.
E.3.4 One-step calibration procedure – coaxial calorimeter method
E.3.4.1 General
Figure E.3 shows a coaxial calorimeter system for calibrating SAR probes by the temperature
rise method at frequencies below 450 MHz, from [57]. It consists of a coaxial line with air
dielectric, having an outer shield diameter 2,3 times the inner conductor diameter, which gives
50 Ω characteristic impedance. Towards the short-circuit end of the line, the inner conductor
is replaced with a plastic pipe containing the tissue-equivalent medium, which is in direct
contact with the metal end of the centre conductor and also the shorting plate at the top of the
system. The probe is inserted through an aperture in the centre of the shorting plate. The
diameter of the liquid in the pipe shall be the same as that of the metal inner conductor, and
this shall be sufficiently large to prevent probe proximity errors from being significant. Power
applied to the input port results in an axially polarized field in the liquid, with a fairly uniform
distribution below 150 MHz.
IEC/IEEE 62209-1528:2020 – 169 –
© IEC/IEEE 2020
Figure E.3 – Schematic of the coaxial calorimeter system
The SAR probe is used to measure the electric field in the liquid at the reference point for
approximately 1 W input power. A temperature probe is then used to measure the rate of
temperature rise in the liquid at the same position, using approximately 30 W. For accuracy, it
is important that the actual temperature rise is limited to 1 °C, since the conductivity of the
medium changes by between 3 % and 5 % per °C. The ratio of the power for the SAR probe
measurement and the temperature rise measurement shall be determined using a directional
power sensor, or directional coupler and power sensor, although the absolute power is not
required. Calibration is by comparing the linear output voltages from the SAR probe sensors
with the electric field in the liquid that is calculated using Formula (E.14):
2 ρ × ch dT Psensor
Eprobe = (E.14)
σ dt Ptemperature
where
E 2 probe is the electric field squared at the probe sensors;

ρ is the density of the phantom liquid;
σ is the conductivity;
ch is the specific heat;
P sensor is the power used in the SAR probe measurement;
P temperature is the power used in the temperature measurement;
dT/dt is rate of increase in temperature in the liquid.
This system provides accurate calibrations over the frequency range 30 MHz to 450 MHz, with
a standard uncertainty of around ±5 %.
E.3.4.2 Uncertainty analysis for coaxial calorimeter probe calibration
When performing the calibration using this method, at least the parameters included in
Table E.6 shall be considered.
– 170 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table E.6 – Uncertainty components for probe calibration using thermal methods

PDF i ±% ±%
Liquid temperature ∞
R √3 1
coefficient
Liquid heat capacity R √3 1 ∞
Liquid density R √3 1 ∞
Liquid conductivity R √3 1 ∞
Power sensor linearity ∞
R √3 1
and drift
Temperature probe ∞
R √3 1
linearity
Thermal effects giving ∞
rise to non-linear R √3 1
temperature rise
Positional errors N 1 1 ∞
Combined standard
RSS
uncertainty
E.4 Isotropy
E.4.1 Axial isotropy
The probe shall be exposed to a wave with incident polarization normal to the probe axis. The
axial isotropy is determined by rotating the probe along its major axis from 0° to 360° with a
step size less than or equal to 15°.
E.4.2 Hemispherical isotropy
E.4.2.1 General
The probe shall be exposed to a reference wave with varying angles of incidences relative to
the plane normal to the probe axis. Assessment of the hemispherical isotropy shall be done at
a location where the SAR gradients are less than 3 % per mm; this can be achieved by using
standing waves. The hemispherical isotropy shall be determined by either tilting the probe, or
changing the polarization of the reference wave. The angles of incidence shall vary from 90°
(axial) to 0° (normal), with a step size less than or equal to 30°. For each incidence angle, the
probe shall be rotated through a range of 360° with a step size less than or equal to 15°.
The following four methods can be used for hemispherical isotropy, each producing similar
results:
a) flat phantom with dipole at side (see E.4.2.2);

b) flat phantom with dipole beneath (see E.4.2.3);
c) spherical phantom with dipole (see E.4.2.4);
d) reference antenna method (see E.4.2.5).
E.4.2.2 Isotropy with flat phantom and dipole at side
The set-up consists of a thin-wall plastic box filled with tissue-equivalent medium exposed to
a half-wave resonant dipole operating at the test frequency. The following protocol shall be
used for evaluating the spherical isotropy of the probe.
IEC/IEEE 62209-1528:2020 – 171 –
© IEC/IEEE 2020
a) Mount the dipole antenna horizontally on a mounting device and position it parallel to the
flat phantom (see Figure E.4). The antenna shall be positioned at a maximum distance of
e = λ/10 from the adjacent wall of the medium container.
b) Insert the probe vertically in the liquid so that the centre of the three probe sensors is
positioned in front of the dipole feed-point at the height of the dipole axis.
c) The horizontal position of the probe shall be, whenever possible, at the maximum of the
standing wave near the back side of the box, at a distance c from the phantom/liquid
interface, where the E-field is partially homogeneous and the H-field is at a minimum.
d) The dipole shall be rotated around the axis of its mounting device from at least 0° to 180°
with incremental steps of less than or equal to 30°.
e) At each step the probe is rotated about its axis from 0° to 360° by the probe positioner,
and measurement data is recorded in steps of less than or equal to 15°.
The deviation from spherical isotropy is then expressed as an uncertainty component with a
rectangular probability distribution limited by the measured response peaks. Details can be
found in [22].
Key
c Distance from probe to backwall

e Maximum distance from the front wall of the liquid container
NOTE The assessment is preferably performed in the location c of minor field gradients resulting from the
standing wave. The dimension e is the distance between the dipole feed-point and the interface of the front wall
and the liquid.
Figure E.4 – Set-up to assess hemispherical

isotropy deviation in tissue-equivalent medium
– 172 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
E.4.2.3 Isotropy with flat phantom and dipole beneath
An alternative set-up for isotropy evaluation is shown in Figure E.5. A half-wave dipole
mounted on a holder is positioned with arms parallel to and below a flat phantom containing
tissue-equivalent medium. A probe is positioned at a point directly above the dipole feed-
point. All probe rotations are performed and controlled by a high-accuracy positioner. The
rotation of the probe about its axis (ϕ rotation from 0° to 360°), along with dipole rotation
(θ rotation from 0° to 180°) and change of probe inclination angle ϑ (from 0° to 75°), produces
a 3D evaluation of a significant portion of the hemispherical receiving pattern.
Key
ϑ probe axis tilt angle

θ rotation angle of the axis of the dipole mounting device
ϕ probe axis rotation angle of the probe
Figure E.5 – Alternative set-up to assess hemispherical

isotropy deviation in tissue-equivalent medium
E.4.2.4 Isotropy with dipole and spherical phantom
The hemispherical isotropy of the probe receiving pattern can be assessed using a set-up
comprised of a spherical flask filled with a tissue-equivalent medium and a dipole [58]
(see Figure E.6). The probe sensors are positioned at the geometrical centre of the flask, and
the probe is rotated about its axis under each excitation condition. The probe is maintained in
its vertical orientation during the procedure.
The excitation conditions are provided by the dipole antenna that is external to the flask and
moved at different locations to provide different incidence angles and polarizations. The
dipole axis is aligned with the tangent plane of the surface of the sphere above the dipole
feed-point. The impedance of the antenna is matched for this condition at the operating
frequency.
NOTE Smaller flasks can be preferred because increased medium conductivity or operating frequency implies a
higher path-loss for the wave transmitted to the probe. Larger flasks feature a locally flatter surface (longer
curvature radius), which produces a locally flatter wave front impinging on the probe tip and relaxes the tolerance
on the position of the dipole.
IEC/IEEE 62209-1528:2020 – 173 –
© IEC/IEEE 2020
Figure E.7 [58] shows the conventions used to describe the antenna position and polarization.
A 3D measurement of the probe receive pattern over a hemisphere is accomplished by
rotating the probe by 360° around its axis (ϕ) while positioning the external dipole at angles
0° ≤ ξ ≤ 360°. Every desired polarization of the impinging wave is achieved by rotating the
dipole around its axis 0° ≤ θ ≤ 180°. Multiple θ-positions are not needed when ξ = 0°. While a
large number of measurement points are possible with this set-up, usually steps of 30° in ξ
and 15° in θ are sufficient for characterization.
The measurements consist of complete 360° rotations of the probe for each dipole position
and polarization (ξ, θ). The power radiated during each rotation shall be monitored and
recorded for every ϕ-position of the probe, and the SAR samples shall be normalized to a
nominal value. Although it cannot be a priori excluded, there is no particular reason why any
positions and polarizations of the external dipole other than what is considered here would
produce significantly worse isotropic performance. However, the matrix of tests can be
expanded to include any particular case of interest, e.g. polarization conditions that match the
sensor orientation in some probes.
As with the flat-phantom isotropy test, a rectangular uncertainty distribution is assumed.

However, the uncertainty can be substantially reduced when probe calibration is performed
under a specific polarization and direction of the impinging wave and the same conditions are
maintained during SAR measurements.
Key
ξ elevation angle to dipole axis

θ dipole rotation angle
ϕ probe axis rotation angle
Figure E.6 – Experimental set-up for the

hemispherical isotropy assessment
– 174 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Key
ξ elevation angle to dipole axis

θ dipole rotation angle
NOTE ξ is the angle between the probe axis and the axis of the dipole holder, and θ is the angle between the
probe axis and the dipole axis.
Figure E.7 – Conventions for dipole position (ξ) and polarization (θ)
E.4.2.5 Isotropy with reference antennas
The following protocol shall be used for evaluating the isotropy of the probe using the
reference antenna set-up shown in Figure E.8.
a) Position one antenna in the tissue-equivalent medium. The antenna shall be at a minimum
distance of 10 cm from the walls of the liquid container, and have dimensions such that it
can be placed in the position indicated in Figure E.8 .
b) Position the probe in the liquid so that the geometric centre of the sensors is at a distance
d from the antenna, where d ≥ 2 D 2 λliquid , D is the largest dimension of the reference
antenna, and λ liquid is the wavelength in the liquid. It is recommended that the SAR value
be set to between 0,5 W/kg and 1 W/kg at this position.
c) Orient the probe axis so that its main axis is orthogonal to the direction of propagation
from the antenna (see Figure E.8).
d) Rotate the probe along its main axis from 0° to 360° with a step size less than or equal to
15°. Record the SAR values. The axial isotropy is expressed as an uncertainty component
with a rectangular probability distribution based on the measured response peaks.
e) Vary the incidence of the reference field by tilting the reference antenna or the probe axis
(see Figure E.8) from 0° to 90° with a step size of 15° or 30°.
f) For each incidence angle, rotate the probe along its main axis from 0° to 360° with a step
size less than or equal to 15°. Record the SAR values.
g) The hemispherical isotropy is expressed as an uncertainty component with a rectangular
probability distribution limited by the measured response peaks.
IEC/IEEE 62209-1528:2020 – 175 –
© IEC/IEEE 2020
Dimensions in centimetres
Key
θ probe axis rotation angle

Φ probe axis tilt angle
d ≥ 2 D 2 /λ liquid
Figure E.8 – Measurement of hemispherical isotropy with reference antenna
E.5 Lower detection limit
The lower detection limit is the minimum local SAR measurable within the overall system
uncertainty. It is related to the noise level and offset of the measurement system and can be
assessed by varying the output power in the same set-up as specified in Clause O.1 (i.e. by
verifying that the probe is still responding linearly at the required lower detection SAR level).
In actual operational conditions of the measurement system, the ambient EM environment
could impair the detection limit. Therefore, verification of the lower detection limit by using the
flat-phantom set-up specified in Clause G.3, or a calibration waveguide, is recommended. The
lower detection limit shall be 10 mW/kg or lower.
– 176 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
E.6 Boundary effect
In the closest vicinity to the inner surface of a phantom shell, probe sensitivity deviates from
that which is established under typical calibration conditions. Boundary effects are evaluated
with the liquid-filled open waveguide set-ups used for probe calibration. The psSAR is
measured using all system components and compensation routines. The uncertainty due to
boundary effects is the deviation from the analytical value at the surface, which is estimated
by extrapolating the trend of measured samples to the interface of the liquid and the dielectric
slab. This measurement shall be performed for each frequency band and for each averaging
volume. For frequencies below 800 MHz, where calibration waveguides might not be available
due to the relatively large size, an experimental set-up comprised of a half-wave dipole
beneath a flat phantom may be used. In this case, extrapolation to the surface is not based on
the known analytical behaviour of guided-wave modes, but on an extrapolation of measured
samples. The boundary effect error is specified as the deviation between the measured SAR
data and the true value in the liquid. When the probe is oriented normal to the phantom
surface, the boundary effect can largely be compensated for as described in [53]. The
uncertainty of the boundary effect shall be evaluated according to Clause O.3.
E.7 Response time
The response time uncertainty is the measurement error caused by movement of the probe
during measurement. Field-probe signal response time uncertainty is evaluated by exposing
the probe to an E-field step response producing at least 100 W/kg. The signal response time
is determined as the time required by the probe and its readout electronics to reach 90 % of
the expected final value produced by the step response through switching the RF power on
and off. The probe shall remain stationary at each measurement location for at least three
times the assessed response time to ensure a negligible probe signal response time
uncertainty. Under these measurement conditions, an uncertainty value of zero may be
entered in Table 9 (see 8.3). Otherwise, the SAR uncertainty due to signal response-time
uncertainty shall be assessed using the signal characteristics of the test device. In this case,
the signal step-response time uncertainty is equal to the percentage difference of the SAR
measured with the chosen measurement time from the SAR measured with a measurement
time of at least three times the assessed response time. A rectangular probability distribution
shall be assumed.
IEC/IEEE 62209-1528:2020 – 177 –
© IEC/IEEE 2020
Annex F
(informative)
Example recipes for phantom tissue-equivalent media
F.1 General
The dielectric properties of the phantom shall be those listed in Table 2 of 6.2.2.
WARNING – To ensure personnel safety, users shall follow the instructions provided in the
material safety data sheet (MSDS) for any material, and/or any local regulations.
F.2 Ingredients
The following ingredients are used in the example liquid recipes for producing the head
tissue-equivalent media:
a) sucrose (sugar) (> 98 % pure);

b) sodium chloride (salt) (> 99 % pure);
c) de-ionized water (16 MΩ cm resistivity minimum);
d) hydroxyethyl cellulose (HEC);
e) bactericide;
f) diethylene glycol butyl ether (DGBE) (> 99 % pure);
g) diethylene glycol monohexylether;
h) polyethylene glycol mono [4-(1,1,3,3-tetramethylbutyl) phenyl ether]. This is available as
Triton™ X-100. 3 The quality of the Triton X-100 shall be ultra-pure to match the
composition of salt. After making the Triton X-100 based liquids, the dielectric properties
are difficult to adjust in order to keep them close to the targets. Therefore, it is
recommended to use alternatives when available;
i) diacetin;
j) 1,2-propanediol;
k) polyoxyethylene (20) sorbitan monolaurate (Tween 20);
l) emulsifiers;
m) mineral oil.
NOTE 1 Viscosity of HEC-based tissue-equivalent medium liquids should be low enough not to affect E-field
probe movement.
NOTE 2 Add salt to water first to make a saline solution, then add the Triton X-100.
NOTE 3 Actual results and mixture percentages may vary from those shown, depending on grade and type of
components used.
NOTE 4 The tolerances of ε r and of σ in Table F.1 and Table F.2 are liquid temperature tolerances described in
O.9.6 based on measurements of the applicable liquid recipes.
NOTE 5 The formulas containing Triton X-100 are under review and verification.
___________
3 Triton is the trademark of a product supplied by The Dow Chemical Company or an affiliated company of Dow.
This information is given for the convenience of users of this document and does not constitute an endorsement
by IEC of the product named. Equivalent products may be used if they can be shown to lead to the same
results.
– 178 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
F.3 Tissue-equivalent medium liquid formulas (permittivity/conductivity)
Table F.1 and Table F.2 provide suggested recipes.
Table F.1 – Suggested recipes for achieving

target dielectric properties, 30 MHz to 900 MHz
Frequency (MHz) 30 50 144 450 835 900

Recipe source 3 3 2 2 3 2 4 2 2 4
number
Ingredients (% by weight)
De-ionized water 48,30 48,30 53,53 55,12 48,30 48,53 56 50,36 50,31 56
Tween 20 44,70 43,31 49,51 48,39 48,34
Oxidized mineral 44 44
oil
Diethylenglycol
monohexylether
Triton X-100
Diacetin 50,00 50,00 50,00
DGBE
NaCl 1,60 1,60 1,77 1,57 1,60 1,96 1,25 1,35
Additives and salt 0,10 0,10 0,10
Measured temperature dependence
Temp. (°C) 21 21 21 20 21 21 20
ε liquid temp. unc. (%) 0,8 0,1 0,1 0,1 0,04 0,04
σ liquid temp. unc. (%) 2,8 2,8 2,6 4,2 1,6 1,6
IEC/IEEE 62209-1528:2020 – 179 –
© IEC/IEEE 2020
Table F.2 – Suggested recipes for achieving target
dielectric properties, 1 800 MHz to 10 000 MHz
Frequency (MHz) 1 800 2 450 4 000 5 000 5 200 5 800 6 000 8 000 10 000
Recipe source number 2 4 4 4 4 1 1 4 5 5
Ingredients (% by weight)
De-ionized water 54,23 56 56 56 56 65,53 65,53 56 67,8 66,0
Tween 45,27 31,1 33,0
Oxidized mineral oil 44 44 44 44 44
Diethylenglycol 17,24 17,24
monohexylether
Triton X-100 17,24 17,24
Diacetin
DGBE
NaCl 0,50
Additives and salt
Measured temperature dependence
Temp. (°C) 21 20 20 20 20 22 22 20 20 20
ε liquid temp. unc. (%) 0,4 1,7 1,8
σ liquid temp. unc. (%) 2,3 2,7 2,6
NOTE 1 Multiple columns under a single frequency indicate optional recipes.
NOTE 2 Recipe source numbers: 1 verified by different labs, 2 Reference [59], 3 developed by IT’IS Foundation, 4
developed by IT’IS Foundation, 5 Reference [60].
NOTE 3 The values of ε liquid temp. unc. and σ liquid temp. unc. are liquid temperature uncertainties described in O.9.6,
based on measurements of the applicable liquid recipes given above. These are not part of the original publications
but have been subsequently developed by the project team.
NOTE 4 The recipes at 8 000 MHz and 10 000 MHz are sufficiently broadband that they cover the frequency
range of 6 000 MHz to 10 000 MHz within a tolerance of ±10 % for permittivity and conductivity.
– 180 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex G
(normative)
Phantom specifications
G.1 Rationale for the phantom characteristics
G.1.1 General
Phantoms that represent the human anatomy are essential components of electromagnetic
exposure evaluation. It is important to define and standardize the features, dimensions and
material properties of the phantoms that are directly relevant to ensuring conservative
exposure. It is unnecessary and impractical to specify all anatomical details. The rationales
for the SAM phantom and flat phantom used in this document are provided in G.1.2 and
G.1.3, respectively. Application specific phantoms are specified in Annex K and Annex L.
G.1.2 Rationale for the SAM phantom
Anthropometric studies of the human population are available to guide the specification of a
conservative and realistic head phantom. Specifically, a large 1988 anthropometric study of
U.S. Army personnel was particularly important in providing data to determine population
coverage requirements for adults of different ages and ethnic groups [61]. Based on this
study, the Specific Anthropomorphic Mannequin (SAM) is specified in this document to
represent exposure of the head. The shape of the standard SAM phantom is established by a
number of lines and curves as illustrated in Figure G.1 (see G.2.1).
Dosimetric studies suggest that larger heads (adult, male user) couple more energy and
constitute a worst-case exposure scenario compared to smaller heads (children) [62], [63],
[64], [65]. The upper 90th percentile (adult male) head dimension shall be used to ensure that
well over 90 % of all users have smaller heads and would therefore sustain a lower exposure.
The back ear protrusion has a direct effect on determining the closeness of the handset and
antenna to the head and is an important parameter for determining SAR. The ear protrusion
was adjusted to simulate the pressure of the device on the ear. The shape of the ear is a
factor in positioning the handset and was designed to enable correct and reproducible
positions. A study on mobile phone exposure of the head using anatomical head phantoms of
adults and children showed that the psSAR assessed with the SAM phantom yields
conservative exposure for both the head and pinna tissue when regarding the pinna as an
extremity that has SAR limits of at least double that of the head tissues [66].
G.1.3 Rationale for the flat phantom
The flat phantom is established to represent the torso and other regions of the body for this
document. The shape and size of the flat phantom are important for accurate and
representative SAR measurements. The part of the user’s body exposed to the RF emissions
of a wireless communication device is not always well specified and might vary with product
design and usage
The flat phantom should not be excessively large compared with the size of a human torso.
Large phantoms are also more difficult to construct, and SAR measurement systems might not
be capable of performing measurements in phantoms that are very large or very deep.
A flat phantom represents a simple engineering structure for which the SAR can be easily and
repeatably measured. It is an open-top thin dielectric shell which is filled with a tissue-
equivalent medium. The physical characteristics of the standard flat phantom are intended to
simulate the exposure characteristics of a human body.
IEC/IEEE 62209-1528:2020 – 181 –
© IEC/IEEE 2020
A flat-bottomed phantom provides maximal exposed surface area of flat regions of the body to
the wireless (DUT). Therefore, the flat phantom generally provides conservative exposure.
Flat phantoms should be large enough to allow complete coupling of the RF radiating antenna
and to allow scanning of 1 g and 10 g volumes.
The use of a flat phantom as a standard phantom for SAR evaluations of devices not
operating next to the ear is intended to represent maximum coupling of variations in energy
absorption through irregular heterogeneous tissue boundaries. The flat phantom will produce
conservative exposure in an equivalent human body, when used with prescribed test
distances.
G.2 SAM phantom specifications
G.2.1 General SAM phantom specifications
The dimensions of the SAM phantom have been derived from selected 90th percentile male
head data reported by Gordon et al. [61], and are listed in Table G.1 and shown in Figure G.1.
Table G.2 shows other pertinent dimensions for comparison. At the ear reference point (ERP),
a thickness of 6 mm including the 2 mm shell is selected to model the external ear (pinna).
This thin ear spacer also simulates users with small ears, and gives conservative exposure. A
system of reference points and reference lines is used to correlate the DUT positioning with
the phantom (Figure G.2 and Figure G.3). The point “M” is the mouth reference point, “LE” is
the left-ear ERP, and “RE” is the right-ear ERP. The ERPs are 15 mm posterior to the
entrance to ear canal (EEC) along the line B-M (back-to-mouth) (Figure G.2).
The plane passing through the two ear reference points and M is specified as the reference
plane. The line N-F (neck-to-front) is along the front truncated edge of the ear. Both N-F and
B-M lines shall be marked on the external phantom shell to facilitate DUT positioning. These
may be etched into the external surface of the phantom without affecting the specifications.
Optional markings shown in Figure G.3 may also be labelled on the phantom. Posterior to the
N-F line, the thickness of the phantom shell with the shape of an ear is a flat surface. The ear
is 6 mm thick at the ERP. Anterior to the N-F line, the ear is truncated as illustrated in
Figure G.2. The ear truncation is introduced to facilitate repeatable DUT positioning.
The phantom shall be filled with head tissue-equivalent medium. To minimize reflections
within the phantom, the depth of the homogeneous medium should be at least 15 cm. Medium
depth of less than 15 cm can be used if it is demonstrated (e.g. using numerical simulations)
that the effect of reflections on psSAR is less than 1 %. If the effect of reflections on psSAR is
more than 1 % but less than 3 %, the maximum uncertainty shall be added to the uncertainty
budget.
– 182 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure G.1 – Illustration of dimensions in Table G.1 and Table G.2

IEC/IEEE 62209-1528:2020 – 183 –
© IEC/IEEE 2020
Table G.1 – Dimensions used in deriving SAM phantom from

the ARMY 90th percentile male head data (Gordon et al. [61])
Ref. Anatomy Gordon report [mm] SAM Deviation

[mm] [%]
Mean value Standard 90th
deviation percentile
H15 Bitragion Chin Arc 325,8 13,4 343,1 329,3 −4,0
H16 Bitragion Coronal Arc 353,3 12,9 369,7 367,3 −0,6
H18 Bitragion Frontal Arc 304,3 10,6 318,2 314,1 −1,3
H19 Bitragion Submandibular Arc 304,2 14,5 323,2 333,5 3,2
H20 Bitragion Subnasale Arc 292 11,1 306,3 305,3 −0,3
H60 Head Breadth 151,7 5,4 158,6 158,4 −0,1
H61 Head Circumference 567,7 15,4 587,3 594,8 1,3
H62 Head Length 197,1 7,1 206 206,0 0,0
H77 Menton-Sellion Length 121,9 6,5 130,4 125,0 −4,1
H80 Neck Circumference 379,6 19,7 405,3 395,4 −2,5
Table G.2 – Additional SAM dimensions compared with selected

dimensions from the ARMY 90th percentile male head data
(Gordon et al. [61])—specialist head measurement section
Ref. Anatomy Gordon report [mm] SAM Deviation

[mm] [%]
Mean Standard 90th
value deviation Percentile
H3 Bigonial Breadth 118,9 7,9 129,2 130,0 0,6
H6 Bitragion Breadth 144,8 6 152,3 152,7 0,2
H10 Chin-Back of Head 194,2 10,3 207,3 206,5 −0,4
H11 Chin-Top of Head 216,8 8,9 228,3 220,4 −3,5
H18 Glabella-Back of Head 199,7 7,2 208,5 209,2 0,3
H19 Glabella-Top of Head 96,2 7,3 105,6 104,4 −1,1
H30 Menton-Top of Head 232,0 8,8 243,3 246,7 1,4
H36 Sellion-Back of Head 197 7,1 205,9 205,3 −0,3
H37 Sellion-Top of Head 112 6,9 120,9 121,7 0,7
H38 Stomion-Back of Head 199,4 9,6 211,9 211,4 −0,2
H39 Stomion-Top of Head 186,3 7,8 196,3 196,3 0,0
H40 Subnasale-Back of Head 203,5 8,3 213,6 213,0 −0,3
H42 Subnasale-Top of Head 161,9 7,7 171,8 177,6 3,4
H43 Tragion-Back of Head 98,9 8,5 106,4 106,4 0,0
H44 Tragion-Top of Head 131 5,7 138,2 138,2 0,0
– 184 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Key
B Direction of B-M line back endpoint

F Direction of N-F line front endpoint
N Direction of N-F line neck endpoint
M Mouth reference point
RE Right-ear ear reference point (ERP)
Figure G.2 – Close up side view of phantom showing the ear region
IEC/IEEE 62209-1528:2020 – 185 –
© IEC/IEEE 2020
Key
B Line B-M back endpoint

M Line B-M front endpoint
N Line N-F neck endpoint
F Line N-F front endpoint
NOTE The full-head model is shown for illustration purposes only. The procedures in this document are derived
primarily for the phantom set-up shown in Figure G.4.
Figure G.3 – Side view of the phantom showing relevant markings
G.2.2 SAM phantom shell specification
The phantom shell with integral solid ear spacer, as specified in the SAM CAD model, shall be
constructed from chemical-resistant, low-permittivity and low-loss material, with relative
permittivity between 3 and 5; however, relative permittivity less than 3 is acceptable for
frequencies up to 3 GHz. The loss tangent of the phantom shell shall be less than or equal to
0,05. It has been shown that at higher frequencies, particularly above 3 GHz, the permittivity
of the phantom shell has an increasing impact on SAR measurement uncertainty [36] .
According to [36], the SAR is affected by the shell even though the shell is thin and has
dielectric properties much lower than the tissue-equivalent medium. Specifically, it was shown
that the variation in the SAR depends not only on the frequency, but also on the relative
permittivity of the shell and the distance between the antenna and the material. It was
recommended that the effects of the shell be taken into account in the higher frequency
range. Therefore, uncertainty evaluation shall be conducted above 3 GHz according to
8.4.2.3.
The shape of the phantom shell shall have a tolerance of less than ±0,2 mm with respect to
the SAM CAD file. In any area within the projection of the DUT, the shell thickness shall be
2 mm ± 0,2 mm, except for the ear (where the thickness is specified in the CAD file with a
tolerance of ±0,2 mm) and the extended perimeter walls for the SAM phantom (see
Figure G.4). The phantom shell shall be made of materials resistive to the tissue-equivalent
media (e.g. those listed in Annex F), to avoid damage and preserve the ±0,2 mm tolerances.
This minimal thickness results in a higher SAR compared to measurements on thicker
phantoms. Smaller thicknesses are not recommended due to problems with mechanical
strength when holding the medium. For non-critical areas, i.e. the central strip containing the
nose in the SAM phantom, as shown in Figure G.5, the thickness tolerance is allowed to be
within ±1 mm.
The SAM phantom CAD file cross section for the reference plane is given in Figure G.6.
– 186 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
NOTE The CAD model is for use in manufacturing the SAM phantom by groups having the necessary expertise
and resources. Four *.IGES files of inner and outer surfaces for the left and right halves extracted from the CAD
model of the SAM phantom are provided with this document at:
http://www.iec.ch/dyn/www/f?p=103:227:0::::FSP_ORG_ID,FSP_LANG_ID:1303,25
The four file names are:

– sam_lft_in_6_11.igs
– sam_lft_out_6_11.igs
– sam_rt_in_6_11.igs
– sam_rt_out_6_11.igs
Figure G.4 – Sagittally bisected phantom with extended perimeter

(shown placed on its side as used for device SAR tests)
Key
B Line B-M back endpoint

M Line B-M front endpoint
N Line N-F neck endpoint
F Line N-F front endpoint
LE Left-ear ear reference point (ERP)
Figure G.5 – Picture of the phantom showing the central strip

IEC/IEEE 62209-1528:2020 – 187 –
© IEC/IEEE 2020
Figure G.6 – Cross-sectional view of SAM at the reference plane
G.3 Flat phantom specifications
The shell of the flat phantom shall be constructed in the form of an open-top container with a
flat bottom. In this document, flat phantoms shall be used for system validation, system
check, and for measurements of SAR for DUTs using positions specified in 7.2.4.1.
Any flat phantoms used shall be large enough to allow the measurement of SAR in 1 g and
10 g volumes (medium density of 1 000 kg/m 3 is used) with an influence from the phantom
shape and dimensions of less than 1 %. Minimum flat phantom dimensions are specified in
6.2.4 and Clause G.4.
NOTE Use of a flat phantom as a standard phantom for SAR evaluations of body-worn and body-supported
devices is intended to represent maximum coupling between the DUT and phantom compared to a significant
majority of exposure situations involving people. Such coupling to the boundary while maintaining prescribed
distances is likely to produce conservative exposure.
The phantom shall be filled with a tissue-equivalent medium. To minimize reflections within
the phantom, the depth of the homogeneous medium should be at least 15 cm ± 0,5 cm.
Medium depth of less than 15 cm can be used if it is demonstrated (e.g. using numerical
simulations) that the effect of reflections on psSAR is less than 1 %. If the effect of reflections
on psSAR is more than 1 % but less than 3 %, the maximum uncertainty shall be added to the
uncertainty budget.
The phantom shell shall be made of low-loss and low-permittivity material, having a loss
tangent ≤ 0,05 and a relative permittivity of
– 188 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
ε r′ ≤ 5 for f ≤ 3 GHz
3 ≤ ε r′ ≤ 5 for f > 3 GHz
The thickness of the bottom-wall of the flat phantom shall be 2,0 mm with a tolerance of
±0,2 mm.
The phantom shell shall be made of materials resistive to the tissue-equivalent media
(e.g. those listed in Annex F), to avoid damage and preserve the ±0,2 mm tolerances.
When filled with the required depth of medium, any sagging of the bottom surface of the
phantom within the projection of the DUT shall be determined and considered in the
uncertainty budget (see 8.4.2.4). For system check and validation testing, the sagging at the
medium interface and inner surface of the phantom directly above the source (e.g. dipole or
waveguide) shall be less than 1 % of the free-space wavelength in the frequency range of
800 MHz to 10 000 MHz, and less than 0,5 % of the free-space wavelength at frequencies
below 800 MHz.
When the preceding requirements are met, the effect of the shape and shell thickness on the
repeatability of SAR measurement results is less than 1 %. Effects on the SAR due to the
influence of the deviation from the shell parameters and thickness shall be included in the
uncertainty estimation.
G.4 Justification of flat phantom dimensions
The influence of the dimensions of the flat phantom (Figure G.7) on the absorbed energy in a
10 g cube inside the medium-only phantom (without box) was assessed numerically using a
commercial FDTD code. The phantom was illuminated with a matched dipole antenna at a
distance of 15 mm (0,042 λ at 840 MHz). The dimensions of the phantom (W and L) were
separately varied between 0,4 λ and 3 λ. The power absorbed in the cube was normalized to a
feed-point current of 1 A and a feed-point power of 1 W. A length and width of at least 0,6 and
0,4 wavelength in air, respectively, was found to ensure that the effect of phantom dimensions
on SAR is less than 1 % for both current and power normalization (Figure G.8). The
dimensions of the phantom set-up can be scaled in terms of the free-space wavelength. The
dependence on the medium properties was found to be insignificant.
Because of its larger size, a 10 g averaging cube will be more sensitive to dimension
changes, i.e. the uncertainty associated with the 1 g average will be smaller than that of the
10 g average. The effects resulting in differences depend on perturbations of the dipole
current magnitude and spatial distribution. Since the dipole dimensions are large compared
with the SAR averaging volumes, the perturbations will increase with volume size. Although
the depth used in this study was 10 cm, rather than the 15 cm required for the flat phantom in
Clause G.3, it is 2,57 times the penetration depth at 840 MHz, and therefore the power
reflection at the medium surface is negligible (less than 1 %).
The numerical target SAR values of Table D.2 (see D.2.2) were calculated using the FDTD
method. The parameters for the lossless phantom shell used in simulations (dimensions, shell
thickness, and permittivity), and the distance s between the reference dipole and tissue-
equivalent medium, are given in Table G.3. The phantom dimensions in this table produce the
same SAR values as a phantom that meets the requirements of Clause G.3, within the
uncertainty of the SAR test system. The phantom dimensions recommended in Clause G.3
are to be used for system check and system validation. The dielectric properties used for the
tissue-equivalent medium are specified in Table 2 (see 6.2.2) and the dimensions of the
reference dipoles are shown in Table D.1 (see D.2.1).
IEC/IEEE 62209-1528:2020 – 189 –
© IEC/IEEE 2020
Key
λ free-space wavelength
NOTE A 10 g cube is shown at the bottom centre of the flat phantom.
Figure G.7 – Dimensions of the flat phantom set-up used for deriving the
minimal phantom dimensions for W and L for a given phantom depth D
– 190 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure G.8 – FDTD predicted error in the 10 g psSAR as a function of the dimensions of
the flat phantom compared with an infinite flat phantom at 800 MHz
Table G.3 – Parameters used for calculation of reference SAR values in Table D.2
Phantom Reference dipole

Phantom shell
Frequency dimensions used distance s from
permittivity
for FDTD models the medium
mm
MHz mm
x; y; z
300 3,7 1 000; 800; 170 15

450 3,7 700; 600; 170 15
750 3,7 700; 600; 170 15
835 3,7 360; 300; 150 15
900 3,7 360; 300; 150 15
1 450 3,7 240; 200; 150 10
1 500 3,7 220; 160; 150 10
1 640 3,7 220; 160; 150 10
1 750 3,7 220; 160; 150 10
1 800 3,7 220; 160; 150 10
1 900 3,7 220; 160; 150 10
1 950 3,7 220; 160; 150 10
2 000 3,7 160; 140; 150 10
2 100 3,7 160; 140; 150 10
2 300 3,7 160; 140; 150 10
2 450 3,7 180; 120; 150 10
2 600 3,7 180; 120; 150 10
IEC/IEEE 62209-1528:2020 – 191 –
© IEC/IEEE 2020
Phantom Reference dipole
Phantom shell
Frequency dimensions used distance s from
permittivity
for FDTD models the medium
mm
MHz mm
x; y; z
3 000 3,7 220; 160; 150 10

3 500 3,7 174; 110; 150 10
3 700 3,7 174; 110; 150 10
5 000 3.7 90; 80; 35 10
5 200 3,7 90; 80; 35 10
5 500 3,7 90; 80; 35 10
5 800 3,7 90; 80; 35 10
7 000 3,7 150; 150; 75 5
9 000 3,7 150; 150; 75 5
NOTE The SAR values in Table D.2 at frequencies above 3 GHz depend on the dipole
spacer and detailed construction of the dipoles and may vary by as much as ±10 %. The
reasons are that the dipole dimensions are short with respect to arm diameter and spacer
dimensions, i.e. the numerical reference values are not generic and need to be determined
for a particular test set-up. Also, the results can be sensitive to the permittivity of the
phantom shell [36].
G.5 Rationale for tissue-equivalent media
The purpose of Clause G.5 is to provide the rationale for the choice of target permittivity and
conductivity for the tissue-equivalent medium used for the SAR measurements. The rationale
is that the SAR induced by the inductively coupled devices is insensitive to the permittivity of
the body tissue [67], [68].
The aim was to provide target properties for a homogeneous phantom that, when used in the
SAM (head) and flat (body) shells, will provide 1 g and 10 g SAR values that are conservative
with respect to the exposure in the human body for a significant majority of the exposed
population and RF sources and device positions that fall within the scope of this document.
The same tissue-equivalent medium is used for both head and body SAR assessment.
The absorption of RF energy in the body is dependent on the nature of the RF-source and its
distance from the body, the region(s) of the body (and therefore the tissue types) that are
exposed, and the penetration depth of the energy into these tissues. The layers of different
tissues within the body can result in enhanced RF absorption at certain frequencies, and there
are anatomical variations in these layers due to factors such as age, sex, BMI, etc.
The homogeneous flat phantom specified in this document is based partly on the tissue
properties obtained from [69], and partly from simulation of homogeneous and layered
structures to represent the body. Over the frequency range 0,15 GHz to 10 GHz, the specified
real relative permittivity for the phantom is approximately 10 % greater than that of brain grey
matter, and conductivity similar to that of brain white matter and muscle tissue (Figure G.9).
Grey matter has lower permittivity than white matter or muscle, so this would be expected to
be conservative for these materials (for a homogeneous phantom, increasing the conductivity
and/or reducing the real relative permittivity will result in higher SAR values, provided that the
loading effect on the transmitter is unchanged). Note that the conductivity at some
frequencies has been increased to account for enhancement of absorption due to tissue
layering in the body [2], and these simulations are described in more detail below.
NOTE Clause G.5 describes the rationale for selecting the tissue-equivalent medium dielectric properties used for
this document. The same set of parameters is used for head and body SAR measurements.
– 192 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Below 0,15 GHz, brain and muscle tissue measurements appear to show a rapid increase in
real relative permittivity as the frequency is decreased [69], and these become larger than the
permittivity of water, so that a liquid phantom with these properties cannot be realized. Thus a
constant target value for relative permittivity of 55 with conductivity of 0,75 S/m has been
adopted for frequencies at and below 0,03 GHz.
In the case of the wireless power transfer systems operating below 30 MHz, and where it has
been shown and documented that the peak normalized magnetic field strength (i.e. the ratio of
the peak magnetic field strength to its reference level) exceeds the peak normalized electric
field strength (i.e. the ratio of the peak electric field strength to its reference level) by 10 dB or
more, a saline solution having identical conductivity (0,75 S/m) within the given tolerance
(±10 %) may be used as a tissue-equivalent medium.
The effect of the tissue layers on the psSAR in the head was investigated in [2] for a region
near the ear, and for 0,3 GHz to 6 GHz. The tissue layers were varied in composition and
thickness to represent the anatomical variation of the exposed head region, covering the user
group including adults and children (between the 10th percentile and 90th percentile). Based
on the worst-case tissue layer compositions with respect to the energy absorption at each
frequency, tissue-equivalent medium liquid dielectric properties for homogeneous modelling
were derived to ensure that the psSAR is the same (or slightly higher) than the psSAR in the
layered models. The dielectric properties of the tissues used to model the human head were
computed by the 4-term Cole-Cole formula [69]. The validity of this approach for near-field
exposure was demonstrated by replacing the plane-wave by half-wave dipole and quarter-
wave monopole sources in the closest proximity to the layered structures, as well as by
comparison with results from magnetic-resonance-imaging (MRI)-based non-homogeneous
human head models of adults and children from the literature. Table 2 (see 6.2.2) shows the
relative permittivity and conductivity of the resulting tissue-equivalent medium liquids as
published in [2].
The study of [2] was extended for general body tissue composition [70]. It was found that
standing wave effects due to reflections in the subcutaneous adipose tissue lead to a
significant increase of SAR in comparison to the findings of [2]. This increase cannot be
compensated by modifying the dielectric properties of the tissue-equivalent medium liquids. A
comprehensive analysis of the coupling mechanism shows that the standing wave effects
need only to be considered in the Fresnel zone and the far-field zone of the DUT [71]. At
close distances, conservative exposure can be achieved using the parameters for tissue-
equivalent medium liquids as proposed in [2]. Therefore, the established liquid parameters
have been retained for the measurements of hand-held and body-worn devices up to 200 mm
from the body.
Extensive large-scale numerical studies have been performed to validate the

conservativeness of the SAM and flat phantoms by comparing with SAR induced in anatomical
head and body models. The first such study [14] demonstrated conservativeness of the SAM
phantom at 835 MHz and 1 900 MHz for wireless devices operating at the head. This work
has been extended to frequencies in the 300 MHz to 6 000 MHz range. The results obtained
from the nine participating organizations show that the SAR produced in SAM for this
frequency range is higher than the head SAR in the anatomical adult and child head models.
The details of this study are reported in Bit-Babik [15]. A similar study has been performed for
the flat phantom in the 30 MHz to 6 GHz frequency range [72] using several different
anatomically correct human body models [73]. The results show the overall conservative
nature of the flat phantom tissue-equivalent medium liquid parameters. Among the hundreds
of exposure conditions simulated as part of this study, there were also some cases where
SAR was underestimated by the flat phantom; follow-up analysis has been performed to
investigate the related temperature rise in the anatomical models that could be caused by
those exposure conditions. Those additional results show that the temperature increase level
is not much different among all exposure conditions, including those selected cases where
SAR was underestimated by the flat phantom.
IEC/IEEE 62209-1528:2020 – 193 –
© IEC/IEEE 2020
Studies also have been performed to assess whether a homogeneous phantom with the
properties specified in this document is conservative with respect to absorption in the layers
of tissues within the head for 6 GHz to 10 GHz. These studies showed that the phantom was
conservative for the frequency range for a region above the ear, and also the cheek region
[60] and [74].
Simulations showed that the tissue-equivalent medium dielectric properties proposed in

Table 2 provide an overestimation of the SAR from exposure to body-worn or hand-held
devices covered under this document used at 4 MHz and 150 MHz. Additional details on
tissue-equivalent medium from 4 MHz to 30 MHz are provided in IEC TR 62905:2018 [18] (on
exposure assessment methods for wireless power transfer systems) .
For dielectric properties of head tissue-equivalent medium liquid at other frequencies within
the frequency range, a linear interpolation method shall be used. Examples of recipes for
liquids having parameters as specified in Table 2 are given in Annex F. The measured
dielectric properties, not the values of Table 2, shall be used in the SAR calculations. This
measurement can be performed using the equipment and procedures specified in Annex H.
a) Real relative permittivity b) Conductivity
NOTE Above 1 GHz, the permittivity decreases approximately linearly with frequency, and the conductivity
increases parabolically.
Figure G.9 – Complex permittivity of human tissues

compared to the phantom target properties
G.6 Definition of a phantom coordinate system and a DUT coordinate system
The offset and rotation between these optional reference coordinate systems can be used to
provide a clear description of device position relative to the phantom. A definition of the
phantom coordinate system for the left-ear ERP is illustrated in Figure G.10.
The x-, y- and z-axes form a right-handed Cartesian coordinate system. For the left-ear ERP,
the axes are specified as follows.
• The z-axis is determined by a connecting line between the left-ear and the right-ear ERPs,
and points from the right to the left from the point of view of the phantom. The z = 0 origin
is at the left-ear ERP.
• The y-axis lies in the reference plane along the B-M line (Clause G.2).
• The x-axis is perpendicular to the reference plane along the N-F line (Clause G.2), and
cuts the reference plane at the left-ear ERP.
For the right-ear ERP, the reference coordinate system can be established analogously. A
definition of a DUT coordinate system is shown in Figure G.11.
– 194 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
For the flat phantom, the x-axis and y-axis are parallel to the flat phantom shell, and the z-axis
is perpendicular to it.
Key
x, y, z axes of the coordinate system relative to the phantom
Figure G.10 – Example reference coordinate system

for the left-ear ERP of the SAM phantom
Figure G.11 – Example coordinate system on a DUT

IEC/IEEE 62209-1528:2020 – 195 –
© IEC/IEEE 2020
Annex H
(informative)
Measurement of the dielectric properties of

tissue-equivalent media and uncertainty estimation
H.1 Overview
Clause H.2 describes the measurement of the dielectric properties of tissue-equivalent media
required to support the SAR measurement procedures. It provides sufficient details to enable
users to select a measurement method based on convenience and performance. The
dielectric properties of some reference media are provided in Clause H.6 to assess the
performance of the measurement procedures. General procedures to evaluate dielectric
properties measurement uncertainties are provided in Clause O.9.
H.2 Measurement techniques
H.2.1 General
The dielectric properties required are the complex relative permittivity εr =

εr′ − jσ ωε0 of the
tissue-equivalent medium at specific test frequencies and temperature. Several well-
established techniques can be used for dielectric properties measurements of tissue-
equivalent medium.
H.2.2 Instrumentation
The following or equivalent instrumentation is required:
• vector network analyser and S-parameter test set.

• the sample container for holding the liquid sample under test.
• application software to calculate sample dielectric properties from S-parameter
measurements.
Three measurement techniques and the corresponding test methodologies are specified in
Clause H.3 to Clause H.5. The achievable measurement accuracy for the different methods
can vary. The dimensions of the transmission line or coaxial probe are functions of the
measurement frequency ranges. The measurement accuracy is validated by measuring the
dielectric properties of reference liquids in Clause H.6.
H.2.3 General principles
The following general principles should be applied for all the procedures.
a) The sample holder is thoroughly clean.

b) All cell, probes, cables, and connectors are undamaged.
c) The procedure for filling the sample holder with the sample fills the volume completely
without air bubbles.
d) The temperature of the sample is recorded, and the report states that the dielectric
properties are applicable at that temperature only.
e) Measurement personnel are acquainted with the nature of the measurement and what to
expect at each stage of the procedure.
f) After calibration, a measurement is made on a reference liquid to validate the system prior
to measuring the sample. Data for several recommended reference materials are given in
Clause H.6.
– 196 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
g) Data reduction methods for relating complex reflection coefficient and complex permittivity
are accurate and appropriate for the sample-holder geometry being used.
H.3 Slotted coaxial transmission line
H.3.1 General
A terminated slotted coaxial line with a moveable probe can be used as a sample container
[75]. A network analyser provides the RF signal at the input of the slotted line, and enables
the magnitude and phase of the signal transmitted into the sample to be determined as a
function of position along the line by means of the moveable probe.
The test procedure should specify the network analyser calibration and settings for the
required frequency range, the starting measurement position, step size along the slot, and
total number of subsequent measurement positions. The application software should interpret
the measured data to yield the dielectric properties of the sample. An example procedure is
given in H.3.3.
H.3.2 Equipment set-up
The test equipment consists of a slotted coaxial transmission line with a probe connected to a
vector network analyser, as shown in Figure H.1. The sensing element of the probe is the
extended centre conductor of the coaxial cable/connector. The log-magnitude and phase of
S 21 should be displayed simultaneously. Source power should be set to a level high enough to
provide good signal-to-noise ratio. Since the measured quantities are magnitude and phase
changes versus distance, the accuracy of the scale is very critical.
The network analyser injects a signal into one end of the slotted coaxial transmission line.
The probe inserted through the slot into the tissue-equivalent medium liquid detects the RF
amplitude and phase for each measurement position along the length of the line. A full two-
port calibration of the network analyser should be carried out prior to filling the line with liquid,
and the following precautions should be observed.
a) Fill the slotted line carefully to avoid trapping air bubbles. This operation should be
performed while the slotted line is horizontal.
b) The probe should be inserted into the slot at the end nearest to the input connector of the
slotted line, so that the tissue-equivalent medium liquid is flush with the inside surface of
the line, and aligned with a well-defined position on the distance scale of the slotted line.
c) The probe should be inserted perpendicular to the slotted-line longitudinal axis until a
stable and adequate amplitude response is achieved. Do not insert the probe too deeply
into the coaxial line, because it can overly perturb the field distribution.
Although not shown, equipment for sample temperature monitoring is recommended.
Figure H.1 – Slotted line set-up

IEC/IEEE 62209-1528:2020 – 197 –
© IEC/IEEE 2020
H.3.3 Measurement procedure
a) Configure and calibrate the network analyser.
b) Measure log-magnitude and phase data at 10 to 20 positions along the slotted line,
corresponding to about a 30 dB change in magnitude.
c) Plot S 21 log-magnitude and phase against measurement distance.
d) Determine if the plotted points closely follow a straight-line approximation, based on the
correlation coefficient or a similar statistical measure. The data should produce a good
linear regression fit (expected correlation coefficient r 2 > 0,99 for lossy materials). If not,
re-measure the liquid by increasing the number of measurement points to extend the
magnitude change from 30 dB to 40 dB.
For low-loss materials, ensure that the slotted line is long enough to avoid reflections from
the load-terminated end.
e) Calculate the conductivity and relative permittivity of the tissue-equivalent medium liquid
using Formulas (H.1), derived from [76].
m ln (10 )
α= m [Np cm]
20
mp π
β= [rad cm]
180
2 (H.1)
ε′ =
( β ) − (α )
2
r
ω2 µ 0 ε 0
2αβ  100 cm 
σ=   [S m]
ωε0  m 
where
m m and m p are the slopes of the least-squares linear fits of the log-magnitude and
phase plots, respectively;
α and β are the average attenuation and propagation coefficients along the line.
H.4 Contact coaxial probe
H.4.1 General
Contact probes are open-ended coaxial transmission line sections, usually with an end flange
serving as a ground plane for fringing fields. Coaxial probe methods can be used over the
4 MHz to 10 GHz frequency range specified in this document. Multiple probes are typically
used to cover different frequency bands. A larger probe size has better sensitivity at lower
frequencies. A smaller probe can measure at higher frequencies due to the suppression of
non-TEM modes in the coaxial line. At frequencies below 100 MHz, special techniques may
be used to mitigate electrode polarization, which is caused by charge accumulation on the
probe conductors.
Measurements are made by placing the probe in contact with the tissue-equivalent medium
sample and measuring the admittance or reflection coefficient with respect to the open-circuit
end, using a network analyser or equivalent instrumentation [77], [78], and [79]. Coaxial
probes can also be used to test solid dielectrics, for example bulk samples of the materials
used to construct the device holder or phantom. To minimize errors with contact probes, solid
surfaces should be highly polished [80].
– 198 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The network analyser should be calibrated and the test procedures should specify the
equipment settings for the frequency range to be measured. The application software should
use the measured data to compute the dielectric properties of the tissue-equivalent medium
sample as a function of frequency. Coaxial probes are commercially available, including the
application software required for use with specific network analysers. The applicable
frequency range is a function of the probe size. An example procedure is given in H.4.3.
The equipment consists of a coaxial probe connected to one of the ports on a vector network
analyser. The open-ended coaxial line construction is shown in Figure H.2. Cylindrical
coordinates (ρ, ϕ, z) are used, where ρ is the radial distance from the probe axis, ϕ is the
angular displacement around the axis, z is the location along the axis, a is the inner conductor
radius, and b is the inner radius of the outer conductor. Coaxial probes typically include a
flange to satisfy the infinite ground-plane assumption used in the admittance calculations.
Key
a inner conductor radius
b outer conductor inner radius
z displacement along the axis
ε1 permittivity of the dielectric medium filling the space between a and b
ρ radial distance from the axis
ϕ angular displacement around the axis
(x, y, z) Cartesian coordinates
Figure H.2 – An open-ended coaxial probe with

inner and outer radii a and b, respectively
A non-metallic container that is sufficiently large compared to the size of the probe immersed
in it should be used to hold the tissue-equivalent medium sample. A probe with an outer
diameter b of 2 mm to 4 mm is suitable for tissue-equivalent medium measurements in the
300 MHz to 10 GHz range. This probe size requires sample volumes of 50 cm 3 or more.
Larger sample volumes are required for larger probes.
IEC/IEEE 62209-1528:2020 – 199 –
© IEC/IEEE 2020
The coaxial probe is connected to a network analyser and typically calibrated with open and
short terminations, and also in a known dielectric medium such as de-ionized water.
Commercially available probes are supplied with high precision shorting plugs for probe
calibration. The calibration procedures are highly automated by the application software
provided with commercially available probes. To ensure measurement accuracy, the probe
end should be clean and free of oxidation [81]. Calibration can be checked by measuring the
reference liquids of Clause H.6, and re-measuring the short circuit to ensure that a reflection
coefficient of Γ = −1,0 (linear units) is obtained consistently.
Large measurement errors due to “flange resonances” can occur when the diameter of the
flange is approximately equal to one half wavelength in the dielectric medium [82]. Such
effects are most pronounced for high-permittivity liquids that have a loss tangent less than
approximately 0,25 (at wireless device frequencies these include water, methanol, and
dimethyl sulfoxide). Therefore, calibration with a liquid having a high loss tangent,
e.g. ethanol, is strongly recommended for larger sensors. There could be problems calibrating
7 mm flanged coaxial sensors with water at some frequencies. Tissue-equivalent media have
a loss tangent of about 0,5, which is high enough to ensure that resonance effects are
practically nonexistent no matter what size of sensor is used.

a) Configure and calibrate the network analyser and probe system as required by the
application software.
b) Place the sample in a non-metallic container and immerse the probe into the liquid. A
fixture or clamp is recommended to stabilize the probe by mounting it with the probe face
oriented at an angle to the liquid surface to ensure that there are no air bubbles trapped
beneath the flange during insertion into the liquid.
c) Activate the application software to measure the complex admittance with respect to the
probe aperture.
d) The complex relative permittivity, εr = εr′ − jσ ωε0 , is automatically computed by the
application software; for example according to Formula (H.2) [79]:
j2ωεr′ ε0 b b π exp  − jω(μ 0 εr′ ε0 )1/ 2 r 

  dφ ′dρ′dρ
Y=
ln ( b a ) 
2 ∫a ∫ a ∫ 0 cosφ ′
r
(H.2)
where Y is the admittance of the probe in contact with the sample, the primed and
unprimed coordinates refer to source and observation points, respectively,
2
r2 =ρ2 + ( ρ′ ) − 2 ρρ′cosφ ′ , ω is the angular frequency, and j = √(−1).
Other numerical approaches may be used, e.g. [83], [84].
H.5 TEM transmission line
H.5.1 General
This method is based on measurement of the complex transmission coefficient of a TEM-

mode coaxial transmission line filled with the test liquid [85]. A transmission measurement
with a vector network analyser is used to determine the magnitude and phase of the
scattering coefficient S 21 , from which the complex permittivity is calculated. The network
analyser should be calibrated and the test procedures should specify the equipment settings
for the frequency range to be measured. The application software should use the measured
data to compute the dielectric properties of the liquid sample as a function of frequency.
– 200 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The measurement set-up is shown in Figure H.3. The liquid sample is contained in the open-
wall stripline consisting of a centre-conductor with a circular cross-section, two planar vertical
(side) ground conductors, one optically transparent plastic bottom wall, an open top, and a
temperature sensor. The length d of the TEM line is chosen for a given frequency range so
that the effect of multiple reflections inside the TEM line is small, and the total attenuation due
to the liquid within the line does not exceed the dynamic range of the network analyser. For
example, two TEM lines with different lengths can be used to cover the frequency range of
800 MHz to 2 000 MHz. The liquid sample should be carefully syringed or poured into the
TEM line through the open top to avoid any air bubbles.
Figure H.3 – TEM line dielectric properties test set-up [85]

a) Configure and calibrate the network analyser.
0
b) Record the magnitude and phase of S21 of the empty cell at the desired frequencies.
c) Inject sample liquid into the cell, and repeat the measurement to obtain the transmission
s
coefficient with the sample S21 .
d) Remove the liquid from the TEM line, thoroughly rinse the inside with water, and dry
carefully.
e) Reconnect the coaxial line to the network analyser. Make sure that the magnitude and
0
phase of S21 are unchanged (deviation less than 0,1 dB in magnitude and 0,5° in phase).
This step is needed only if multiple samples are to be measured without recalibration.
s 0
f) Calculate the complex permittivity of the liquid from the magnitude and phase of S21 S21
by a numerical solution of the Formulas (H.3).
IEC/IEEE 62209-1528:2020 – 201 –
© IEC/IEEE 2020
s
S21
=
(1− Γ2 ) exp  − j ( k − k0 ) d 
0
S21 1− Γ2 exp ( − j2kd )
1− εr
Γ=
1+ εr (H.3)
2πf
k= εr
c0
2πf
k0 =
c0
where
Γ is the reflection coefficient at either end of the TEM line;
k is the wave number in the liquid;
k0 is the free-space wave number;
d is the length of the sample holder (TEM line);
f is the frequency;
c0 is the free-space speed of light;
εr= εr′ − j σ ωε0 is the complex relative permittivity of the sample.
H.6 Dielectric properties of reference liquids
The procedures in Clause H.2 require measurements of reference liquids that have well-
established dielectric properties in order to validate the dielectric properties measurement
system. It is recommended to use a reference liquid listed in Table H.1 for validating the
dielectric properties measurement system. The difference between the measurement results
and the calculated dielectric properties (normalized for frequency and temperature) should be
within the expanded uncertainty of the dielectric properties measurement system. Two
reference liquids are required, one for calibration, and one for checking the calibration. A
general formula for calculating frequency-dependent dielectric properties is given by
Formula (H.4):
εs − ε∞ σ
εr =
ε∞ + + i (H.4)
β jωε0
1+ ( jωτ )1−α 
 
which corresponds to the Debye formula for α = 0 and β = 1, to the Cole-Cole formula for β = 1
and 0 < α < 1, and to the Cole-Davidson relationship for α = 0 and 0 < β < 1 [86], and where:
ε s is the static or low frequency permittivity;

ε ∞ is the asymptotic high-frequency permittivity;
τ is the relaxation time;
σi is the ionic conductivity.
The parameters for calculating the properties of several reference liquids using this formula
are given in Table H.1. These liquids are dimethyl sulfoxide, methanol, ethanediol, de-ionized
water. The parameters α and σ i are zero for all liquids listed in Table H.1. The model
parameters were obtained using various test methods described in the literature [86]. In
general, results obtained with two-port closed-transmission-line cells [87] are expected to be
the most accurate, but results from all sources listed in the table agree relatively well.
Table H.2 contains target 20 °C relative permittivity and conductivity values. If other reference
liquids are used, target values and literature references should be provided in the
measurement report.
– 202 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
It is a good practice to choose reference liquids with dielectric properties similar to tissue-
equivalent media. Any reference liquid mixture recipes should be followed exactly, and
dielectric properties should be measured at the specified temperatures. In order to avoid
contamination and evaporation, mixtures should not be left exposed to air. Reference liquids
should be high-purity grade, e.g. analytical grade or better. After opening, reagent bottles
should be stored in accordance with the manufacturer’s recommendation for the duration of
the recommended shelf life.
All personnel should be familiar with and apply any special handling procedures according to
the Material Safety Data Sheet (MSDS) for each particular liquid.
Table H.1 – Parameters for calculating the

dielectric properties of various reference liquids
Reference Temperature
Ref. Model εs ε∞ τ [ps] β
liquid °C
DI water a 20 [88] Debye 80,21 5,60 9,36 1
DI water a 25 [88] Debye 78,36 5,20 8,27 1
DMS b 20 [89] c Debye 47,13 6,80 21,07

1
DMS b 25 [89] c Debye 46,49 6,50 19,12

1
Cole-
Ethanediol 20 [89] 41,89 4,75 165,44 0,856
Davidson
Cole-
Ethanediol 25 [89] 41,75 4,70 133,70 0,859
Davidson
Methanol 20 [89] c Debye 33,64 5,65 56,40 1
Methanol 25 [89] Debye 32,66 5,56 50,70 1

a DI water is deionized water.
b DMS is dimethyl sulphoxide.
c Parameters from Gregory and Clarke [89] were derived from measurements to 5 GHz only; the
accuracy of these parameters above 5 GHz has not been evaluated.
IEC/IEEE 62209-1528:2020 – 203 –
© IEC/IEEE 2020
Table H.2 – Dielectric properties of reference liquids at 20 °C
Methanol DMS b DI water c Ethanediol

Frequency a a
[89] [89] [88] [89] a
MHz
εr′ σ (S/m) εr′ σ (S/m) εr′ σ (S/m) εr′ σ (S/m)
4 33,64 0,000 47,13 0,000 80,21 0,000 41,89 0,000
10 33,64 0,000 47,13 0,000 80,21 0,000 41,89 0,000
30 33,64 0,000 47,13 0,000 80,21 0,000 41,86 0,002
150 33,56 0,012 47,11 0,007 80,20 0,005 41,19 0,041
300 33,33 0,05 47,07 0,03 80,19 0,02 39,26 0,15
450 32,94 0,11 46,99 0,06 80,16 0,05 36,51 0,31
750 31,78 0,29 46,74 0,17 80,07 0,14 30,32 0,68
835 31,37 0,35 46,64 0,20 80,03 0,17 28,66 0,78
900 31,04 0,41 46,57 0,24 80,00 0,20 27,45 0,85
1 450 27,77 0,92 45,70 0,60 79,67 0,51 19,62 1,35
1 500 27,45 0,97 45,60 0,64 79,63 0,54 19,11 1,38
1 640 26,54 1,11 45,31 0,76 79,52 0,65 17,79 1,47
1 750 25,83 1,22 45,08 0,86 79,43 0,74 16,88 1,53
1 800 25,51 1,27 44,96 0,91 79,38 0,78 16,50 1,56
1 900 24,88 1,37 44,73 1,01 79,29 0,87 15,79 1,61
2 000 24,25 1,47 44,49 1,11 79,19 0,96 15,15 1,66
2 100 23,63 1,57 44,24 1,22 79,09 1,06 14,57 1,70
2 300 22,43 1,75 43,71 1,44 78,87 1,27 13,56 1,78
2 450 21,57 1,89 43,29 1,61 78,69 1,44 12,91 1,83
2 600 20,75 2,02 42,86 1,80 78,51 1,61 12,33 1,87
3 000 18,76 2,33 41,64 2,31 77,96 2,13 11,10 1,98
3 500 16,65 2,66 40,00 3,00 77,18 2,87 10,00 2,08
4 000 14,93 2,93 38,30 3,71 76,30 3,70 9,20 2,17
4 500 13,53 3,15 36,57 4,44 75,33 4,62 8,60 2,24
5 000 12,40 3,32 34,84 5,16 74,27 5,62 8,13 2,30
5 200 12,01 3,38 34,16 5,45 73,83 6,04 7,98 2,32
5 400 11,64 3,44 33,49 5,73 73,38 6,47 7,83 2,34
5 600 11,31 3,49 32,83 6,01 72,91 6,91 7,70 2,36
5 800 11,00 3,54 32,17 6,29 72,43 7,36 7,58 2,38
6 000 10,72 3,59 31,53 6,56 71,95 7,81 7,47 2,39
7 000 9,56 3,77 28,50 7,83 69,40 10,23 7,02 2,47
8 000 8,75 3,91 25,81 8,96 66,69 12,79 6,69 2,54
9 000 8,17 3,98 23,47 9,94 63,88 15,45 6,45 2,59
10 000 7,72 4,07 21,45 10,79 61,04 18,14 6,26 2,64

a Parameters from Gregory and Clarke [89] were derived from measurements to 5 GHz only; the accuracy
of these parameters above 5 GHz has not been evaluated.
b DMS is dimethyl sulphoxide.
c DI water is deionized water.
– 204 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex I
(informative)
Studies for potential hand effects on head SAR
I.1 Overview
This document is developed based on the principle of providing conservative exposure

measurement protocols to assess the head SAR of handset users according to presently
available science. Conservativeness of the SAM phantom is established by using the 90th
percentile head dimensions of adult male population [61], [63] to ensure that the exposure
corresponds to well over 90 % of the user population [62], [63], [65]. The dielectric properties
of the homogeneous tissue-equivalent medium required for SAR measurement have been
chosen to provide SAR overestimation when compared to results obtained from numerical
simulations using anatomical models [2]. In addition, a lossless spacer with the thickness of a
compressed ear is used on the SAM phantom to represent the external ear (pinna) so that the
handset is positioned at a very close distance to the head; therefore, SAR is not
underestimated.
For IEC 62209-1:2015, it was assumed that excluding the hand in the measurement
procedures, i.e. using a lossless holder with minimal disturbance of the RF fields, provides
conservative exposure. Since the publication of IEC 62209-1:2005, there have been some
studies reporting an increase in head psSAR when the hand is included in the SAR
assessment. The question of hand effects on head SAR was brought to the attention of
IEC TC 106 in 2008 [90], and IEC TC 106 has considered two options to address the
concerns of hand effects on head SAR.
a) Continue with the current test protocol and add a clause or annex to address hand effects
on head SAR.
b) Revise the head SAR measurement protocol to use a modified lossy device holder or hand
phantom.
A third option, to apply a correction factor to increase the measured head SAR according to
influences of the hand, was also discussed within IEC TC 106. However, there was not
enough data to establish an appropriate correction factor. In addition, the correction factor
methodology was not pursued, because it would affect the SAR assessment of all handsets
regardless of phone designs and actual hand influences on the head SAR.
The IEC National Committees were asked to vote on the preceding two options (or abstain).
The voting results favoured option a) by a majority of 15 to 8. Some National Committees also
requested continuing research be considered for this, to avoid delay in the completion of this
document. Thus, IEC TC 106 will continue to review and analyse contributions for hand
effects on head SAR. If the outcome of additional studies leads to a decision to modify the
current test protocol, future revisions of this document will be amended accordingly. Following
the National Committees voting results for Option a), Annex I was drafted to identify the
current knowledge on hand effects on head SAR.
I.2 Background
I.2.1 General
IEC TC 106 took action to address questions raised regarding hand effects on head SAR. Two
high priority large scale studies were initiated:
• Hand Effect Measurement Interlab (HEMI) study by six participating laboratories;

• computational study by six participating laboratories on SAM conservativeness using
different hand phantoms.
IEC/IEEE 62209-1528:2020 – 205 –
© IEC/IEEE 2020
Final results and conclusions for these studies [91], [92], [93] were presented in multiple
IEC TC 106 face-to-face meetings. The final report of the investigation on psSAR sensitivity to
hand and finger positions was presented in January 2013 [94]. Besides these studies, results
from several other projects were also presented at IEC TC 106 meetings [90], [95], [96], [97],
[98], [99], [100], [101], [102], [103], or published elsewhere [5], [104], [105], [106], [107],
[108], [109], [110], [111].
Both experimental and computational studies indicate that SAR variations are dependent on
hand grips, DUT position, form factor of the DUT, antenna design and placement, as well as
DUT operating frequency bands.
I.2.2 Hand phantoms
Four homogeneous hand phantoms were developed for the CTIA Certification Program Test
Plan V3.0 [112] to evaluate the over-the-air (OTA) performance of DUTs using pre-defined
grips. The physical dimensions of these four hand phantoms were derived from the average of
50th percentile hand data for men and women in anthropometric research publications ([113],
[114], [115], and [116]). The hand grip configurations used for the phantoms were established
based on studies of device size, usage, and form factor. The average wet and dry palm
dielectric properties reported in [87] were used for the CTIA hand phantoms. The average
anatomical characteristics were considered and were used in several experimental studies,
e.g. the HEMI study as well as [96], [98], and other computational studies.
I.3 Summary of experimental studies
I.3.1 Experimental studies using fully compliant SAR measurement systems
IEC TC 106 members conducted the HEMI study to investigate hand effects on head SAR.
Thirty-three DUTs with a diverse range of physical characteristics, marketed between June
2006 and June 2010, operating in the GSM 900 MHz and WCDMA 1 950 MHz bands, were
tested using a single hand phantom and grip configuration, involving 42 total test
configurations. Only six out of the 42 configurations show an increase in measured head SAR;
the increases ranged from 5 % to 40 %, with most of the increases less than 15 %. None of
the increases result in measured SAR above regulatory limits. All increases were in the
WCDMA 1 950 MHz band. The remaining WCDMA and GSM configurations showed
decreases of up to 78 %. The observed SAR increases were for phones exhibiting
comparatively low SAR, for instance the highest increase for 10 g psSAR, including the 40 %
increase, was 0,41 W/kg.
I.3.2 Experimental studies using other SAR measurement systems
The data collected in [96] using a fast SAR testing system (not fully compliant with all
normative clauses of this document) have showed both increases and decreases in SAR
when tested with a realistic hand phantom for 46 handsets in the 1 800 MHz frequency band;
where SAR decreases outnumbered SAR increases by a ratio of approximately 2 to 1. The
results in [96] and [108] also suggest that SAR increases occur more often at 1 800 MHz and
above when compared to lower frequency bands. In [98] and [117], cases are reported where
the head SAR increase introduced by hand effects can lead to both an unchanged local SAR
distribution or a modified distribution with a shift in peak SAR location. It was also shown that
the SAR is comparable for a DUT held by a CTIA hand phantom and a human hand
replicating the same grip [96].
– 206 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
I.4 Summary of computational studies
Several computational studies have been performed and reviewed by IEC TC 106 to
investigate hand effects on head SAR [91], [94], [100], using several simplified and detailed
CAD phone models at 900 MHz, 1 800 MHz, and 2 000 MHz. Several adult and child hand
models were used to simulate realistic hand grip conditions for exposures in several realistic
head models. The results were in general consistent with experimental data, showing a
considerable prevalence of cases with psSAR decreases due to presence of the hand, and
most of the SAR increases were observed at 1 800 MHz and 2 000 MHz.
The computational studies have also provided additional details on certain parameters that
can introduce noticeable influences on head psSAR. One study, [118] and [119], showed
higher SAR increases at frequencies around 1 800 MHz when the palm of a hand model was
positioned 1 cm to 2 cm above the antenna or peak SAR location. In another study [94],
involving parametric search for maximum SAR increase in varied hand grip conditions, a
significant SAR increase in both the SAM phantom and anatomical head models was
observed when fingers were in contact with the head. This was also observed in one of the
experimental studies using a fast SAR testing system (which is not fully compliant with all
normative clauses of this document) [96]. For frequencies below 2 GHz, the maximum
increase in head SAR is similar for both 1 g and 10 g psSAR [96] [120].
I.5 Conclusions
a) The data show that psSAR in the head can increase due to presence of the hand for some
configurations.
b) Both experimental and computational studies show:
1) head SAR increase was observed for a small percentage of the configurations at all
studied frequencies;
2) most head SAR increases were found at frequencies between 1 800 MHz and
2 000 MHz. Data for frequencies above 2 GHz are limited;
3) in a considerable majority of the cases where SAR increase was observed, the effect
was small.
c) No particular hand configuration or grip position could be identified which yields more
conservative results for all DUTs and/or for all testing conditions.
d) For a given DUT, methods do not seem to exist for predicting the transmission frequency
and realistic use condition that would result in the highest possible SAR enhancement.
e) The dependency of SAR increase on phone or antenna design cannot be easily predicted.
f) There is insufficient data to scientifically justify the implementation of correction factor
proposed by some National Committees.
g) IEC TC 106 will continue to review and analyse contributions related to hand effects on
head SAR.
IEC/IEEE 62209-1528:2020 – 207 –
© IEC/IEEE 2020
Annex J
(informative)
Skin enhancement factor
J.1 Background
In the course of the development of this document, several studies ([70], [71], [121]) indicated
that the measurements in a homogeneous phantom could result in localized SAR values lower
than the maximum values in a heterogeneous and anatomically realistic body model at
1 900 MHz. Since then, several research projects and simulations were undertaken within the
IEC TC 106 to verify these results. Simulations were undertaken with homogeneous phantoms
and layered models simulating skin, fat, and muscle. These findings were confirmed in
simulations for the metric SAR.
The exposure standard of ICES IEEE Std C95.1 and the recommendations by ICNIRP are
designed to protect against established adverse health effects. The ICNIRP guidelines in [3],
page 17, state:
“Established biological and health effects in the frequency range 10 MHz to a few GHz are
consistent with responses to a body temperature rise of more than 1 °C.”
Nevertheless, simulations for temperature result in different distributions compared to SAR

distributions (Figure J.1), with the temperature rise distribution being much smoother than the
corresponding SAR distribution. This result indicates that local SAR enhancements do not
typically produce equivalent temperature rise enhancement, due to the different nature of the
underlying physics: wave phenomena for SAR, and diffusion phenomena for temperature.
Figure J.1 – SAR and temperature increase (ΔT) distributions

simulated for a three-layer (skin, fat, muscle) planar torso model
– 208 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
J.2 Rationale
The reasons for the higher localized SAR values in heterogeneous tissue are standing-wave
effects which can occur under far-field-like exposure conditions if a tissue layer with low water
content, such as fat, bone or breast tissue, is enclosed between two tissues with high water
content, e.g. muscle, skin, most inner organs, etc. The most typical structure consists of a fat
layer between skin and muscle. Incident waves which pass the skin experience almost no
attenuation in the subsequent fat tissue layer. If the thickness of the fat layer corresponds to
approximately λ/4, the phase of the wave reflected at the muscle layer will lead to a standing
wave, with its maximum in the skin layer. This will lead to a significant increase of the local
SAR in the skin. Even if the SAR is averaged over a cubical volume, which in this case will
contain a comparatively large amount of fat tissue, the SAR measured in the same volume of
homogeneous tissue-equivalent medium will not yield conservative exposure. A detailed
discussion and quantification of this effect can be found in [70], [71].
J.3 Simulations
The project team investigated the following in the existing literature with respect to age, sex,
and race:
• statistical distribution of epidermis plus dermis thickness;

• statistical distribution of the fat layer thickness;
• thickness given by mean and standard deviation.
Skin anatomy and physiology undergo modifications throughout each person’s entire lifespan.
Diller [122] assumes that the skin thickness of children is 72 % of adult skin thickness.
Body-worn device positions according to a report of Carnegie Mellon University [123] were
chosen for the study. This report from Carnegie Mellon University identified regions of the
human body where radios can be carried without impeding the person’s movements. Dynamic
wearability is confined to body positions such as:
– arm – triceps;
– forearm;
– upper chest;
– back – subscapular;
– back – above pelvis;
– ankle;
– leg – calf;
– thigh – front;
– waist – lateral.
Body tissue thickness varies with individuals. The various thicknesses have a probability
distribution. The SAR for a given skin and fat tissue (t s , t f ) is weighted according to P(t s ) P(t f )
by the thickness probability.
Simulations were done for the different wearability areas with the data from the literature
study on the skin and fat thicknesses of the human body [124]. Antennas with electrical
lengths of 0,5 λ, 0,47 λ, 0,35 λ, 0,23 λ, 0,10 λ and 0,05 λ were analysed, at frequencies
between 30 MHz and 6 GHz.
The increased SAR also varied with the distance of the antenna from the body. This distance
depends on the frequency.
IEC/IEEE 62209-1528:2020 – 209 –
© IEC/IEEE 2020
J.4 Recommendation
The results of all simulations for the different thicknesses of skin, different thicknesses of fat
layers, different dielectric properties, different antennas, and different distances from the body
were analysed. From all the simulations, adopting a statistical coverage factor protective of
90 % of the population (Figure J.2), skin enhancement factors can be derived as shown in
Table J.1 and Figure J.3.
The project team asked ICNIRP and ICES for advice whether this factor is necessary for the
evaluation of the exposure. ICES came to a very clear consensus.
• It is not necessary to apply a scaling factor to limit the skin SAR.

• The measurement of SAR in a homogenous phantom without a scaling factor is adequate
to protect the public.
ICNIRP responded that this issue will be considered in the general revision of their RF
guidelines.
In order to quantify the possible SAR underestimation and to specify the conditions under
which this underestimation will occur, Annex J is included as an informative part of this
document.
Figure J.2 – Statistical approach to protect 90 % of the population
Table J.1 –psSAR correction factors
a,b
Frequency band Device to phantom distance Skin enhancement factor
300 MHz to 800 MHz 0 mm to 200 mm 1,0
800 MHz to 1 000 MHz 0 mm to 40 mm 1,0
40 mm to 45 mm linearly interpolate between 1,0 and 1,1
45 mm to 200 mm 1,1
1 700 MHz to 3 000 MHz 0 mm to 20 mm 1,0
20 mm to 35 mm linearly interpolate between 1,0 and 1,5
35 mm to 200 mm 1,5
5 000 MHz to 6 000 MHz 0 mm to 200 mm 1,0
a This represents the enhancement of the psSAR (1 g and 10 g). It is referred to as the skin enhancement
factor because the SAR enhancement is in the skin layer.
b For all other frequencies and distances, use linear interpolation.
– 210 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure J.3 – psSAR skin enhancement factors

IEC/IEEE 62209-1528:2020 – 211 –
© IEC/IEEE 2020
Annex K
(normative)
Application-specific phantoms
K.1 General
Some wireless communication devices (e.g. smart glasses, hearing aids, smart watches) have
form factors and/or usage scenarios such that assessments performed with the standard
phantoms of Clause 6 (sagittally-bisected SAM, and the elliptical or rectangular flat phantoms)
are either not applicable or are inappropriate. For example, some wireless communication
devices expose the top of the head, the front of the face, or specific body parts, e.g. wrist,
hands, or ankles. Specific phantoms are proposed that allow the exposed region of the body
to be evaluated in accordance with the intended use of the device, and reduced measurement
uncertainty. Annex K describes the requirements for appropriate phantoms, scanning
measurements, and reporting.
K.2 Phantom basic requirements
The phantom and procedure shall enable conservative exposure. The shape and tissue-
equivalent medium shall be selected such that the assessed psSAR is higher than the 90th
percentile of the highest exposure in the user population. This is usually derived by numerical
parameter variations using anatomical models that represent the user population [73].
The phantom shell shall ensure stable shape within 2 % relative to the provided CAD file of
the phantom, even when filled with tissue-equivalent medium and with a mounted device
under test (DUT).
Justification for the appropriateness of phantom and tissue-equivalent medium for the
exposure conditions, and evidence that the test system (phantom and tissue-equivalent
medium) is conservative, shall be provided. These items are typically demonstrated, using
validation studies and other results, by the equipment manufacturer, including the SAR
measurement set-up; for example, test device and phantom positioning, field probe
positioning and scanning, or post-processing requirements, etc.
K.3 Examples of specific alternative phantoms
K.3.1 Face-down SAM phantom
The SAM face-down phantom shown in Figure K.1 is recommended for devices where the
exposure is at the front side of the head. It is the same as the SAM phantom specified in
Annex G, but oriented with the front of the face facing downward. This phantom is truncated
along a plane behind the ear reference point. Above this plane, an upper extension is added
to ensure that the tissue-equivalent medium liquid is deep enough to measure in the relevant
regions of the SAM phantom. The upper extension is flanged, to allow measurement probe
access. The shell thickness shall be 2 mm ± 0,2 mm (not including the upper extension). The
phantom is filled with head tissue-equivalent medium having dielectric properties as specified
in 6.2. The same phantom could be used for some transmitters mounted on eyeglasses or
eyewear. If the psSAR location is in the eye region, the user shall either show that the tissue-
equivalent medium provides conservative exposure or propose an alternative medium. The
conservativeness of the assessment shall be determined by numerical evaluation, or with an
equivalent method (e.g. with the Virtual Family phantoms).
– 212 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Figure K.1 – SAM face-down phantom
K.3.2 Head-stand SAM phantom
The head-stand phantom (Figure K.2) is recommended for devices where the exposure is at
the top side of the head (e.g. head-mounted devices). It is the same as the SAM phantom
specified in Annex G, but oriented with the top of the head facing downward. It is truncated
along a plane above the bottom of the ear reference point. Above this plane, an upper
extension is added to ensure that the tissue-equivalent medium liquid is deep enough to
measure in the relevant regions of the SAM phantom. The upper extension is flanged, to allow
better measurement probe access for the top of the head (bottom of the head-stand
phantom). The shell thickness shall be 2 mm ± 0,2 mm (not including the upper extension).
The phantom is filled with tissue-equivalent medium having dielectric properties as specified
in 6.2.
Figure K.2 – SAM head-stand phantom
K.3.3 Wrist phantom
The wrist phantom (Figure K.3), to test wrist-worn wearables (watches, bracelets, etc.) that
are equipped with transmitters, is based on CTIA forearm phantom with a shell thickness of
2 mm. The conservativeness of the tissue-equivalent medium shall be demonstrated. The
wrist phantom is recommended when the flat phantom is inappropriate (e.g. when wrist-worn
wearables do not conform to the flat phantom, when test distances or gaps between the
device and the phantom are inconsistent with intended use conditions, or when wrist band
straps operate correctly only when closed).
IEC/IEEE 62209-1528:2020 – 213 –
© IEC/IEEE 2020
Figure K.3 – Wrist phantom
K.4 Scanning and evaluation requirements
For application-specific phantoms, the measurement requirements of Clause 7 shall be

applied where possible. If the geometry of the specific phantom does not allow the
requirements to be applied (e.g. due to probe orientation, scan volume), the details of the
assessment method shall be documented, and associated additional uncertainty analysis shall
be reported.
If the probe orientation can be maintained within ±30° of the normal to the phantom surface
(see 7.4), the positioning accuracy of the probe geometrical centre (centre of the probe
detector) relative to the phantom shell shall be better than ±0,2 mm, as specified in 6.4.
However, if it is not possible to maintain the probe orientation within the ±30° tolerance, a
larger positioning tolerance of up to ±0,5 mm is acceptable, if the additional uncertainty
analysis is provided. In this case, the boundary effect compensation is expected to have a
higher uncertainty, and the full hemispherical isotropy shall be evaluated (Clause E.4); the
method for extrapolation to the surface shall be assessed for all angles of incidence relative
to the probe orientation, and uncertainty assessed and documented (Clause K.5 and
Clause K.6). The scan area shall be larger than the projection of the device on the phantom.
If the SAR averaging volume includes portions outside the tissue-equivalent medium, the
psSAR evaluation shall be performed in compliance with methods from the numerical
standard IEC/IEEE 62704-1 [41], and the same uncertainty as in IEC/IEEE 62704-1 shall be
applied.
K.5 Uncertainty assessment
All uncertainty parameters specified in Clause 8 shall be determined for assessment using
application-specific phantoms. Additional uncertainties related to the specific phantom and
associated measurement procedures shall also be addressed. When an alternative phantom
is used, it is the user’s responsibility to ensure proper validation and justification is available,
including clear documentation and acceptable uncertainty assessment.
K.6 Reporting
In addition to the reporting requirements of Clause 9, the report shall include the justification
of the choice of phantom used, and the uncertainty assessment of the evaluation, especially
variations from Clause 8.
– 214 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex L
(normative)
Fast compliance evaluations using a flat-bottom

phantom with a curved corner (Uniphantom)
L.1 General
In Annex L, the main focus is methods to reduce the number of tests for evaluating the
compliance of a DUT by using a flat-bottom phantom with a curved edge, known as a
“Uniphantom,” and specified in Clause L.2. The techniques utilize a conventional SAR
measurement system, which complies with the basic requirements of this document, apart
from the shape of the phantom. Two handset form factors, straight and clamshell, may be
tested, as specified in Clause L.3 and Clause L.4.
L.2 Uniphantom
The Uniphantom is a flat-bottom container based on the rectangular flat phantom used for
body-worn devices (Clause G.3), and has a curvature that is representative of the SAM
phantom (Clause G.2) at a bottom edge, as shown in Figure L.1. The curvature cross-section
along the edge is a quarter elliptic arc with horizontal axis dimension of 90 mm and vertical
axis dimension of 75 mm. The ear reference point (ERP) is set at 135 mm from the adjacent
inner sidewall of the phantom. The curvature of this phantom approximately follows the line
along the ERP and the cheek-chin curve of the cross section of the SAM phantom (see
Figure G.6). The Uniphantom shall satisfy the specifications on dielectric properties of the
material and the mechanical requirements are the same as for the standard phantoms. The
thickness and width requirements of the flat bottom, curved edge, and sidewalls of the
Uniphantom are the same as the flat phantom for body-worn devices.
Figure L.1 – Cross section of the unified phantom (Uniphantom) with its dimensions
L.3 Device positions for compliance testing and definitions of handset shapes
L.3.1 General
Different device positions apply depending on the form factor of the handset. Subclauses
L.3.2 and L.3.3 describe devices with straight and clamshell form factors.
IEC/IEEE 62209-1528:2020 – 215 –
© IEC/IEEE 2020
L.3.2 Handsets with a straight form factor
Typical handsets categorized under this type include slider phones, bar phones, and smart
phones. With slider phones the keyboard can slide out. A phone of this category is placed at
the flat-bottom position or at the curved-edge position as shown in Figure L.2 a). The centre
of the speaker is at the ERP, same as in case of the full-compliance SAR measurement set-
up using the SAM phantom.
L.3.3 Handsets with a clamshell form factor
These handsets consist of two or more sections, and can be folded via one or more hinges. A
phone of this category is set at the curved-edge position as shown in Figure L.2 b). The
centre of the speaker is at ERP, same as in the case of the full-compliance SAR
measurement set-up using the SAM phantom.
a) Straight form factor b) Clamshell form factor
Figure L.2 – Measurement positions of handsets with

straight and clamshell form factors
L.4 Testing procedure
L.4.1 General
When comparing the measured SAR with the SAR limit, a margin shall be considered, to
prevent underestimation of the SAR measured with the Uniphantom. This margin shall be
appropriately determined based on statistical data. Examples of the statistical determination
of the margin are specified in Annex U.
L.4.2 Handsets with straight form factors
Figure L.3 shows a flowchart of the testing procedure for handsets with straight form factors.
In detail, the steps are as follows:
a) Step 1: Go to step 5 if measurement at the flat bottom of the Uniphantom will be omitted.
b) Step 2: Set the DUT at the flat bottom of the Uniphantom, with the front side in direct
contact with the phantom shell, as shown in Figure L.2 a).
c) Step 3: Measure psSAR (SAR flat ).
d) Step 4: If SAR flat is lower than the compensated SAR limit (SAR limit −
SAR margin_straight_flat ), the phone is deemed to be compliant and the evaluation may stop;
otherwise go to step 5.
e) Step 5: Set the phone at the curved-edge (cross-section corner) of the Uniphantom, with
the front side in direct contact with the phantom shell, as shown in Figure L.2 a).
f) Step 6: Measure psSAR (SAR corner ).
– 216 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
g) Step 7: If SAR corner is smaller than the compensated SAR limit (SAR limit −
SAR margin_straight_corner ), the phone is deemed to be compliant and the evaluation may
stop; otherwise go to step 8.
h) Step 8: Perform full-compliance psSAR evaluation using the SAM phantom.
Figure L.3 – Flow chart of testing procedure for handsets with straight form factors
L.4.3 Handsets with clamshell form factors
Figure L.4 shows a flowchart of the testing procedure for handsets with clamshell form
factors. In detail, the steps are as follows:
a) Step 1: Set the phone at the curved-edge (cross-section corner) of the Uniphantom, with
the front side in direct contact with the phantom shell, as shown in Figure L.2 b).
b) Step 2: Measure psSAR (SAR corner ).
c) Step 3: If SAR corner is lower than the compensated SAR limit (SAR limit +
SAR margin_clamshell_corner ), the phone is deemed to be compliant and the evaluation may
stop; otherwise go to step 4.
d) Step 4: Perform full-compliance psSAR evaluation using the SAM phantom.
IEC/IEEE 62209-1528:2020 – 217 –
© IEC/IEEE 2020
Figure L.4 – Flow chart of testing procedure for handsets with clamshell form factors
L.5 Uncertainty of SAR measurement results using Uniphantom
The uncertainty evaluation of SAR measurement results using the Uniphantom can be
performed by the methodology specified in Clause 8. When the measurement is carried out
only at the flat bottom of the Uniphantom, and the same device holder as that in the full-
compliance SAR evaluation is used, the same uncertainty table as in Clause 8 can be applied.
Notwithstanding the uncertainty evaluation scheme for specified in Clause 8, for compliance
evaluation using the Uniphantom, several Type A uncertainty items may be evaluated
separately.
The uncertainties related to the items listed below are identical to those of the full-compliance
SAR measurement specified in Clause 8:
• mechanical restrictions of the probe positioner;

• phantom shape and thickness;
• postprocessing;
• test sample positioning;
• device holder.
– 218 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex M
(informative)
Wired hands-free headset testing
M.1 Concept
Annex M describes a procedure that is suitable for assessing the SAR in the head from a
wired personal hands-free headset. Note that other procedures following the same principles
can also be suitable.
• An experimental test set-up designed in accordance with the principles set out in this
document shall be used.
• For correct comparison, the configurations shown in Figure M.1 and Figure M.2 shall be
used.
Annex M only describes how to make an informed assessment of hands-free kits.
Figure M.1 – Configuration of a personal wired hands-free headset

IEC/IEEE 62209-1528:2020 – 219 –
© IEC/IEEE 2020
Figure M.2 – Configuration without a personal wired hands-free headset
The tissue-equivalent medium at the appropriate operating frequency should be used, as

specified in this document.
The two conditions to be tested are:
– the head region of the head-and-torso phantom is filled with the tissue-equivalent medium;
– the entire head-and-torso phantom is filled with tissue-equivalent medium.
In both set-ups, the DUT should be placed next to the ear, and then at the waist with the
headset connected to the phone via lead-wires and taped at the ear. Scanning shall be
focused on finding SAR changes close to the ear. The psSAR in the head shall also be
measured.
Human phantoms used for measurements involving a hands-free accessory shall include the
torso; i.e. measurements shall not be performed on the head phantom alone. Use of only a
head phantom alone can have significant (and anomalous) impact on the results, because the
RF energy coupled into the leads of hands-free accessories is strongly attenuated by the
body.
M.2 Example results
The example results are given for the phantom partially filled with tissue-equivalent medium,
and the DUT speaker centred at the phantom ear. The maximum 1 g psSAR of 1,2 W/kg is
located below the ear. The maximum 1 g psSAR at the ear position is about 0,4 W/kg. There
is no significant change in the results of this measurement if the torso is also filled with tissue-
equivalent medium, because the DUT interacts mainly with the head.
When the headset accessory is used and the phone is worn at the chest, the following results
have been obtained. The headset earpiece is centred at the phantom ear canal and the DUT
at the belt (Figure M.1). In this case (with the unfilled torso), the maximum 1 g psSAR
measured at the ear (32 mm × 32 mm) is 0,05 W/kg, showing an 87 % (9 dB) reduction
compared to the with the DUT at the ear position. If the phantom chest is completely filled
with tissue-equivalent medium, a maximum 1 g psSAR of 0,02 W/kg near the ear is obtained,
which corresponds to a 95 % (13 dB) reduction from the case with the DUT at the ear.
– 220 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
M.3 Discussion
From these example measurements it is clear that: 1) the SAR near the ear where the
headset earpiece is located is low compared to the peak SAR produced by the DUT; 2) the
whole-phantom maximum psSAR is not located near the ear with the headset accessory; 3)
the presence of the torso attenuates the field from the headset accessory, resulting in a lower
SAR compared to that measured without the torso. It should be noted that in the preceding
situations, the coupling between the DUT and headset accessory was very strong, as shown
by the additional measurement of the whole-phantom maximum SAR in the head from the
headset without the presence of the torso. These measurements yield a peak 1 g psSAR of
0,9 W/kg. Again, it should be noted that the location of this peak is not near the headset
earpiece, but it is shifted down to the cheek where the corresponding current on the headset
cable is stronger. This is close to the results of computations with best coupling between the
wire and RF source, which indicates that the presence of the torso leads to a strong decrease
of the whole-phantom maximum SAR in the head, even when the wire is strongly coupled to
the RF source (DUT).
The measurements also show that the SAR from the headset accessories near the ear is
always lower than when the phone is near the head. It should be noted that if the primary RF
source is removed from the ear, the secondary radiator (headset) produces a much lower
exposure at this location.
Though the attenuation effect of the body can vary depending on the position of the wire
relative to the body, if a large part of the hands-free accessory wire is close to the user’s body
then the psSAR measured in the ear region can be more than 10 dB lower than psSAR
measured from the DUT alone when placed near the head.
IEC/IEEE 62209-1528:2020 – 221 –
© IEC/IEEE 2020
Annex N
(informative)
Applying the head SAR test procedures
Annex N illustrates an example of how to apply the SAR test protocol of 7.3. A DUT
supporting GSM mode in 850 MHz, 900 MHz, 1 800 MHz, and 1 900 MHz bands is used for
illustration. The device has a sliding keypad that can be used for voice calls with the slide in
open and closed positions. The test procedures in 7.4.2 are applied according to the following
steps.
Step 1: SAR measurements are performed at the channel having the highest output power for
each transmit frequency band according to procedures in 7.4. The corresponding channel
numbers are 190 at 836,6 MHz for GSM 850, 38 at 897,6 MHz for GSM 900, 699 at 1
747,6 MHz for GSM 1800, and 661 at 1 880 MHz for GSM 1900. All device test configurations
specified in 7.2.4.2.2 and 7.2.4.2.3 for cheek and tilt positions on the left and right sides of the
SAM phantom are tested, with the sliding keypad open and closed. The device supports only
GSM voice mode for head use; therefore other GSM modes are not applicable for this
example.
Step 2: SAR measurements are performed for additional frequency channels in each
frequency band and mode as required by 7.2.8 for the highest psSAR test configuration of
step 1. In addition, SAR is also measured at other frequency channels as required by 7.2.8 for
all configurations tested per step 1 where the psSAR is greater than or equal to half of the
applicable SAR limit.
Step 3: The SAR measurement results are shown in Table N.1, Table N.2, Table N.3, and
Table N.4. The highest 10 g psSAR among all GSM frequency bands is 1,205 W/kg, on
channel 251 at 848,8 MHz for GSM 850.
Table N.1 – SAR results tables for example test results in GSM 850 band
10 g psSAR [W/kg] a
Device
Mode and
configuration SAM phantom test position Ch 128 b Ch 190 b Ch 251 b
Band
(slide position)
824,2 MHz 836,6 MHz 848,8 MHz
Cheek 0,776 0,653 c 0,552
Left
Tilt - 0,492 -
closed
Cheek - 0,626 -
Right
Tilt - 0,448 -
GSM
850
Cheek 1,011 1,192 d,e 1,195
Left
Tilt - 0,430 -
open
Cheek 0,892 1,120 e 1,205 f
Right
Tilt - 0,418 -
a The applicable regulatory SAR limit is 2,0 W/kg averaged over a 10 g tissue volume.
b According to 7.2.8, the number of channels to be tested is N c = 3 for all of the modes (see step 2).
c This is the highest psSAR measured for the mid-band channel with slide closed. Hence, the other two
channels in this frequency band also require SAR measurement (see step 2).
d This is the highest psSAR measured for the mid-band channel with slide open. Hence, the other two channels
in this frequency band also require SAR measurement (see step 2).
e The measured psSAR is within 3 dB of the applicable SAR limit. Hence the other two channels in this
frequency band also require SAR measurement (see step 2).
f This is the highest psSAR measured among all GSM frequency bands and test configurations (see step 3).
– 222 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
SAR, averaged over 10 g [W/kg] a

Mode & Device
Device position Ch 975 b Ch 38 b Ch 124 b
Band configuration
880,2 MHz 897,6 MHz 914,8 MHz
Cheek 0,766 0,730 c 0,652
Left
Tilt - 0,522 -
closed
Cheek - 0,631 -
Right
Tilt - 0,482 -
GSM
900
Cheek 0,618 0,723 d 0,833
Left
Tilt - 0,443 -
open
Cheek - 0,620 -
Right
Tilt - 0,406 -

Mode & Device
Band configuration
1 710,2 MHz 1 747,6 MHz 1 784,8 MHz
Cheek - 0,254 -
Left
Tilt - 0,230 -
closed
Cheek 0,336 0,343 c 0,466
Right
Tilt - 0,215 -
GSM
1800
Cheek - 0,154 -
Left
Tilt - 0,162 -
open
Cheek - 0,174 -
Right
Tilt 0,188 0,192 d 0,211
IEC/IEEE 62209-1528:2020 – 223 –
© IEC/IEEE 2020

Mode & Device
Band configuration
1 850,2 MHz 1 880,0 MHz 1 909,8 MHz
Cheek - 0,240 -
Left
Tilt - 0,233 -
closed
Cheek 0,246 0,353 c 0,387
Right
Tilt - 0,221 -
GSM
1900
Cheek - 0,158 -
Left
Tilt - 0,178 -
open
Cheek - 0,169 -
Right
Tilt 0,188 0,195 d 0,301
– 224 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex O
(normative)
Uncertainty analysis for measurement system

manufacturers and calibration laboratories
O.1 Probe linearity and detection limits
Because the non-linear response of the probe does not depend on the surrounding media but
only on the sensor diodes, evaluation of the linearizing functions f i (V i ) for a CW signal may be
done either in free space or in the tissue-equivalent medium (for modulated signals, use the
procedure specified in Clause R.3). This is performed by a power sweep covering the
specified psSAR detection range, or a range at least from 0,12 W/kg to 100 W/kg in steps of
3 dB or less. Small TEM cells or waveguides as field generators are preferable, because high
field strengths can be produced with medium power amplifiers. The linearity uncertainty is
specified as the maximum deviation in the SAR versus power characteristics from the best-fit
straight reference line going through zero (SAR = a ∙ P, where a is the best-fit slope of the line,
using the method of least-squares) specified over the interval 0,12 W/kg to 100 W/kg. The
linearity uncertainty for CW signal is determined in Formula (O.1):
 SAR ( Pi )CW, j =100W/kg  SAR( Pi )CW, j 

SARCW linearity
= uncertainty [%] max  max  100 × − 1 
 (O.1)
j∈{ x, y , z}  SAR ( Pi )CW, j = 0,12W/kg  Pi ⋅ a 

Here, SAR(P i ) CW,j is the SAR measured for a CW signal at the i-th power level, P i . The index j
refers to the field sensor for each of which the procedure shall be repeated. A rectangular
probability distribution is assumed for the probe linearity uncertainty.
At SAR levels as high as 100 W/kg, the tissue-equivalent medium temperature can rise
significantly, causing deviations in the dielectric properties. To avoid this problem, care
should be taken to ensure that the tissue-equivalent medium is exposed to high SAR levels
only for short periods of time.
The probe detection limit uncertainty is determined as in Formula (O.2) by comparing the
measured psSAR (SAR meas ) at a reference power level P ref to the calculated SAR, SAR ref =
a ∙ P ref , using the best-fit slope as in the preceding paragraph.
SARmeas SARmeas
U detection limit [%] =
100 × 100 ×
= (O.2)
SARref a ⋅ Pref
The reference power level P ref shall be chosen such that the signal to noise ratio (determined
at the measurement time) is 6 dB. This shall be verified using the data from the power sweep.
A rectangular probability distribution shall be assumed for detection limit uncertainty in
Table 9.
The uncertainty posed by the non-linear response can be rather high and depends on the
various probe components, i.e. it shall be determined for each probe individually. A
rectangular probability distribution has been assumed for probe linearity uncertainty in
Table 9.
IEC/IEEE 62209-1528:2020 – 225 –
© IEC/IEEE 2020
O.2 Broadband signal uncertainty
Electric field-strength probes used for SAR measurements are electrically small dipoles that
do not have a constant frequency response (Annex E). In addition, the probes are immersed
in a tissue-equivalent medium with frequency dependent dielectric properties. This frequency
dependence can be neglected when measuring narrowband signals. If the signal bandwidth
exceeds 100 MHz, an additional uncertainty term shall be included in the uncertainty budget.
As a first approximation, the frequency dependence of the probe sensitivity S f can be

assumed to be linear with frequency f. The uncertainty can be estimated using Formula (O.3):
S f +1 − S f −1 BW
SARuncertainty [ % ] =
kw × × (O.3)
Sf f +1 − f −1
where
BW is the bandwidth of the signal being measured (in hertz);

f +1 is the closest calibrated frequency that is above the centre frequency of the signal being
measured (in hertz);
f −1 is the closest calibrated frequency that is below the centre frequency of the signal being
measured (in hertz);
f +1 − f −1 > BW;
S f+1 is the sensitivity of the sensor at frequency f +1 ;
S f−1 is the sensitivity of the sensor at frequency f −1 ;
Sf is the sensitivity of the sensor at the centre frequency of the signal being measured;
kw is a weighting factor of the spectral non-symmetry of the signal time average envelope
with respect to the centre frequency. A conservative value shall be applied that is
determined by measurements.
This uncertainty is assumed to have a rectangular distribution.
O.3 Boundary effect
The probe boundary effect introduces measurement uncertainty. For the purposes of this
document, this uncertainty is negligible if the closest distance between the probe tip and the
phantom inner surface is always larger than the probe-tip diameter.
In some cases, measurements at distances closer than the probe tip diameter can be
required, in order to reduce interpolation and extrapolation uncertainties. Then the boundary
effect uncertainty of Clause E.6 shall be assessed preferably by using the waveguide system
specified in E.3.2.3.4. Alternatively, the temperature method specified in E.3.2.3.2 may be
used below 800 MHz. The following method is applicable assuming that the angle between
the probe axis and the surface normal line is within the requirements of Table 3 and Table 4.
Because the boundary effect is a characteristic of a specific probe, it shall be determined
during probe calibration (i.e. influence of the probe tip diameter). If algorithms are applied to
compensate for the boundary effect, then the SAR uncertainty shall be determined with the
same evaluation hardware and software that is used for SAR measurements. The boundary
effect uncertainty can be estimated according to Formula (O.4) and Formula (O.5), as based
on linear and exponential extrapolations between the surface and d be + d step along lines that
are approximately normal to the surface.
– 226 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
=SARuncertainty [%] ΔSARbe [%]

( dbe + dstep )
2
(e − dbe ( δ 2)
); ( dbe + dstep ) < 10 mm, f ≤ 3 GHz (O.4)
2d step δ 2
δ
SARuncertainty [%] ΔSARbe [%]
= ; dbe < δ , f > 3GHz (O.5)
δ − dbe
where
SAR uncertainty is the uncertainty, as a percentage, of the probe boundary effect;

d be is the distance between the surface and the closest measurement point;
d step is the separation distance between the first and second measurement points
that are closest to the phantom surface, assuming that the boundary effect at
the second location is negligible;
δ is the penetration depth of the tissue-equivalent medium;
ΔSAR be is the deviation of the measured SAR value, as a percentage, at the distance
d be from the boundary, from the analytical reference SAR value for waveguide
systems SAR ref ; SAR ref is calculated using Formula (E.5) (see E.3.2.3.4).
If waveguide systems are not available for certain frequency ranges, temperature probes shall
be used to assess the reference values SAR ref at the locations d be and (d be + d step ), and the
SAR uncertainty of the temperature probe shall be accounted for. If temperature methods are
used, then SAR ref is the value at this location determined using the temperature probe. Note
that the actual calibration shall be performed at distances between the probe tip and boundary
that are larger than the probe tip diameter, where the boundary effect is negligibly small.
Enter the uncertainty of the probe boundary effect in the appropriate row and column in
Table 9 using rectangular distribution.
O.4 Field-probe readout electronics uncertainty
All uncertainties related to the probe readout electronics, including the gain and linearity of
the instrumentation amplifier, its loading effect on the probe, and accuracy of the signal
conversion algorithm, shall be evaluated to estimate the maximum SAR uncertainty. One
method to determine these uncertainty components is by replacing the probe with an
equivalent source having the same source impedance as the probe under consideration,
according to the manufacturer’s specifications for the probe. This evaluation is generally
performed by the system manufacturer. Each uncertainty shall be converted to a standard
uncertainty using normal probability distribution. The RSS value of these uncertainties shall
then be used to determine the overall readout electronics uncertainty.
O.5 Signal step-response time uncertainty
Field-probe signal step-response time uncertainty is evaluated by exposing the probe to an

E-field step response producing at least 100 W/kg near the phantom surface. The signal
response time is evaluated as the time required by the probe and its readout electronics to
reach 90 % of the expected final value produced by the step response by switching the RF
power on and off. During the SAR measurement, the probe shall remain stationary at each
measurement location for at least twice the assessed response time, so that the probe signal
response time uncertainty is negligible. Under these measurement conditions, a tolerance
value of zero may be entered in Table 9 Otherwise, the SAR uncertainty due to signal
response-time uncertainty shall be assessed, using the signal characteristics of the test
device. In this case, the signal step-response time uncertainty is equal to the percentage
difference between the SAR measured at the chosen measurement time and the SAR
measured at twice the chosen measurement time. A rectangular probability distribution shall
be assumed.
IEC/IEEE 62209-1528:2020 – 227 –
© IEC/IEEE 2020
O.6 Probe integration-time uncertainty
O.6.1 General
Probe integration-time uncertainties can arise when test devices do not emit a continuous
signal. When the integration time and discrete sampling intervals used in the probe
electronics are not synchronized with the pulsed characteristics of the measured signal, the
RF energy at each measurement location might not be fully or correctly captured. This
uncertainty shall be evaluated according to the signal characteristics of the test device prior to
the SAR measurement.
O.6.2 Probe integration-time uncertainty for periodic pulsed signals
For signals with periodic pulse modulation and a pulse period greater than 1 % of the probe
integration time, additional SAR uncertainty shall be considered when the probe integration
time is not an exact multiple of the longest periodicity. The uncertainty shall be assessed
according to the maximum uncertainty expected for unsynchronized probe integration time,
with an assumed rectangular probability distribution. For a signal with an envelope s(t), the
average signal read by the probe during the integration time t int starting at time t 0 is given by
s int (t 0 ,t int ) in Formula (O.6):
t0 +tint
1
sint (t0 , tint )
=
t int ∫ s(t ) dt, 0 ≤ t0 ≤ T (O.6)
t0
Formula (O.6) assumes that the filtering by the probe does not significantly alter the signal
envelope s(t). If t 0 is not synchronized with the longest period T of s(t), the probe integration-
time uncertainty can be specified as shown in Formula (O.7):
max ( sint ( t ) ) − sint (t )

SARuncertainty [%
=] 100 × (O.7)
sint (t )
Here ⋅ denotes the average value. Formula (O.6) and Formula (O.7) can be used to derive
the probe integration-time uncertainty of any pulsed signal.
A simple alternative formula for the uncertainty of a TDMA system is shown in Formula (O.8):
tframe slotidle
SAR uncertainty [=
%] 100 × ∑ t slottotal
(O.8)
all sub-frames int
Formula (O.8) is an approximation that typically overestimates the uncertainty. Here slot idle is
the number of idle slots in a frame, with slot total being the total number of slots. The frame
duration is t frame , with t frame < t int . The total probe integration-time uncertainty is the sum of
the errors for all subframes in the frame structure that have idle slots. For example, the basic
frame for GSM systems has a frame duration t frame = 4,6 ms, with 7 idle slots in an 8-slot
frame, and the multiframe duration is t multiframe = 120 ms, i.e. 1 idle slot in a 26-slot frame.
For a probe integration time of 0,2 s, the uncertainty is estimated to be 0,020 1 + 0,023 1 =
0,043 2, or 4,32 % for GSM using Formula (O.8), compared with 3,84 % using Formula (O.6)
and Formula (O.7). GPRS is the same as GSM, except that the number of idle slots can be 6,
5, …, 1, where 7 idle slots is the worst case.
A rectangular probability distribution shall be assumed for probe integration-time uncertainty.

For continuous or CW-equivalent signals, an uncertainty value of zero shall be entered.
– 228 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
O.6.3 Probe integration-time uncertainty for non-periodic signals
For other than periodic pulsed signals, the probe integration time should be determined from
SAR measurements using a stable source with the same signal characteristics and the same
probe type that is used for DUT measurements. Measurements at a single point (where the
SAR is at least 1 W/kg) shall be made consecutively using the chosen integration time and
successively larger integration times. The integration time shall be doubled for each
consecutive measurement, until the last and next to last measurement results are within 0,5 %
of each other. Each measurement at a given integration time shall be repeated several times
to verify that the measurement result is stable, and the average SAR of the repeated
measurements shall be used for that integration time. For the probe integration time, a
rectangular probability distribution shall be assumed. The uncertainty is the percentage
difference between the average SAR at the selected integration time and the average SAR at
the longest integration time.
O.7 Contribution of mechanical constraints
O.7.1 Mechanical tolerances of the probe positioner (directions parallel to phantom

surface)
The mechanical tolerances of the field probe positioner can introduce deviations in the
accuracy and repeatability of probe positioning, which add to the uncertainty of the measured
SAR. The uncertainty may be estimated from the specifications of the probe positioner
relative to the position required by the actual measurement location determined by the
geometrical centre of the field probe sensors, and is expressed as maximum deviation d ss . By
assuming a rectangular probability distribution, the psSAR uncertainty contributions due to
mechanical tolerances of the probe positioner may be calculated using d ss , according to a
first-order error approximation as described in Formula (O.9):
d ss
%
SARuncertainty = × 100 (O.9)
δ 2
where
SAR uncertainty is the uncertainty as a percentage;

d ss is the maximum position uncertainty between the calculated position at the
centre of the probe sensors and the actual position with respect to a
reference point specified by the system manufacturer;
δ is the penetration depth of the tissue-equivalent medium specified in this
document.
If the system manufacturer does not specify the mechanical tolerances of the probe
positioner, this shall be evaluated to determine the contribution to SAR measurement
uncertainty. This can be done by evaluating the relative accuracy of probe movements within
the area-scan region. The largest distance difference between the target and actual positions
shall be used to assess the SAR uncertainty. The SAR uncertainty shall be entered in Table 9
assuming rectangular distribution.
O.7.2 Probe positioning with respect to phantom shell surface
The uncertainty of the probe positioner relative to the phantom shell shall be estimated as the
maximum deviation of the distance between the probe tip and the phantom surface d ph . By
assuming a rectangular probability distribution, the psSAR uncertainty contribution is
calculated using Formula (O.10):
dph
SARuncertainty [=
%] × 100 (O.10)
δ/2
IEC/IEEE 62209-1528:2020 – 229 –
© IEC/IEEE 2020
where

d ph is the maximum deviation of the distance between probe tip and phantom
shell, i.e. the uncertainty of determining the probe tip location;
δ is the penetration depth of the head tissue-equivalent medium specified in
this document.
O.7.3 First-order approximation of exponential decay
Formula (O.9) and Formula (O.10) are first-order error approximations of the SAR error
corresponding to a positioning error ∆z at location z. They are derived according to
Formula (O.11):
−2α ( z +∆z )
− SAR0e −2α z
∆SAR
=
SAR
SAR0e
SAR0e −2α z
= (e −2α z
)
−1
(O.11)
1 2
= 1 − 2α ∆z + ( 2α ∆z ) +  + ( −1) ≈ −2α ∆z
2
Here α is the reciprocal of the field penetration depth δ, and SAR 0 is the SAR value at the
surface of the phantom (z = 0). ∆SAR/SAR is an estimate of the SAR sensitivity for the
exponential decay.
O.8 Contribution of post-processing
O.8.1 General
The estimation of the uncertainty resulting from the post-processing of the discrete measured
data to determine the 1 g and 10 g psSAR is described, i.e. the combined uncertainty of
interpolation, extrapolation, averaging, and maximum-finding algorithms. These algorithms
can add uncertainty due to assumptions on the general field behaviour and therefore might
not perfectly predict the SAR distribution in the tissue-equivalent medium for a specific DUT.
The algorithm uncertainty is a function of the resolution chosen for the measurement, and the
post-processing methods used in the area scans and zoom scan.
The actual SAR distribution at the peak location is strongly dependent on the operating
frequency and design of the DUT, test position, and proximity to the tissue-equivalent
medium. SAR distributions can have a rather flat gradient when a low frequency source is a
large distance away, or can have a very steep gradient when a small high frequency source
(such as a helix antenna) is placed next to the tissue. In some cases, the maximum SAR is
offset into the medium rather than at the surface of the phantom, due to cancellation of
magnetic fields at the surface [64].
The analytical SAR distribution functions presented in O.8.2 are intended to simulate the
preceding conditions, and were developed for the purpose of this uncertainty estimation.
These empirically derived reference functions are used to create artificial or “dummy” SAR
data sets, for testing the system software post-processing subroutines. Computed reference
function values at coarse and fine grid spacing, the same as those used in measurements,
are input to the SAR system software. SAR values at grid points corresponding to the area-
scan and zoom-scan measurement grids are computed according to the three SAR distribution
functions given in O.8.2, then processed by the system interpolation, extrapolation, and
integration algorithms as if they were actually measured. The resulting 1 g and 10 g psSAR
values are compared with reference SAR values listed in O.8.2. Procedures for evaluating the
SAR uncertainty of the area-scan and zoom-scan post-processing algorithms are specified in
O.8.3. The test functions assume a planar tissue-equivalent medium and phantom interface.
The applicability of these functions for curved interfaces is discussed in O.8.3.3. This
uncertainty concept assumes that there are no errors in location of the grid points calculated
– 230 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
with the analytical distribution functions, and probe positioning and measurement
uncertainties are not included.
O.8.2 Evaluation test functions
Three analytical functions, f 1 , f 2 , and f 3 , are used to represent the possible range of SAR
distributions expected for DUTs tested according to the procedures of this document. The
distribution f 1 in Formula (O.12) is based on the evaluation of SAR footprints of actual
wireless communication devices, and is applicable for frequencies up to 2 GHz. Because f 1
takes into account devices placed in close proximity to SAM phantom at frequencies above
900 MHz, f 1 is also used to model very steep SAR gradients. Two parameter sets are given
for f 1 , such that SAR distributions with single and double maxima can be evaluated.
The function f 2 in Formula (O.13) is used at frequencies up to 3 GHz to account for exposure
conditions with H-field cancellation at the phantom-tissue surface.
The third reference function f 3 in Formula (O.14) is specified for testing in the frequency range
3 GHz to 10 GHz. It is also valid for capacitively-coupled sources. Because noise can affect
the extrapolation at these frequencies, a noise term is included. The distribution functions are
specified for the phantom surface at z = 0, and the half-space tissue-equivalent medium for all
z > 0 in Formula (O.12):
2 z
 ( x′ + x / 2)
d

 exp  −
( y ′)2  − a
A1 exp  −
f1( x, y , z ) = 2
e
 2σ x , peak   2σ 2y , peak 
 
(O.12)
 ( x ′ − x / 2 )2   ( y ′ )2  − z
d
+ A2 exp  −  exp  − e a
 2σ x2,sec   2σ 2y ,sec 
 
z  2z
a2 − 
 3 − e a  cos2  π y ′ 
−
f 2 ( x, y , z ) A e
= a
2
(O.13)
a 2 + ( x ′)    2 3a 

A exp  −
f3 ( x, y , z ) =
( )
 ( x ′ )2 + ( y ′ )2 2  8 z
 e− a + N
2  rms rnd ( ζ ) (O.14)
 ( a 4 ) 
Note that due to the noise term in the function f 3 , it should be evaluated at least 4 000 times,
to verify that the standard deviation converges to the correct value. x, y, and z are the spatial
Cartesian coordinates of the SAR distribution. The SAR averaging cube shall be assumed to
be directly against the surface of a planar phantom. The z-axis is normal to the phantom
surface, and the axes of the averaging cube are aligned with the axes of the coordinate
system.
In Formula (O.12), the parameters σ x,peak , σ y,peak , σ x,sec , and σ y,sec are set as described in
Formula (O.15) through Formula (O.18):
σ x ,pp , x ′ ≥ − xd / 2
σ x ,peak =  (O.15)
σ x ,pn , x ′ < − xd / 2
IEC/IEEE 62209-1528:2020 – 231 –
© IEC/IEEE 2020
σ y ,pp , y ′ ≥ 0
σ y ,peak =  (O.16)
σ y ,pn , y ′ < 0
σ x ,sp , x ′ ≥ xd / 2
σ x ,sec =  (O.17)
σ x ,sn , x ′ < xd / 2
σ y ,sp , y ′ ≥ 0
σ y ,sec =  (O.18)
σ y ,sn , y ′ < 0
These parameters have been selected based on the evaluations of different DUTs at
1 950 MHz, and are given in Table O.1. The decay parameter a was derived by fitting
experimental zoom scan data at 1 950 MHz with an exponential decay curve, and by
averaging the results of different DUTs.
Table O.1 – Parameters for the reference function f 1 in Formula (O.12)
No. A1 A2 a xd σ x,pp σ y,pp σ x,sp σ y,sp σ x,pn σ y,pn σ x,sn σ y,sn

of
peaks [W/kg] [W/kg] [mm] [mm] [mm] [mm] [mm] [mm] [mm] [mm] [mm] [mm]
1 1,2 0,0 11,9 n.a. 19,6 15,5 n.a. n.a. 21,9 17,2 n.a. n.a.
2 1,2 1,0 11,9 60,47 22,6 19,7 19,4 19,6 22,0 15,5 17,9 24,2
n.a. = not applicable
In Formula (O.12), Formula (O.13), and Formula (O.14), the parameters are set as follows:
x′= x + d y ′= y + d
Where d is the offset parameter, a is 20 mm, and A is1 W/kg.
The preceding parameters a and A are used for the generation of the appropriate SAR
distributions. A value of d = 2,5 mm, for example, provides a lateral shift of the SAR
distribution, so that the peak location is not aligned with a measurement grid having a 5 mm
increment. This offset is used to test the software peak search subroutines and uncertainty.
N rms represents the variance of the system noise in SAR (in W/kg) in the tissue-equivalent
medium for A = 5. This parameter is system dependent, and corresponds to random SAR
values measured in the absence of an RF signal [rnd(ζ)]. Suitable functions are available in
typical math applications. The variable ζ is an arbitrary seed.
The reference SAR values from the distribution functions f 1 , f 2 , and f 3 for 1 g and 10 g
averaging cubes aligned with the (x, y, z) coordinate axes are given in Table O.2. These SAR
values were calculated with an accuracy of 0,01 %.
When function f 1 is considered, the maximum deviation from reference values obtained
considering the single-peak and two-peaks cases shall be used for post-processing
uncertainty computations.
These reference values are applied in O.8.3 for testing the post-processing algorithms used
by the area scans and zoom scans.
– 232 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table O.2 – Reference SAR values from the distribution functions f 1 , f 2 , and f 3
Function (condition) SAR 1 g, W/kg SAR 10 g, W/kg

f 1 (single peak) 0,791 0,494
f 1 (two peak, cube centred
0,796 0,503
on primary peak)
f 1 (two peak, cube centred
0,686 0,438
on secondary peak)
f2 1,796 1,375
f3 0,015 7 0,026 8
O.8.3 Data-processing algorithm uncertainty evaluations
O.8.3.1 Evaluation of the coarse area scan
The area-scan interpolation algorithms shall be capable of locating psSAR coordinates with an
accuracy of ±L z /2 mm or better, where L z is the side length of the zoom-scan volume. If this
precondition is satisfied, which is tested with the procedures of O.8.3.1, the evaluation of the
area scan does not contribute to the uncertainty budget.
The reference function values calculated at the usual area-scan grid points are input to the
system software. The interpolation algorithm treats these data points as if they were
measured to complete the area scan and determine the psSAR location (x eval , y eval ). This is
compared with the actual peak location determined by the analytical functions at (x ref , y ref ),
with an offset parameter d, as specified in O.8.2 for x′ and y′ for Formula (O.12),
Formula (O.13), and Formula (O.14). The subscripts "eval" and "ref" refer to evaluated and
reference, respectively. In other words, Formula (O.19) shall be satisfied:
xref − xeval ≤ Lz / 2, yref − yeval ≤ Lz / 2 (O.19)
The following procedure shall be used to assess the uncertainty of the interpolation
algorithms used in the area scan for determining the peak SAR location.
a) Choose the measurement resolution (Δx, Δy), and number of evaluation (corresponding to
measurement) points (N x , N y ). The centre of the area scan shall be set to (x 0 , y 0 ) = (0, 0).
b) SAR values are computed using the functions f 1 , f 2 , and f 3 at the evaluation grid points of
the area scan within the ranges of Formula (O.20) and Formula (O.21):
x0 − Δx × ( N x − 1) / 2 ≤ x ≤ x0 + Δx × ( N x − 1) / 2 , (O.20)
 ( )   (
y0 − Δy ×  N y − 1 / 2 ≤ y ≤ y0 + Δy ×  N y − 1 / 2 ,
 ) (O.21)
where N x and N y are odd integers. A value of z = 0 is assumed, because the peak location is
independent of z for these three functions.
The SAR values computed by the three distribution functions are interpolated by the SAR
measurement system with a spatial resolution of (Δx i , Δy i ), according to the interpolation
functions g i (x) and g i (y) used by the system to determine the psSAR location (x eval , y eval ).
If the measurement system does not allow SAR values to be imported to perform the
assessment, the same algorithm shall be implemented independently by other means, to
determine the interpolation and peak-search uncertainties.
IEC/IEEE 62209-1528:2020 – 233 –
© IEC/IEEE 2020
The psSAR location determined by the interpolation algorithms shall satisfy the requirements
of the inequalities in Formula (O.19). Otherwise, the data-processing and measurement
systems shall use a finer grid resolution and/or a larger number of interpolation points to
repeat the assessment starting at step b).
The centre of the area scan (x 0 , y 0 ) shall be shifted in 1 mm steps within the range
0 < x 0 ≤ Δx/2 and 0 < y 0 ≤ Δy/2 to repeat the assessment starting at step b) for each of the
shifted (x 0 , y 0 ) in these ranges.
O.8.3.2 Evaluation of the zoom scan
The zoom scan is evaluated by comparing the highest 1 g or 10 g SAR values with the
reference SAR values in O.8.2. From the area scan procedure in O.8.3.1, the true peak
location (x ref , y ref ) will be displaced from the estimated peak location (x eval , y eval ) by an
amount given by inequalities in Formula (O.19). This displacement is accounted for in the
reference functions f 1 , f 2 , and f 3 in O.8.2 by incorporating the distance d. Because this
displacement will vary in practice, the value of d shall be varied over the range of
Formula (O.22):
d ≤ ( Lz − Lc ) / 2 (O.22)
where L c is the averaging cube side length (10 mm for 1 g, 21,5 mm for 10 g). For each
distance d, the largest uncertainty produced by any of the three functions is recorded.
The RMS of the largest uncertainty values for several distances d is entered as the
uncertainty due to extrapolation, interpolation, and integration.
Although the requirement for the area scan is that the local psSAR be located within |d| ≤
L z /2, a smaller range of |d| ≤ (L z − L c )/2 is used here to ensure that the 1 g or 10 g cube can
be computed on the first attempt. For values of (L z − L c )/2 and |d| ≤ L z /2, the measurement
software shall alert that the 1 g or 10 g cube is not captured and the measurement shall be re-
attempted. This will not affect the uncertainty, so it is not necessary to consider this case
here.
a) Choose a displacement d for the evaluation of the functions f 1 , f 2 , and f 3 . d shall vary from
−(L z − L c )/2 to +(L z − L c )/2 in small increments (e.g. 1 mm steps). The displacement shall
also vary separately in the x- and y-directions.
b) SAR values are computed according to the functions f 1 , f 2 , and f 3 at the evaluation grid
points that correspond to measured zoom-scan volume points. The zoom-scan volume
shall be centred at the point in Formula (O.23):
( x, y , z )
= (0, 0, ( Lh / 2 + dbe ) ) (O.23)
where
Lh is the height of the zoom-scan volume, and
d be is the distance of the closest measurement point from the inner surface of the
phantom.
c) The computed SAR values are extrapolated to the phantom surface at z = 0 by the system
software, to obtain the additional points in the zoom-scan volume that cannot be measured
due to probe constraints. Both the computed and extrapolated data points are then
interpolated to a finer resolution by the system software, which subsequently applies the
integration algorithms as well as the search algorithm for finding the psSAR within the
zoom-scan volume, to determine the highest 1 g or 10 g psSAR. Alternatively, other
procedures are possible. If the system does not allow SAR values to be imported to
perform the assessment, the same algorithm shall be implemented independently by other
means to test the extrapolation, interpolation, and integration algorithms.
– 234 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
d) The 1 g and 10 g psSAR values determined by the system or data processing software
(SAR eval ) are compared to the reference SAR values given in O.8.2. The standard
deviation caused by the random noise (SAR stdev (N rms )) is determined by evaluating f 3 at
least 4 000 times, and with each of the 4 000 or more evaluations using different random
noise parameters. The SAR uncertainty for distribution functions f 1 and f 2 is calculated
using Formula (O.24):
SAReval − SARref
SARuncertainty [ %
=] 100 × (O.24)
SARref
The SAR uncertainty for distribution function f 3 is calculated using Formula (O.25):
SAReval − SARref SARstdev ( Nrms )

SARuncertainty [ %] =
100 × + 100 3 × (O.25)
SARref SARref
e) The highest SAR uncertainty estimated by the three distribution functions is recorded.
f) Repeat steps b) to e) for other displacement values d.
g) Compute the RMS value of the uncertainties calculated in step d) for each displacement d
above. This value shall be entered as the uncertainty due to extrapolation, interpolation,
and integration in the corresponding row and column of Table 9, and a rectangular
probability distribution shall be used.
h) Record the following parameters used to estimate the zoom-scan uncertainty:
1) the dimension of the grid used to sample the reference functions, both in terms of
number of points and sample steps in the three dimensions;
2) the number of interpolation points included between two test points, or the
interpolation resolution in the three directions, for the reference functions;
3) the dimension of the extrapolation region, i.e. the distance between the probe sensor
location at the first measurement point and the phantom surface (also considering that
measurement point is behind the probe tip);
4) the interpolation, extrapolation, and averaging algorithms used.
The computational conditions (such as the number of grid points, the grid increments, and the
number of interpolation points in the three directions) shall be the same for all the functions.
O.8.3.3 SAR evaluation with curved surfaces
The procedures in O.8.3.1 and O.8.3.2 assume that the tissue-equivalent medium and
phantom boundary is planar. However, the uncertainty estimated with these functions for flat
tissue-equivalent medium and phantom boundaries is also valid for smoothly-varying curved
surfaces, provided that the four side faces are parallel to the line normal to the phantom at the
centre of the cube face next to the phantom surface. The fact that the function is based on a
flat surface does not impose any restrictions with respect to applicability for the test, provided
the procedure is based on equivalent distances of the grid from surfaces. This produces
averaging volumes such as illustrated in Figure O.1. Figure O.1 illustrates an acceptable
method for shaping the cube during SAR post-processing. The front face of the volume facing
the phantom/medium interface conforms to the curved boundary, to ensure that all SAR peaks
are captured. The back face shall be equally distorted to maintain the correct averaging mass.
IEC/IEEE 62209-1528:2020 – 235 –
© IEC/IEEE 2020
Figure O.1 – Orientation and surface of averaging volume relative to phantom surface
O.9 Tissue-equivalent medium properties uncertainty
O.9.1 General
Details of dielectric properties test methods are given in Annex H, and uncertainty estimation
methods are given in O.9.5.
In accordance with usual metrological practices, the measurement uncertainty for each of the
dielectric properties is required to be less than or equal to the allowable variations (see 7.2.1)
from the target values of the measured dielectric properties.
O.9.2 Medium density
A tissue-equivalent medium density of 1 000 kg/m 3 shall be applied for the SAR calculation
and averaging, regardless of the actual density of the medium used. This ensures that the
SAR and the volumes of the 1 g and 10 g averaging masses are independent of the chosen
recipe, where only the permittivity and conductivity are relevant. Therefore, no uncertainty is
associated with the density.
O.9.3 Medium conductivity uncertainty
The uncertainty due to the medium conductivity arises from two different sources. The first
source of uncertainty is the allowable tolerance from the target value of Table 2 (see 6.2.2),
and the second source of uncertainty arises from the measurement procedures used to
assess conductivity. The uncertainty shall be estimated using a normal probability. See 7.2.1
for applicable tolerances, and 7.8 for corrections of the measured SAR for the deviations in
conductivity.
O.9.4 Medium permittivity uncertainty
The uncertainty due to the medium relative permittivity arises from two different sources. The
first source of uncertainty is the deviation from the target value of Table 2 (see 6.2.2), and the
second source of uncertainty arises from the measurement procedures used to assess
relative permittivity. The uncertainty shall be estimated using a normal probability. See 7.2.1
for applicable tolerances, and 7.8 for corrections of the measured SAR for the deviations in
permittivity.
O.9.5 Assessment of dielectric properties measurement uncertainties
The measurement procedures specified in Annex H use vector network analysers for dielectric
properties measurements. Network analysers require calibration in order to account for and
remove inherent losses and reflections. The uncertainty budget for dielectric properties
measurements derives from inaccuracies in the calibration data, analyser drift, and random
errors. Other sources of errors are the uncertainties of the sample holder hardware,
deviations from the optimal dimensions for the specified frequencies, and sample properties.
This applies regardless of the type of sample holder and the nature of the scattering
parameters being measured. Uncertainties due to the straight-line fit in the slotted-line
method can be evaluated using a least-squares analysis [125].
– 236 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
An example uncertainty template is shown in Table O.3. All uncertainty parameters might or
might not apply to a specific test set-up or procedure, and other components not listed might
be relevant in some test set-ups. Measurement of well-characterized reference liquids can be
used to estimate the uncertainty of the measurement of the dielectric properties [126], [127],
[128], [129], using the following procedure.
a) Configure and calibrate the network analyser in a frequency span large enough around the
centre frequency of interest, of the tissue-equivalent medium used in the SAR
measurement.
b) Measure a reference material.
c) Repeat steps a) and b) at least n times (where n is at least three and is sufficient such that
the measurements have stabilized. n should be large enough to keep the repeatability in
step d) within the applicable tolerances as specified in 7.2.1 at all frequencies of interest.
Make the measurements at the same medium temperature at which the target dielectric
properties of the reference liquids are known. At each frequency, perform steps d) to g).
d) Calculate the repeatability as the sample standard deviation divided by the mean value.
For the permittivity, this is given by Formula (O.26):
n
1 1
∑( )
2
repeatability [%] =
100 × εr,′ i − εr′ (O.26)
εr′ n − 1 i =1
where the mean value is specified in Formula (O.27):
n
1
εr′ = ∑
n i =1
εr,′ i (O.27)
Do the same for the conductivity.

e) Enter the repeatability in Table O.3, and the number of degrees of freedom ν i = n − 1.
Determine the deviation of the dielectric properties from the target values ε r,ref and σ ref .
For the permittivity, this is given by Formula (O.28):
εr′ − εr,ref
′
deviation [%
= ] 100 × (O.28)
′
εr,ref
f) Enter the deviation in Table O.3, and the number of degrees of freedom ν i = n − 1. Do the
same for the conductivity.
g) Estimate the Type B uncertainties for the other components of Table O.3 (and other
relevant components if needed) in the frequency range under consideration.
h) Determine the combined standard uncertainty as the RSS of the uncertainty components
from steps d), e), and f). Enter this value in Table O.3.
i) For the relative permittivity, choose the frequency that gives the largest value for the
combined standard uncertainty in step g). Enter this uncertainty and the corresponding
degrees of freedom ν i into the appropriate row of Table 9. Do the same for the
conductivity.
In Table 9, the sensitivity coefficients c i for the measurement uncertainties of the conductivity
and permittivity of tissue-equivalent media are needed. These sensitivity coefficients are c σ for
conductivity and c ε for permittivity. They are calculated using Formula (9) through
Formula (12) (see 7.8.2).
For inductively-coupled sources, the highest sensitivity coefficients over the 30 MHz to
6 000 MHz frequency range were found to be c σ = 0,79 (at 180 MHz) and c ε = 0,23 (at
2 000 MHz) for 1 g averaging, and c σ = 0,77 (at 30 MHz) and c ε = 0,26 (at 5 500 MHz) for 10 g
averaging. For capacitively-coupled sources, the sensitivity coefficients can be different from
this, and should be evaluated using an approach similar to that of [28].
IEC/IEEE 62209-1528:2020 – 237 –
© IEC/IEEE 2020
These maximum values are entered into Table 9. Alternatively, maximum values for specific
tested frequency ranges may be entered.
Table O.3 – Example uncertainty template and example numerical

values for permittivity ( ε r′ ) and conductivity ( σ ) measurement

Input quantity X i Divisor u(x i ) = ci
distribution a(x i ) c i ∙u(x i ) νi
(source of uncertainty) qi a(x i )/q i (1 g, 10 g)
PDF i ±% ±%
Repeatability of εr′ or σ
N 5,2 1 1 5,2 4
(N repeats)
Deviation from reference
R 3,0 √3 1 1,73 4
liquid target εr′ or σ
Network analyser-drift,
R 0,5 √3 1 0,29 ∞
linearity, etc.
Test-port cable variations U 0,5 √2 1 0,35 ∞
Combined standard
5,50 5
uncertainty
O.9.6 Medium temperature uncertainty
The measurements of both SAR and tissue-equivalent medium dielectric properties shall be
performed at an ambient temperature between 18 °C and 25 °C, inclusively. These two
measured temperatures shall not differ by more than 2 °C. The following evaluation shall be
performed for each recipe to determine the uncertainty caused by the temperature tolerance.
This evaluation is typically done only once per recipe at the frequencies of interest. This
evaluation shall be performed for every new recipe, or whenever modifications are made to a
recipe.
Measurements of the dielectric properties at medium temperatures T low = 18 °C and

T high = 25 °C shall be used to compute temperature uncertainty, according to Formula (O.29).
The conditions that apply here only concern the evaluation of the medium temperature
uncertainty, but do not affect the temperature requirements during testing.
2 ×  εr (Thigh ) − εr (Tlow )  2 °C
εliquid temperature uncertainty [%] =
100 × ×
εr (Thigh ) + εr (Tlow ) Thigh − Tlow
(O.29)
2 × σ (Thigh ) − σ (Tlow )  2 °C
σliquid temperature uncertainty [%] =
100 × ×
σ (Thigh ) + σ (Tlow ) Thigh − Tlow
where
ε liquid temperature uncertainty is the temperature uncertainty for the medium permittivity as a
percentage;
σ liquid temperature uncertainty is the temperature uncertainty for the medium conductivity as a
percentage;
ε r (T high ) is the relative permittivity at temperature T high ;
ε r (T low ) is the relative permittivity at temperature T low ;
σ(T high ) is the conductivity at temperature T high ;
σ(T low ) is the conductivity at temperature T low ;
T high is the highest temperature in °C at which the dielectric properties
were measured;
– 238 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
T low is the lowest temperature in °C at which the dielectric properties
were measured.
These formulas can be used to derive the temperature uncertainty for the particular tissue-
equivalent medium. The uncertainty of the measurements for T low and T high shall be less than
0,1 °C. Note that this tolerance is applicable only for this uncertainty evaluation; this does not
affect nor apply for the temperature requirements during DUT testing.
The values of ε liquid temperature uncertainty and σ liquid temperature uncertainty are entered into the
appropriate rows and column in Table 9. Calculated values for some recipes are provided in
Annex F. A rectangular probability distribution is assumed for the temperature uncertainty in
Table 9.
IEC/IEEE 62209-1528:2020 – 239 –
© IEC/IEEE 2020
Annex P
(normative)
Post-processing techniques
P.1 Extrapolation and interpolation schemes
P.1.1 General
The local SAR inside the phantom shall be measured using small dipole sensing elements
enclosed in a probe cover/case. Probe calibration (involving E-field measurement) may be
carried out with respect to the geometric centre of the internal dipoles set, in which case the
spatial offset of the dipoles of a few millimetres from the physical tip of the probe shall be
taken into account when defining the measurement positions. In order to minimize probe
boundary effect uncertainties, the probe tip shall not be in contact with the phantom surface
although this usually is where the highest local SAR values usually occur. The assessment of
these highest local SAR values is essential to determine the psSAR, and shall therefore be
obtained by extrapolation from measurements carried out at a range of distances from the
phantom surface. The accurate assessment of the psSAR requires a fine spatial resolution
according to Table 4 within a three-dimensional scanned volume. The measured data shall be
extrapolated and interpolated to provide a data array with sufficient resolution to accurately
calculate the overall psSAR. The measurement uncertainty resulting from these interpolation,
extrapolation, and other numerical procedures (integration, averaging, etc.) shall be
determined. The location uncertainty of measurement points is determined as a separate
uncertainty component.
P.1.2 Extrapolation schemes
Since the actual measurement location of the field probe corresponds to the geometrical
centre of the dipole sensors, which is displaced from the tip of the probe, the SAR values
between the surface of the phantom and the closest measurable points required for computing
the 1 g or 10 g psSAR shall be determined by extrapolation.
Extrapolation can be performed using splines, biharmonic splines, wavelets, polynomials, or

rational functions. Various computational mathematics books describe how to implement
some of these methods.
P.1.3 Interpolation schemes
The measured and extrapolated SAR values within the zoom-scan volume are interpolated to
a 1 mm grid in order to determine the 1 g or 10 g psSAR, for example by using three
staggered one-dimensional cubic splines [130].
Interpolation can be performed using various mathematical techniques, such as statistics

[131], basis function curve-fits [132], Fourier analysis [133], and wavelet [134], polynomial, or
spline curve fits [135]. Various numerical analysis textbooks, e.g. [130], describe how to
implement some of these methods.
P.2 Averaging scheme and maximum finding
P.2.1 Volume average schemes
After extrapolation and interpolation, the cubical volumes evaluated to average the local SAR
shall extend to the phantom surface, in order to include the highest values of local SAR. The
contour of the bottom surface of the averaging volume shall follow the contour of the phantom
surface (Figure O.1). The top surface of the averaging volume shall be the same shape as the
bottom surface. The other four side surfaces of the averaging volume shall each be parallel to
– 240 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
the normal of the centre of the bottom surface. This ensures that the extruded volume is close
to a cubical shape and conformal to the phantom surface.
P.2.2 Finding the psSAR and estimating the uncertainty
The psSAR is found by moving the averaging volume specified in P.2.1 over the zoom scan
volume. At each point (with a resolution of 1 mm or better) the SAR is averaged over the
averaging volume. Averaging may use the trapezoidal algorithm or other methods. The psSAR
is the highest of these averaged SAR values over all locations.
If the highest 1 g or 10 g cube touches the boundary of a zoom-scan volume, the entire zoom
scan shall be repeated with the new centre located at the maximum psSAR location indicated
by the preceding zoom scan measurement.
The uncertainty contribution of the averaging scheme and maximum finding is included in the
evaluation methods of Clause O.8; the methods of Clause O.8 serve not only as benchmarks
for the interpolation and extrapolation schemes, but also for the averaging and maximum-
finding schemes.
IEC/IEEE 62209-1528:2020 – 241 –
© IEC/IEEE 2020
Annex Q
(informative)
Rationale for time-period averaged SAR test procedure
The SAR measurement procedures specified in 7.4 refer to handsets, laptops, tablets, and
other devices within the scope of this document that do not have autonomous and real-time
capability to calculate and control their transmission power and/or duty factor. Recent devices
have evolved significantly in this respect, and the measurement procedures shall take these
capabilities into account.
RF exposure guidelines specify basic restrictions in terms of the psSAR integrated over a
certain period of time. A six-minute averaging time is specified by the ICNIRP Guidelines [3]
and by IEEE Std C95.1-2019 [4]. Until now the SAR measurement and testing standards have
not taken these integration times into account, because historically the DUTs were not able to
calculate or limit their transmit power or duty factor.
Traditionally there are two basic components needed to ensure that the exposure of the user
does not exceed SAR exposure limits.
• The detection component detects when the exposure of a user could exceed the limits.
• The mitigation component provides a method to reduce the exposure when high exposure
has been detected.
Typically devices have utilized a proximity sensor for body detection, and look-up tables for
static power reduction as a means of SAR mitigation. Typically such an implementation
detects the physical proximity of the user to the antenna. When the proximity sensor is
triggered, the device reduces its transmit power if needed to ensure that the psSAR levels are
within the specified limits.
More recent devices apply algorithms which accurately track, calculate, and control the
transmit power level and/or duty factor over the averaging period with high precision and
accuracy. Such devices typically limit the transmission duty factor, output power, or both
when necessary to reduce the time-period averaged SAR.
These algorithms and techniques can be tested for compliance according to the procedure
specified in 7.6.
– 242 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex R
(normative)
Measurement uncertainty analysis for testing laboratories
R.1 RF ambient conditions
The ambient RF level shall be evaluated by performing SAR measurements using the same
equipment set-up as used for testing a DUT, but with the RF power switched off. For both 1 g
and 10 g averaging mass, the SAR due to RF ambient shall be less than or equal to
0,012 W/kg (i.e. 3 % of 0,4 W/kg).
It is not necessary to check RF ambient noise prior to each SAR test, provided the laboratory
can demonstrate that there are no new RF sources.
The test configurations specified in Clause A.2 are used to assess the effects of reflections
from nearby objects at the test site. The total amount of reflections shall comply with
requirements of 6.1. In addition, the RF ambient noise shall be determined by performing SAR
measurements with all local RF sources switched off. The allowable variation shall be within
±3 % of 0,4 W/kg, and the uncertainty shall be estimated assuming a rectangular probability
distribution.
When SAR is measured in a controlled environment, such as an anechoic chamber, RF

ambient effects shall be assessed at least once per year. Otherwise, RF ambient effects
should be assessed before performing any SAR measurement on a DUT, and ambient
conditions shall be monitored during the measurement so that any nearby high output
sources, such as walkie-talkies, do not affect the SAR measurements.
The rationale for the non-controlled environment RF check evaluation is that there is no
reason to assess this uncertainty contribution before every SAR measurement if it can be
demonstrated that RF sources are sufficiently far from the SAR measurement system, even if
the measurement system is placed in a non-controlled environment, due to the near field
nature of the SAR measurement. The rationale on calibration intervals is specified in
ISO 10012:2003 [136]. It is recommended to periodically evaluate the RF ambient effects on
SAR measurements.
R.2 Device positioning and holder uncertainties
R.2.1 General
A device holder is used to keep the test position of the DUT at the phantom during a SAR
measurement. Because the device holder can influence the characteristics of a DUT, the SAR
uncertainty due to the device holder perturbation shall be estimated using the procedures in
R.2.2. Procedures to assess the SAR uncertainties due to positioning variations due to
mechanical uncertainties of the device holder are discussed in R.2.2.2 and R.2.2.3. Both
subclauses include procedures for device-specific and predetermined uncertainties. If
predetermined uncertainties are used, in most cases multiple repetitions of device-specific
tests are required to further reduce the predetermined standard deviations.
IEC/IEEE 62209-1528:2020 – 243 –
© IEC/IEEE 2020
R.2.2 Device holder perturbation uncertainty
R.2.2.1 General
The device holder shall be made of low-loss dielectric material, with a loss tangent ≤ 0,05 and
relative permittivity ≤ 5. (These material parameters can be determined, for example, using
the coaxial contact probe method.) Nevertheless, some holders might still affect the DUT, so
the uncertainty due to the holder (i.e. the deviation relative to a set-up without the holder)
shall be estimated. The uncertainty of the device holder for testing a specific DUT shall be
estimated according to the Type B method specified in R.2.2.2. Alternatively, the Type A
method specified in R.2.2.3 can be used to assess the uncertainty for a group of DUTs with
similar SAR characteristics, if they are tested with the same device holder.
R.2.2.2 Device holder perturbation uncertainty for a specific DUT: Type B
The uncertainty for a specific DUT operating in a specific device holder shall be estimated by
performing the following two tests using a flat phantom:
a) assessment of the 1 g or 10 g psSAR (SAR w/ holder ) by placing the device in the holder in
the way the DUT would be held when tested next to the head, then by positioning the DUT
in direct contact with a flat phantom (horizontal and vertical centreline of the DUT parallel
to the bottom of the flat phantom);
b) assessment of the 1 g or 10 g psSAR (SAR w/o holder ) by placing the device in the same
position as in a), but held in place using foamed polystyrene or an equivalent low-loss and
non-reflective material (permittivity not greater than 1,2, and loss tangent not greater than
10 −5 ).
The SAR uncertainty to be used in Table 9 is in Formula (R.1):
 SARw/ holder − SARw/o holder 

SARuncertainty % 
=  × 100 (R.1)
 SARw/o holder 
 
where

SAR w/ holder is the SAR with device holder in W/kg;
SAR w/o holder is the SAR without device holder in W/kg.
This uncertainty has an assumed rectangular probability distribution, and ν i = ∞ degrees of

freedom.
R.2.2.3 Device holder perturbation uncertainty for a specific test device: Type A
The specific-device holder-disturbance uncertainty data evaluated according to the Type B

method described in R.2.2.2, for a group of DUTs with the same shape and with substantially
equivalent SAR distribution characteristics, shall be statistically compiled and applied to
selected groups of devices tested in the same device holder and the same configurations. The
statistical procedures shall include at least six devices, each evaluated according to the
Type B method described in R.2.2.2. As more DUTs with similar SAR characteristics are
evaluated with the same device holder, the perturbation error may be included in a Type A
uncertainty analysis, and applied to future SAR evaluations of similar device and holder
configurations. Increasing the number of test devices will increase the degrees of freedom (ν i )
and lower the coverage factor (k p ). The coverage factor (k p ) for such conditions shall be
determined as k p = t p (ν eff ) where t p (ν eff ) is the coverage factor of the t-distribution, and ν eff is
the effective number of degrees of freedom estimated using the Welch-Satterthwaite formula
(see ISO/IEC Guide 98-3). The effect of the device holder for N different models of DUTs in
the different configurations shall be estimated by performing the tests of R.2.2.2 according to
the Type B method described in this paragraph for each model (N shall be at least 6).
– 244 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
The corresponding uncertainty for Table 9 shall be estimated by using the RMS of the
individual uncertainties, with ν i = N – 1 degrees of freedom. It is recommended that the
database be updated yearly, in order to account for new DUT designs.
R.2.3 DUT positioning uncertainty with a specific test device holder: Type A
R.2.3.1 General
The deviation of the actual DUT position from the positions specified in 7.2.4 depends on the
precision of the DUT positioner, as well as on the interpretation and handling by the person
performing the test. In addition, the magnitude of this deviation on the psSAR values depends
on the DUT design. Since these parameters cannot be separated, the Type A tests specified
in R.2.3.2 and R.2.3.3 shall be performed.
R.2.3.2 Positioning uncertainty of a specific DUT in a specific device holder
The positioning uncertainty of a specific DUT tested in a specific device holder is assessed by
repeating measurements of the 1 g or 10 g psSAR. This positioning uncertainty shall be
evaluated using the antenna position, frequency channel, and device position for the
operating mode that produced the highest SAR among all frequency bands. In addition to the
original SAR measurement, the DUT shall be repositioned and the tests repeated at least four
more times. This minimum of five tests is sufficient to establish a reasonable value for the
degrees of freedom. If the positioning uncertainty of an individual device is suspected to be
large, more tests might need to be performed, to reduce the impact on the total measurement
uncertainty. Increasing the number of tests will increase the effective degrees of freedom
(ν eff ) and decrease the coverage factor. The average SAR for the total number of
measurements (N) is used to determine the SAR uncertainty, according to the standard
deviation and degrees of freedom (ν i = N − 1) of the number of tests performed.
R.2.3.3 Positioning uncertainty of specific types of DUTs in a specific device holder
A Type A uncertainty analysis can be applied for a group of DUTs with predominantly the
same shape and substantially equivalent physical dimensions and SAR distributions. The
tests shall include at least six separate devices, each evaluated according to the procedures
for a specific DUT in a specific device holder described in R.2.3.2. The number of tests n shall
be at least 5, and each of the n tests shall be performed for all M devices. For the sagitally-
bisected SAM phantom, the n tests shall include the cheek position and the tilt position. For
other phantoms, relevant device positions should be used. The corresponding uncertainty
shall be estimated by applying the RMS of the M individual standard deviations. The value to
be filled in the uncertainty table shall be the standard uncertainty with k = 1. The degrees of
freedom are determined according to the total number of tests N = n × M. For the M devices
included in the specific group of DUTs, ν i = N − 1. If this procedure is applied to determine
uncertainty, it might be unnecessary to apply the procedures for a specific DUT in a specific
device holder described in R.2.3.2 to individual DUTs. It is recommended that the database
be updated yearly, in order to account for new DUT designs.
R.3 Probe modulation response
Probe sensor responses to modulated signals that are based on diode detectors can be
complex because the diodes are non-linear elements. The diode response theory for complex
signals is described in Nadakuduti et al. [20]. The linearization parameters for a particular
operating mode shall be determined by relative experimental calibration, i.e. by evaluating a
power sweep for a particular operating mode, as specified in Clause E.2. The linearization
parameters shall be determined for each sensor separately.
IEC/IEEE 62209-1528:2020 – 245 –
© IEC/IEEE 2020
The following uncertainty can be determined by using any source (e.g. waveguide or dipole)
with a set-up equivalent to that specified in Figure A.1. The signal generation set-up shall
simulate the operating mode for which the uncertainty is determined, according to the
specification of the operating mode standard. The power shall be increased from a probe
sensor voltage equivalent to lower than P 0 – the power transmitted to produce a SAR of
0,1 W/kg – to the equivalent of higher than P 1 – the power transmitted to produce a SAR of
10 W/kg – for the investigated sensor, in 5 dB steps. At each power level, the SAR shall be
measured with the modulated signal, and with a CW signal at the same RMS power. (It shall
be verified that the power meter uses a true RMS detector, and that the amplifier is
sufficiently linear over the entire dynamic range of the signals.) This procedure shall be
repeated for each field sensor.
Formula (R.2) is used to derive the modulation uncertainty of a particular modulation

operating mode mod X.
 P + 20 dB 
0 SAR ( Pi )mod X , j 
[%]
SARmodulation uncertainty,mod X= max  max  100 × −1 (R.2)
j∈{x , y , z}  Pi = P0  SAR ( Pi )cal X , j 
  
where
SAR modulation uncertainty,modX is the uncertainty for the particular operating mode mod X as a
percentage;
SAR(P i ) modX,j is the SAR measured with the modulated signal at power level P i
for sensor index j;
SAR(P i ) cal X,j is the SAR measured with the calibrated signal at the same RMS
power for sensor index j.
The SAR uncertainty is determined as the maximum of all SAR(P i ) modX at each step for all
three sensors. The index j refers to the field sensor, for each of which the procedure shall be
repeated. A rectangular probability distribution has been assumed for probe modulation
response uncertainty in Table 9.
R.4 Time-period averaged SAR
R.4.1 General
The uncertainty of time-period averaged SAR (TPAS) is the combined uncertainty of the TX
factor uncertainty assessment (TXF) and the uncertainty of SAR measurement of a DUT
(∆SAR).
R.4.2 TX factor uncertainty

a) When the TX factor is assessed using SAR measurement, its uncertainty (TXF) is
evaluated by measuring a CW signal with a pulse modulated amplitude. The amplitude
variation shall be appropriate for the signal being measured. As an example, a signal
amplitude may be used that has a square pulse shape (with 1 ms rise time) that varies in
sequence such that is produces SAR results between three levels: 0,3 W/kg, 0,5 W/kg,
and 1 W/kg, with amplitude durations of 1 s, 5 s, and 10 s. The measured TX factor is
compared to its theoretical value (0,6 W/kg in this case). The worst-case deviation is
taken as the sampling rate uncertainty assuming rectangular distribution.
b) When the TX factor is assessed using conducted power measurements (e.g. with a power
meter) the uncertainty for pulse signals (applicable to the DUT testing) is considered to be
the TX factor uncertainty assessment (TXF).
NOTE The requirement on the signal amplitude variation might change as more DUTs applying time averaging are
developed (see Annex Q).
– 246 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
R.5 Measured SAR drift
R.5.1 General
The SAR measurement drift of the DUT is accounted for by the first and last step of the
measurement process specified in 7.4.2, by using the following two methods.
a) As the preferred method, local SAR measurements are made by the SAR measurement
system prior to the area scan. A secondary measurement is made by the system at the
same point after completion of the psSAR measurement. The measurement is performed
within the tissue-equivalent medium at a reference point where the SAR values for both
primary and secondary measurements exceed the lower detection limit of the
measurement system. The distance from the reference point to the inner phantom surface
in the normal direction shall be less than or equal to 10 mm.
b) Alternatively, and if the preferred method in a) is not sensitive enough, conducted power
measurements can be made on the device at the antenna port, using equipment capable
of measuring RF power prior to device placement for SAR testing. A secondary RF power
measurement shall be made after the SAR testing is completed.
c) In either case, the drift is recorded as the percentage difference of the secondary
reference measurement, Ref secondary (SAR or conducted power), from the primary
reference measurement, Ref primary as described in Formula (R.3):
drift 100 % ×
=
( Refsecondary − Refprimary ) (R.3)
Refprimary
Commercial DUTs shall have SAR drifts within ±5 %. Some devices could have significant
fluctuations in output power that are not classifiable as undesirable power drift, but rather
are a characteristic of the normal operating behaviour of the device. In this case, other
methods, such as SAR scaling, shall be considered to ensure that an accurate and
conservative SAR is obtained.
d) If the SAR drift cannot meet the 5 % threshold while performing the SAR testing in
accordance with 7.4.2, then a drift measurement for the longest intended measurement
evaluation time shall be performed without recharging the battery. This is done by
performing a measurement according to method a) or b) of this subclause continuously
over the evaluation time (at least once every 5 s). This time sweep measurement shall be
performed at each frequency band for the operating mode with the highest time-averaged
output power.
– If the difference between the maximum and minimum in the time sweep is less than
5 % of the average value, or if the difference is less than 10 % and the SAR primarily
decreases during the time sweep (i.e. if it does not increase by more than 2 % at any
time during the time sweep), it is sufficient to perform reference measurements at the
beginning of the area scan and at end of the last zoom scan, as specified in 7.4.2 a)
and f).
– Otherwise, additional reference measurements shall be taken during the zoom scan,
and the zoom scan measurements shall be corrected prior to performing extrapolation,
integration, and averaging. Linear interpolation between the reference measurements
is performed prior to the correction. The SAR values measured during the zoom scan
shall be corrected by the difference between the interpolated values and the first
reference value measured before the area scan. The time between reference
measurements during the zoom scan shall be sufficiently small, such that correction of
the time sweep curve specified in this paragraph shall be conservative for all points.
R.5.2 Accounting for drift

a) If the SAR drift is within 5 %, then it can be treated either as an uncertainty (i.e. random
error), or a systematic offset. If the drift is larger than 5 %, the measurement drift shall be
considered a systematic offset rather than an uncertainty.
IEC/IEEE 62209-1528:2020 – 247 –
© IEC/IEEE 2020
b) If treated as an uncertainty, the absolute value of the drift shall be entered into the
uncertainty table. The drift shall not be added to the assessed SAR value. The uncertainty
value reported in the uncertainty budget shall either be the value corresponding to the
highest SAR measurement drift reported, or the maximum allowed 5 %.
c) If treated as a systematic offset, compensation to the measured SAR shall be applied by
adding the absolute difference to the determined SAR value if the drift is either negative or
positive, as described in Formula (R.4):
= SARmeasured × (1+ drift / 100 % )

SARcompensated (R.4)
In this case it is not necessary to record the drift in the uncertainty budget (i.e. u i = 0 %).
d) To maintain a conservative value for the resulting SAR, drifts shall not be subtracted from
the assessed SAR. If different assessed operating modes of the device exhibit different
drift ratios, all the corresponding measured SAR values may be compensated by the same
ratio, provided that the applied drift ratio is the largest one detected during the SAR
testing over all operating modes of the device. The uncertainty shall be estimated
assuming a rectangular probability distribution.
R.6 SAR scaling uncertainty
Uncertainties of the SAR scaling (7.2.3.2) are associated with non-linearities of the signal and
RF amplifier stages, the modulation signal bandwidth, and antenna impedance.
For SAR scaling uncertainty, a rectangular probability distribution shall be assumed. The
uncertainty is evaluated by determination of the SAR of mod Y at the psSAR location (x p, y p, z p )
using the following procedure.
a) Perform an area scan with mod X according to 7.4.2.

b) Move the probe to the peak location of the area scan.
c) Take the SAR reading at the peak location with mod X.
d) Switch the device to mod Y (without moving the device).
e) Take the SAR reading with mod Y .
f) Calculate the ratio of the measured and scaled SAR modY using Formula (R.5):
(
 SAR xp , yp , zp ) 
mod Y − 1 × 100 %
SARscaling uncertainty 
= (R.5)
 ( )
 SAR xp , yp , zp ⋅R
mod X p


with R p as the time-averaged RF output power ratio of mod X and mod Y , according to
7.2.3.2.
If SAR scaling uncertainty > 5 %, do not use the scaling; instead perform full SAR testing for mod Y .
– 248 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex S
(normative)
Validation antenna SAR measurement uncertainty
S.1 Deviation of experimental antennas
The system validation antennas in Annex D are well specified, and were simulated to obtain
numerical reference values using validated numerical codes. However, the mechanical and
electrical uncertainties of each specific antenna affect the resulting psSAR values, due to, for
example, different feed-point impedance and current distribution as function of distance,
phantom shell, tissue-equivalent medium, etc. In other words, the numerical target values are
valid for the specific antennas and set-up configuration requirements. The deviation from the
target values shall be determined either by Type A or Type B evaluations. Type A evaluations
should use statistical evaluation of several measurements using different tissue-equivalent
media, probes, and phantoms. For Type B evaluations, all parameters need to be assessed
experimentally. Numerical target values are established by numerical simulations and
validated laboratory calibration.
For the standard waveguide, Table S.1 provides the contributions to the deviations of the
waveguide from theory.
For the standard dipole, confined loop antenna, meander dipole, and VPIFA, contributions to
the deviation of experimental antennas from theory include variations in the physical
parameters specified in Annex D.
The combined uncertainty from the deviations in experimental antennas from theory is
entered into Table 9, with an assumed normal probability distribution.
Table S.1 – Uncertainties relating to the deviations of

the parameters of the standard waveguide from theory
Prob. Uncert'nty u(y) =

Input quantity X i Div. u(x i ) = ci
Symbol Dist. a(x i ) c i ∙u(x i ) νi
(source of uncertainty) qi a(x i )/q i (1 g, 10 g)
PDF i ±% ±%
∆TE 01 Variation from theoretical

R 5,0 √3 2,9 1 2,9 ∞
waveguide mode in medium
L, W Waveguide dimensions N 1,0 1 1,0 1 1,0 ∞
u c (y) Combined uncertainty RSS 3,1 ∞
S.2 Other uncertainty contributions when using system validation antennas
In addition to the uncertainty terms for the system validation antenna discussed elsewhere,
there are additional terms to be added in the uncertainty budget. These terms depend on the
type of system validation antenna used. For a dipole antenna, meander dipole, or VPIFA,
such a term can be the distance of the feed-point to the tissue-equivalent medium, and for a
waveguide or confined loop antenna, such a term can be the mismatch error. Table S.2 and
Table S.3 show the uncertainty terms for the dipole antenna and the waveguide, respectively.
The expanded measurement uncertainty shall be within ±10 % for k = 2. These additional
uncertainty terms are entered into the row VAL of Table 9. The numerical values in this table
are examples only, and should not be assumed to represent the values for specific antennas.
NOTE The waveguide is placed directly against the phantom, as specified in Annex D. Therefore, no uncertainty
term is needed to account for the distance to the phantom, like with the dipole antenna.
IEC/IEEE 62209-1528:2020 – 249 –
© IEC/IEEE 2020
Table S.2 – Other uncertainty contributions relating
to the dipole antennas specified in Annex D.

Input quantity X i Divisor u(x i ) =
distribution a(x i ) ci c i ∙u(x i ) νi
(source of uncertainty) qi a(x i )/q i
PDF i ±% ±%
Dipole axis to medium distance R 2,0 √3 1 1,2 ∞
Table S.3 – Other uncertainty contributions relating to

the standard waveguides specified in Annex D

Input quantity X i Divisor u(x i ) =
distribution a(x i ) ci c i ∙u(x i ) νi
(source of uncertainty) qi a(x i )/q i
PDF i ±% ±%
a
Mismatch error for system U-shaped 5,0 √2 1 3,5 ∞
Uncertainty in measurement of
power transmission loss of adapter N 1,0 2 1 0,5 ∞
and waveguide.
Combined uncertainty RSS ± 3,5 ∞
a This is calculated for 8 dB return loss to the waveguide, 30 dB return loss for the power sensor, and 25 dB
return loss for the output port of the coupler.
– 250 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex T
(normative)
Interlaboratory comparisons
T.1 Purpose
The purpose of the interlaboratory comparison is to validate various SAR measurement

systems by comparing results from several different laboratories. The general concept is to
use reference DUTs and the required phantoms specified in this document. The SAR
assessment methodologies specified in Clause 7 are used.
The measured data from various laboratories shall be compared. Also, a comparison of
measurement uncertainty analyses among the test laboratories shall be performed. If the
differences and discrepancies between the measured data can be attributed to differences in
the measurement uncertainties, the interlaboratory comparison is deemed to be successful. If
not, all sources of uncertainties shall be investigated in accordance with the recommendations
of Clause 8. Also, other possible sources of uncertainties different from those specified in
Clause 8 shall be considered.
NOTE General guidelines for accomplishing interlaboratory comparison tests are given for example in [137] and
[138].
T.2 Phantom set-up
The required phantoms are specified in this document. The phantom shall have a mounting
structure made of a rugged material, with low permittivity (less than 5) and low loss tangent
(less than 0,05). Metallic parts closer than 50 cm from the structure shall be avoided. The
phantom shall be irradiated using a reference DUT, mounted according to the test positions
specified in 7.2.4.
T.3 Reference devices
Reference devices are commercial wireless DUTs. The set of reference devices may also be
extended to include non-commercial sources. Each reference device shall have evaluated
target SAR values with associated measurement uncertainty. The designation of the supplier
is based on the agreement between the participating parties.
T.4 Power set-up
The power set-up is dependent on the precision of the output power measurement of the
reference device used for the interlaboratory comparison. Each device is individually checked
by the participating laboratory, so that the maximum conducted output power of each
reference device is in the range of ±0,3 dB from the reference value. The measured output
power shall be reported to the monitor laboratory. The device output power shall be adjusted
via test software to a predetermined value. The device battery shall be fully charged.
Careful effort shall be expended to ensure an accurate output power measurement. Some
devices may allow easy and accurate output power measurements at the antenna port. The
ease of output-power measurement shall be considered when designating reference devices
for use in the comparison tests. If deemed necessary, the reference devices shall be returned
to a monitor laboratory periodically during the comparison campaign, for verification of SAR,
output power, and frequency. Also, the battery condition shall be checked by the monitor
laboratory, to ensure uniform power output at all laboratories.
IEC/IEEE 62209-1528:2020 – 251 –
© IEC/IEEE 2020
T.5 Interlaboratory comparison – procedure
Measurement procedures for SAR assessment used in the interlaboratory comparison are the
same as those used for compliance testing, in accordance with Clause 7. When a full SAR
testing system or a fast SAR testing system is used, the interlaboratory comparison shall be
performed with a calibrated and validated system. Each laboratory shall provide a full
measurement report in accordance with the requirements of Clause 8 and Clause 9, including
system validation, dielectric properties measurements, system-contributed uncertainty data,
and measured output power data. If a fast SAR testing system that does not meet all of the
specifications of this document is used to perform SAR measurements, the uncertainty should
follow general guidance provided in Clause C.1, if applicable.
Results of the interlaboratory comparison shall be compared to their respective target values.
Each result shall fall within the expanded measurement uncertainty of the measured value
and target value. The reports from different laboratories are to be evaluated and compared by
the designated monitor laboratory.
– 252 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex U
(informative)
Determination of the margin for

compliance evaluation using the Uniphantom
U.1 General
When comparing the psSAR measured with the Uniphantom with the psSAR limit, it is
necessary to consider a margin in order to compensate the possibility of the underestimation
of measured psSAR. This margin shall be appropriately determined based on statistical data.
Some examples of the determination of the margin are specified in Annex U.
U.2 Deviation of the psSAR measured using the Uniphantom from the psSAR
measured using the SAM phantom
The Uniphantom is designed to obtain a more conservative SAR value in comparison to the
SAM phantom. In some cases, however, the SAR value measured using the Uniphantom can
be lower than the psSAR measured with the SAM phantom. The compared results of the
measured psSAR between the Uniphantom and the SAM phantom are can be categorized into
four “Classes,” as shown in Figure U.1.
It is important to avoid a false negative as shown by the Class 4 results of Figure U.1,
whereas a false positive can be checked by a full-compliance SAR measurement following
after the evaluation using the Uniphantom. A margin is therefore necessary to compensate
the measured SAR using the Uniphantom. However, an overly-large margin is inappropriate,
because then most cases would be categorized as false positives, which would need full SAR
testing using the SAM phantom.
The 95 % confidence interval of the deviation of the measured SAR between the Uniphantom
and the SAM phantom is used to determine the margin in Annex U. This margin means that
5 % of the cases are misclassified if the measured SAR of the Uniphantom is higher than the
compensated SAR limit, which generally is very rare if the 95 % confidence interval is
relatively small.
NOTE The psSAR limit is also shown as a dashed red line.
Figure U.1 – Categories (classes) for comparison of the measured psSAR between the
Uniphantom (SAR uni ) and the SAM phantom (SAR SAM )
IEC/IEEE 62209-1528:2020 – 253 –
© IEC/IEEE 2020
U.3 Determination of margin based on 95 % confidence interval
a) Measure maximum psSAR values using the Uniphantom (SAR flat and SAR corner ) and the
SAM phantom (SAR SAM ) for handsets of the same shape. SAR flat is the measured SAR for
the handset positioned at the flat bottom of the Uniphantom, and SAR corner is the
measured SAR for the handset positioned at the curved edge (cross-section corner) of the
Uniphantom. SAR SAM is the highest value of the maximum condition among the left and
right sides and the cheek and tilt positions, and so on. The number of handsets is N.
b) Calculate the deviations of the measured SAR between the Uniphantom and the SAM
phantom (Dev) for each handset as in Formula (U.1):
Dev SARuni − SARSAM

= (U.1)
where SAR uni is SAR flat or SAR corner .

c) Calculate the average (Dev ave ) and the standard deviation (Dev std ) of Dev for N handsets.
d) Calculate the 95 % confidence interval, based on the t-distribution with N − 1 degrees of
freedom (Int 95 % ).
e) Calculate the margin (SAR limit margin ) as in Formula (U.2):
SARlimit margin = Devstd × Int95 % − Devave (U.2)
f) If SAR limit margin is negative, do not apply the margin (i.e. SAR limit margin = 0).
If SAR limit margin is larger than the psSAR limit value (SAR limit ), do not apply the evaluation
using the Uniphantom.
g) Test for compliance by comparing the SAR uni and the compensated psSAR limit as follows:
1) compliant if SAR uni ≤ SAR limit − SAR limit margin ;
2) not compliant if SAR uni > SAR limit − SAR limit margin .
NOTE Some national regulatory authorities might have other or different requirements for compliance
demonstration.
The uncertainty shall also be considered into SAR uni if the uncertainty is larger than 30 %,
same as for the full-compliance SAR measurement using the SAM phantom.
U.4 Examples of the determination of the margin factor
U.4.1 Margin for handsets with straight form factors at flat-bottom position
In Clause U.4, 45 handsets with straight form factors are considered to determine the margin
for the 10 g psSAR limit (2 W/kg), and 40 handsets are considered for the 1 g psSAR limit
(1,6 W/kg). The information used for the determination of the margin is summarized in
Table U.1. The statistical distributions are also shown in Figure U.2 and Figure U.3.
Figure U.2 and Figure U.3 show that the margin based on the 95 % confidence interval is
conservative. Note that this high ratio of the number of false-positive handsets is caused by
the inclusion of the SAR results of the smart phones. This result suggests that the smart
phones and the other handsets could be separately considered to apply the SAR margin, or
the measurement of the flat bottom of the Uniphantom could be skipped. The margins for the
smart phone and the other handsets are determined in U.4.2 to U.4.5.
– 254 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table U.1 – Summary of information to determine the margin for handsets
with straight form factors positioned at the flat bottom of the Uniphantom
Contribution 10 g psSAR 1 g psSAR

Number of handsets (N) 45 40
Average of the deviation (Dev ave ) (W/kg) 0,53 1,30
Standard deviation of the deviation (Dev std ) (W/kg) 0,67 1,58
t value of 95 % with N − 1 degree of freedom 1,68 1,68
Margin (SAR limit margin ) (W/kg) 0,60 1,36
Compensated psSAR limit (SAR limit − SAR limit margin ) 1,40 0,24
(W/kg)
Ratio of false-positive handsets to all of the measured
29 % 98 %
handsets (Class 3 in Figure U.1) a
a The handsets should be measured at the curved edge (corner) of the Uniphantom or using
the SAM phantom in the same way as in the full-compliance SAR measurement.
NOTE The normal distribution and its 95 % confidence interval with the same average and standard deviation are
also shown.
Figure U.2 – Histogram of the deviation of the 10 g psSAR of 45 handsets

IEC/IEEE 62209-1528:2020 – 255 –
© IEC/IEEE 2020
also shown.

U.4.2 Margin for handsets with straight form factors (except smart phones at flat-
bottom position)
In U.4.2, 25 handsets with straight form factors, including slide-type handsets and bar
phones, but not smart phones, are considered to determine the margin for the 10 g psSAR
limit (2 W/kg), and 20 handsets are considered for the 1 g psSAR limit (1,6 W/kg). The
information used for the determination of the margin is summarized in Table U.2. The
statistical distributions are also shown.
conservative, which results in a lower ratio of false-positive handsets than in the cases
wherein all handsets with straight form factors including smart phones are considered (U.4.1).
– 256 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table U.2 – Summary of information to determine the margin for handsets
with straight form factors, including slide-type and bar handsets (except
smart phones), positioned at the flat bottom of the Uniphantom

(W/kg)
12 % 40 %
also shown.

straight form factors positioned at the flat bottom of the Uniphantom
IEC/IEEE 62209-1528:2020 – 257 –
© IEC/IEEE 2020
also shown.
Figure U.5 – Histogram of the deviation of the 1 g psSAR from 20 handsets with
straight form factors positioned at the flat bottom of the Uniphantom
U.4.3 Margin for smart phones at flat-bottom position
In U.4.3, 20 handsets with straight form factors or smart phones are considered to determine
the margin for the 10 g psSAR limit (2 W/kg), and 20 handsets are considered for the 1 g
psSAR limit (1,6 W/kg). The information used for the determination of the margin is
summarized in Table U.3. The statistical distributions are also shown.
conservative. Note that the ratio of the false-positive handsets for 10 g psSAR is more
reasonable than that in the cases where all handsets with straight form factors, and with
including smart phones, are considered (U.4.1). However, the ratio for 1 g psSAR is still
considerably high, which means that the SAR measurement at the flat bottom of the
Uniphantom should be omitted.
– 258 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table U.3 – Summary of information to determine the margin for the
smart phones positioned at the flat bottom of the Uniphantom

(W/kg)
30 % 85 %
also shown.
form factors or smart phones positioned at the flat bottom of the Uniphantom
IEC/IEEE 62209-1528:2020 – 259 –
© IEC/IEEE 2020
also shown.
form factors or smart phones positioned at the flat bottom of the Uniphantom
U.4.4 Margin for smart phones at corner position
In U.4.4, 20 smart phones are considered to determine the margin for the 10 g psSAR limit
(2 W/kg), and 19 smart phones are considered for the 1 g psSAR limit (1,6 W/kg). The
information used for the determination of the margin is summarized in Table U.4. The
statistical distributions are also shown.
Figure U.8 clearly shows that the margin based on the 95 % confidence interval is
conservative for the 10 g psSAR limit. On the other hand, there is one case where the
deviation is outside the 95 % confidence interval, as shown in Figure U.9. However, this is not
a false-negative case, i.e. Class 2 but not Class 4 per Figure U.1. Note that the ratio of the
false-positive handsets, which require further measurements, is extremely good; i.e. 0 %,
which is a significant improvement with respect to the cases of the measurement at the flat
bottom of the Uniphantom (Table U.5), because the deviation is significantly small, which
implies that the Uniphantom is an appropriate substitute for the SAM phantom.
– 260 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table U.4 – Summary of information to determine the margin
for smart phones positioned at the corner of the Uniphantom

Average of the deviation (Dev ave ) (W/kg) 0,03 −0,03
(W/kg)
0% 0%
also shown.
form factors or smart phones positioned at the corner of the Uniphantom
IEC/IEEE 62209-1528:2020 – 261 –
© IEC/IEEE 2020
also shown.
form factors or smart phones positioned at the corner of the Uniphantom
Table U.5 – Statistical analysis results of P(Tilt/Cheek > x) for various device types
x P(Tilt/Cheek > x)
Bar Clamshell Slide (Open) Slide (Closed)
1,0 6,853 × 10 −3 2,992 × 10 −8 5,441 × 10 −4 1,771 × 10 −3
1,1 1,334 × 10 −3 3,769 × 10 −10 2,138 × 10 −5 1,739 × 10 −4
1,2 1,982 × 10 −4 2,801 × 10 −12 4,385 × 10 −7 1,131 × 10 −5
1,3 2,240 × 10 −5 1,221 × 10 −14 4,658 × 10 −9 4,844 × 10 −7
2,0 1,998 × 10 −15 0,000 × 10 0 0,000 × 10 0 0,000 × 10 0
U.4.5 Margin for handsets with clamshell form factors at corner position
In U.4.5, 19 handsets with clamshell form factors are considered to determine the margin for
the 10 g psSAR limit (2 W/kg), and 19 handsets are considered for the 1 g psSAR limit
(1,6 W/kg). The information used for the determination of the margin is summarized in
Table U.6. The statistical distributions are also shown.
Figure U.10 and Figure U.11 clearly show that the margin based on the 95 % confidence
interval is conservative, which results in a good or low ratio of the false-positive handsets, i.e.
0 % for 10 g psSAR, and 11 % for 1 g psSAR. This also means that the curved-edge (cross-
section corner) of the Uniphantom is an appropriate substitute for the SAM phantom for
handsets with clamshell form factors.
– 262 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Table U.6 – Summary of information to determine the margin for handsets with
clamshell form factors positioned at the corner of the Uniphantom

(W/kg)
0% 11 %
also shown.

with clamshell form factors at the corner of the Uniphantom
IEC/IEEE 62209-1528:2020 – 263 –
© IEC/IEEE 2020
also shown.

with clamshell form factors at the corner of the Uniphantom
– 264 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex V
(informative)
Automatic input power level

control for system validation
V.1 General
The input power level (i.e. forward power to the antenna) in system validation and calibration
should be stable enough for reliable measurements. Typically, the input power for system
validation or probe calibration is controlled manually during the process, and the equipment
needs time (e.g. half an hour) to warm up such that the power from the signal generator and
amplifier stabilizes. For the reduction of the power uncertainty contribution, it is recommended
to apply automatic input power level control (AIPLC) methodology. The AIPLC is applied to
the system validation and probe calibration process, and helps to keep the input power
constant during the tests by controlling the signal generator and/or the amplifier.
The expanded uncertainty for the SAR measurement system is limited to 30 %; therefore, for
example a 5 % power drift contributes a significant amount to the total measurement
uncertainty.
Figure V.1 shows the power drift before applying AIPLC (left) and after applying AIPLC (right),
which can be compared with the power drift to operation time after its application.
a) Without AIPLC b) With AIPLC
Figure V.1 – Generated RF input power variations to

operation time without and with application of AIPLC
V.2 Operational mechanism of AIPLC
A schematic diagram of AIPLC scheme is shown in Figure V.2. A computer reads the power
meter, and controls the signal generator or the amplifier automatically through GPIB or a
RS-232C cable.
Two adjusting methods are available to control the RF power output level: one is to adjust the
amplifier, and the other is to adjust the signal generator. Figure V.3 shows the power variation
patterns when AIPLC is applied by adjusting the amplifier or the signal generator. The power
variation by adjusting the amplifier is about ±0,125 dB (shown in Figure V.3 a)), whereas it is
less than ±0,05 dB by adjusting the signal generator (shown in Figure V.3 b)).
IEC/IEEE 62209-1528:2020 – 265 –
© IEC/IEEE 2020
Figure V.2 – The system block diagram of the AIPLC
a) By adjusting the amplifier output b) By adjusting the signal generator output
NOTE Control step-size of signal generator is higher resolution than the amplifier.
Figure V.3 – Power variation characteristics by

adjusting the amplifier or signal generator outputs
– 266 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
Annex W
(informative)
LTE test configurations supporting information
W.1 General
The relationship between RF conducted power and psSAR is investigated using the following
formula, where m indicates a specific LTE mode in a frequency band:
SAR ( m ) = α ( m ) P ( m ) (W.1)
The psSAR is proportional to conducted RF power P. However due to antenna mismatch,

phantom coupling, and other conditions, the ratio between SAR and conducted power,
designated as α, is expected to vary among the LTE modes (m). In order to evaluate the
deviation of α among LTE modes, two independent studies were conducted, as described in
Clause W.2 and Clause W.3.
W.2 Study 1
This study is fully described in [139], [140] and [141]. This study was performed according to
the protocol in IEC 62209-1, and therefore refers to low, middle, and high channels. The
protocol of 7.9.3.6 to commence SAR testing with the LTE mode having the highest conducted
RF power was recommended after the results of Study 1 and Study 2 were combined. The
following is a summary of the test configurations.
• Device type and characteristics: smart phones.

• Frequency and channels tested: band 1 (1 920 MHz to 1 980 MHz); low, middle, and high
channels.
• Phantoms and test positions: SAM and flat.
• Conducted power measurement configurations: Figure W.1.
NOTE Green colour bars in Figure W.1 mean measurements were not performed for these conditions,
because these are the same as those with 20 MHz bandwidth (i.e. largest channel bandwidth used in device).
• Number of DUTs: four different devices from three different manufacturers.

• Number of measurements: 42, for both conducted power and SAR.
The results shown in Figure W.2 are summarized in the following list.
– Deviations: α has a relative standard deviation of less than 11,5 % (Table W.1).
– RF conducted power is the highest without maximum power reduction (MPR) for the
devices with the highest psSAR.
– The deviation of α for the SAM phantom is slightly larger than for the flat phantom. This
might be due to DUT placement uncertainty.
IEC/IEEE 62209-1528:2020 – 267 –
© IEC/IEEE 2020
Table W.1 – Relative standard deviation of α found in Study 1 (without MPR)
a
Coefficient of variation
Phantom (test position) SAR
1g 10 g
Flat 7,90 % 7,90 %
SAM (Cheek) 9,90 % 9,40 %
SAM (Tilt) 11,40 % 10,90 %
a Coefficient of variation (CV) is the standard deviation divided by the average value.
Figure W.1 – Low, middle, and high channels at 2 GHz band (Band 1)
– 268 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
NOTE The three clusters of results along the horizontal axis are due to differences in maximum output power or
MPR associated with the LTE modes tested for the LTE conditions.
Figure W.2 – RF conducted power versus 10 g psSAR
W.3 Study 2
This study is fully described in [33]. The test configurations are as follows.
• Two DUTs.
• Positions: body rear 10 mm, SAM left-side cheek/tilt, SAM right-side cheek/tilt.
• Bands: LTE FDD band 4, LTE FDD band 17.
• Channels:
– Band 4: 19975 (1 712,5 MHz), 20000 (1 715 MHz), 20025 (1 717,5 MHz), 20050
(1 720 MHz), 20175 (1 732,5 MHz), 20300 (1 745 MHz), 20325 (1 747,5 MHz), 20350
(1 750 MHz), 20375 (1 752,5 MHz);
– Band 17: 23755 (706,5 MHz), 23780 (709 MHz), 23790 (710 MHz), 23800 (711 MHz),
23825 (713,5 MHz), 23825 (713,5 MHz).
• Bandwidths:
– Band 4: 5 MHz, 10 MHz, 15 MHz, 20 MHz;
– Band 17: 5 MHz, 10 MHz.
• RB sizes:
– Band 4: 1, 12, 25, 36, 50, 75, 100;
– Band 17: 1, 12, 25, 50.
• RB offsets:
– Band 4: 0, 6, 12, 13, 18, 24, 25, 36, 39, 49, 50, 74, 99;
– Band 17: 0, 6, 12, 24, 25, 49.
• Modulations: QPSK, 16-QAM.
• Number of measurements: 2 510.
IEC/IEEE 62209-1528:2020 – 269 –
© IEC/IEEE 2020
A summary of the measurements including LTE modes, conducted power measurements,
device positions, and test conditions can be found in [33]. Figure W.3 and Table W.2 show the
results.
Table W.2 – Maximum relative standard deviation of α found in Study 2 (with MPR)
Coefficient of variation
Phantom (test position)
1 g SAR
Flat 12,90 %
SAM (Cheek) 17,10 %
SAM (Tilt) 16,80 %
NOTE The three clusters of results along the horizontal axis are due to differences in maximum output power or
MPR associated with the LTE Modes tested for the LTE Conditions. Dashed lines indicate linear regressions.
Figure W.3 – 1 g SAR as a function of RF conducted power in various test conditions
W.4 Justifications of relative standard deviations
The probability density function of the ratio of two normal random variables was derived by
Hinkley in [142]. Assuming the peak spatial-average SAR for a given LTE mode m is a random
normal variable with a maximum of 15 % relative standard deviation (30 % k = 2 uncertainty),
and the conducted power P(m) is a random normal variable with a maximum 5 % relative
standard deviation, then α(m) follows a distribution as described by Hinkley such that:
a) if SAR(m) and P(m) are uncorrelated, then the 95 % confidence interval of α(m) is about
31 % maximum, or ±15,5 % on average;
b) if SAR(m) and P(m) correlate with a coefficient of more than 0,9, then the 95 % confidence
interval of α(m) is about 22 % maximum, or ±11 % on average.
– 270 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
In other words, for the preceding assumption, the k = 2 uncertainty of α(m) determination for a
given LTE mode m can be as high as 15,5 %. In addition, α(m) varies with the mode m, so that
relative standard deviation of α across modes is likely to be higher than 8 %.
To investigate the highest relative standard deviation that can be expected, 100 samples of
α(m) were artificially generated as the output of a ratio of two uncorrelated normal random
variables, SAR(m) and P(m), with respective relative standard deviations of 15 % and 5 %. For
a mode m1, SAR(m1) was chosen with a mean value of 0,66 W/kg and a mean conducted
power P(m1) of 137 mW. For a mode m2, the average of SAR(m2) and P(m2) were chosen
respectively equal to 0,66 W/kg and 124,5 mW. In fact, α(m1) and α(m2) were logically found
to exhibit an average relative difference over the 100 realizations of about 10 %. However, the
maximum relative difference on a specific set of realizations of α(m1) and α(m2) was observed
to be as high as 54 %. This approach is conservative, because the entire uncertainty budget
of the measurement system is not exploited when comparing SAR values across LTE modes
in a given position. In addition, averaging the uncertainty contributions across a variety of
modes decreases the overall observed dispersion of α. Based on this analysis, a 95th
percentile value of 1,35 for α has been determined to be conservative, as applied in 7.9.3.6.3.
IEC/IEEE 62209-1528:2020 – 271 –
© IEC/IEEE 2020
Bibliography
[1] IEC TR 62630:2010, Guidance for evaluating exposure from multiple electromagnetic
sources
[2] Drossos, A., Santomaa, V., and Kuster, N., “The dependence of electromagnetic
energy absorption upon human head tissue composition in the frequency range of 300-
3000 MHz,” IEEE Trans. Microwave Theory Tech., Nov. 2000, vol. 48, no. 11,
pp. 1988–1995
[3] ICNIRP, “International Commission on Non-Ionizing Radiation Protection guidelines for

limiting exposure to time-varying electric, magnetic and electromagnetic fields (Up to
300 GHz),” Health Physics, Vol. 74, No. 4, pp. 494–522, 1998
[4] IEEE Std C95.1-2019, IEEE Standard for Safety Levels with Respect to Human
Exposure to Radio Frequency Electromagnetic Fields, 3 kHz to 300 GHz
[5] Francavilla, M., Schiavoni, A., Bertotto, P., and Richiardi, G., “Effect of the hand on
cellular phone radiation,” IEE Proceedings Microwaves, Antennas and Propagation,
Vol. 148, No. 4, pp. 247–253, Aug. 2001
[6] Gandhi, O. P., Lazzi, G., and Furse, C. M., “Electromagnetic absorption in the human
head and neck for mobile telephones at 835 and 1900 MHz,” IEEE Transactions on
Microwave Theory and Techniques, Vol. 44, No. 10, pp. 1884–1897, Oct. 1996
[7] Jensen, M. A., and Rahmat-Samii, Y., “Performance analysis of antennas for hand-
held transceivers using FDTD,” IEEE Transactions on Antennas and Propagation, Vol.
42, No. 8, pp. 1106–1113, Aug. 1994
[8] Jensen, M. A. and Rahmat-Samii, Y., “EM interaction of handset antenna and a human
in personal communications,” Proceedings of the IEEE, vol. 83, no.1, pp. 7–17, Jan.
1995
[9] Kuster, N., Kästle, R., and Schmid, T., “Dosimetric evaluation of mobile
communications equipment with known precision,” IEICE Trans. Commun., May 1997,
vol. E80-B, no. 5, pp. 645–652
[10] Watanabe, S., Taki, M., Nojima, T., and Fujiwara, O., “Characteristics of the SAR
distribution in a head exposed to electromagnetic fields radiated by a hand-held
portable radio,” IEEE Transactions on Microwave Theory and Techniques, Vol. 44, No.
10, pp. 1874–1883, Oct. 1996
[11] IEEE Std 1528, Recommended Practice for Determining the Peak Spatial-Average
Specific Absorption Rate (SAR) in the Human Head from Wireless Communications
Devices: Measurement Techniques, New York: Institute Electrical and Electronics
Engineers, Sep. 2013
[12] Balzano, Q., Garay, O., and Manning, T., “Electromagnetic energy exposure of the
users of portable cellular telephones,” IEEE Transactions on Vehicular Technology,
Vol. 44, No. 3, pp. 390–403, Aug. 1995
[13] Meyer, F. J. C., Palmer, K. D., and Jakobus, U., “Investigation into the accuracy,
efficiency and applicability of the method of moments as numerical dosimetry tool for
the head and hand of a mobile phone user,” Applied Computational Electromagnetics
Society Journal, Vol. 16, No. 2, pp. 114–125, July 2001
– 272 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
[14] Beard, B. B., et al., “Comparisons of Computed Mobile Phone Induced SAR in the SAM
Phantom to that in Anatomically Correct Models of the Human Head,” IEEE
Transactions on Electromagnetic Compatibility, Vol. 48, No. 2, May 2006
[15] Bit-Babik, G., “Computational comparison of SAR in SAM phantom and anatomically
correct head models at 300 MHz–5.8 GHz: Summary comparison of the results,” report
to ICES TC34 SC1, October 2008
[16] Kainz, W., et al., “Dosimetric comparison of the specific anthropomorphic mannequin
(SAM) to 14 anatomical head models using a novel definition for the mobile phone
positioning,” Phys. Med. Biol., Vol. 50, pp. 3423–3445, July 2005
[17] ISO/IEC 17025:2017, General requirements for the competence of testing and
calibration laboratories
[18] IEC TR 62905:2018, Exposure assessment methods for wireless power transfer
systems
[19] Di Nallo, C., and Faraone, A., “Effect of amplitude modulation of the CDMA IS-95
signal on SAR measurements,” IEEE Transactions on Electromagnetic Compatibility,
vol. 48, no. 3, pp. 552–562, August 2006
[20] Nadakuduti, J., Kühn, S., Fehr, M., Douglas, M., Poković, K., and Kuster, N., "Effect of
Diode Response of Electromagnetic Field Probes for the Measurements of Complex
Signals," IEEE Transactions on Electromagnetic Compatibility, vol. 54, no. 6,
pp. 1195–1204, December, 2012
[21] IEC 106/465/INF, Supporting Information for Amendment 1 to IEC 62209-2 Ed.1
(Document 106/460/CDV), 2018-07-27
[22] Poković, K., “Advanced Electromagnetic Probes for Near Field Evaluation,” Doc. Tech.
Sci. Diss. ETH Nr. 13334. Switzerland, Zurich: Swiss Federal Institute of Technology,
1999
[23] Bolomey, J. C., “Efficient near-field techniques for human exposure evaluation:
Applications to mobile and fixed antennas,” presented at the Electromagnetic
Environment and Human Exposure Evaluation Workshop of EMC, Sorrento, Italy, 2002
[24] Kanda, M. Y., Douglas, M.G., Mendivil, E. D., Ballen, M., Gessner, A. V. Chou, C-K.,
“Faster Determination of Mass-Averaged SAR From 2-D Area Scans,” IEEE
Transactions on Microwave Theory and Techniques, vol. 52, no. 8, pp. 2013–2020,
August, 2004
[25] Manning, M., Massey, P., “Rapid SAR testing of mobile phone prototype using a
spherical test geometry,” in IEE Tech. on Antenna Measurements and SAR Seminar,
Loughborough, U.K., May 28–29, 2002
[26] Merckel, O., Fleury, G., Bolomey, J.-C., “Rapid SAR measurement via parametric
modeling,” Proc. 5th International Congress of the European BioElectromagnetics
Association (EBEA), p. 75–77, Helsinki, Finland, Sep. 2001
[27] Merckel, O., Bolomey, J.-Ch., Joisel, A., “Near-field approach to Rapid SAR
Measurement of Mobile Phones,” Symp. of the Association for Measurement and
Testing of Antennas (AMTA 2003), Irvine, Denver, USA, Oct. 2003
IEC/IEEE 62209-1528:2020 – 273 –
© IEC/IEEE 2020
[28] Douglas, M.G., Kanda, M.Y., Luengas, W.G., Ballen, M., Babij, T.M. and Chou, C-K.,
“An Algorithm for Predicting the Change in SAR in a Human Phantom due to
Deviations in its Complex Permittivity,” IEEE Transactions on Electromagnetic
Compatibility, 2007, vol. 51, no 2, May 2009
[29] Douglas, M.G., and Chou, C-K., “Enabling the Use of Broadband Tissue Equivalent
Liquids for Specific Absorption Rate Measurements,” IEEE Electromagnetic
Compatibility Symposium, July 2007
[30] Kuster, N. and Balzano, Q., “Energy absorption mechanism by biological bodies in the
near field of dipole antennas above 300 MHz,” IEEE Transactions on Vehicular
Technology, Vol. 41, No. 1, pp. 17–23, Feb. 1992
[31] Bit-Babik, G., Faraone, A., Ballen, M. and Chou, C-K., “Sensitivity of the Spatial-
Average Peak SAR to the Dielectric properties of Media Used for Compliance Testing
in the Frequency Range 0.3 – 3 GHz,” Antennas and Propagation Society International
Symposium Digest, Vol. 3, pp. 722–725, June 2002
[32] IEC 62479:2010, Assessment of the compliance of low-power electronic and electrical
equipment with the basic restrictions related to human exposure to electromagnetic
fields (10 MHz to 300 GHz)
[33] Derat, B., “Experimental Study on the Relationship between Specific Absorption Rate
and RF Conducted Power for LTE Wireless Devices,” Proc. European Microwave
Conf., Paris, France, Sep. 2015
[34] IEC 62311:2019, Assessment of electronic and electrical equipment related to human
exposure restrictions for electromagnetic fields (0 Hz to 300 GHz)
[35] ISO/IEC Guide 98-1:2009, Uncertainty of measurement – Part 1: Introduction to the

expression of uncertainty in measurement
[36] Onishi, T. and Uebayashi S., “Influence of phantom shell on SAR measurement in 3-6
GHz frequency range,” IEICE Trans. Commun., vol., E88-B, no. 8, pp. 3257–3262,
2005
[37] Niskala, K., “Multimode SAR test reduction,” Final report, IEC MT1 meeting, Tokyo-
Japan, May 2014
[38] Schiavoni, A., Francavilla, M., Forigo, D., Deplano, M., “A Procedure to Reduce the
Qualification Time in SAR Compliance Tests,” BEMS 2005, June 2005
[39] Francavilla, M., “Time reduction in SAR compliance of GSM/UMTS mobile phones,”
BEMS 2011, Halifax, Canada, June 2011
[40] Montgomery, D.C., Design and Analysis of Experiments (4th edition), New York: John
Wiley and Sons. (1997)
[41] IEC/IEEE 62704-1:2017, Determining the peak spatial-average specific absorption rate
(SAR) in the human body from wireless communications devices, 30 MHz to 6 GHz –
Part 1: General requirements for using the finite-difference time-domain (FDTD)
method for SAR calculations
[42] IEC 60154-2, Flanges for waveguides – Part 2: Relevant specifications for flanges for
ordinary rectangular waveguides
– 274 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
[43] IEEE Std C95.3-2002, IEEE Recommended Practice for Measurements and
Computations of Radio Frequency Electromagnetic Fields With Respect to Human
Exposure to Such Fields, 100 kHz–300 GHz
[44] Loader, B., “Computer Simulation of WR159 Waveguide against a Flat Dielectric
Phantom at 5.2 GHz and 5.8 GHz,” NPL Report DEM EM 008, 2007
[45] Choi, D.-G., Kim, K.-H., Choi, J.-H., “New Numerical Target SAR Values and an
Optimized Flat Phantom for SAR Validation Test in the 150 MHz Band,” IEICE TRANS.
COMMUN., Vol. E98-B, No. 7, July 2015
[46] Kim, K.-H., Jang, J.-D., Chung, S.-Y., and Gimm, Y.-M., “A Study on the Meander-Type
Dipole Antenna for SAR Validation Test at 150 MHz Band,” Proceedings of 2016 URSI
AP-RASC (Asia-Pacific Radio Science Conference), pp. 911–913, August 21–25, 2016,
Grand Hilton Seoul Hotel, Seoul, Korea
[47] Kim, K.-H., Choi. D.-G., Gimm, Y.-M., “Automatic RF Input Power Level Control
Methodology for SAR Measurement Validation,” Journal of Electromagnetic
Engineering and Science, Vol. 15, No. 3, pp. 181–184, Jul. 2015
[48] Meier, K., Burkhardt, M., Schmid, T., and Kuster, N., “Broadband calibration of E-field
probes in lossy media,” IEEE Trans. Microwave Theory Tech., Oct. 1996, vol. 44, no.
10, pp. 1954-1962
[49] Person, C., Tanné, G., Ahlonsou, L.N, Ngounou, C., and Grangeat, C., “New reference
antennas for SAR probe calibration in tissue equivalent liquid,” Millennium Conference
on Antennas and Propagation, AP2000, Davos, Switzerland: April 9–14, 2000
[50] Kuster, N., and Balzano, Q., “Experimental and numerical dosimetry,” in Kuster, N., Q.
Balzano, and J. C. Lin, eds., Mobile Communications Safety, New York: Chapman &
Hall, pp. 13–64, 1997
[51] Jokela, K., Hyysalo, P., and Puranen, L., “Calibration of specific absorption rate (SAR)
probes in waveguide at 900 MHz,” IEEE Trans. Instrumen. Meas., Apr. 1998, vol. 47,
no. 2, pp. 432-438
[52] Magee, J. W., “Molar heat capacity (Cv) for saturated and compressed liquid and vapor
nitrogen from 65 to 300 K at pressures to 35 MPa,” Journal of Research of the National
Institute of Standards and Technology, Vol. 96, No. 6, pp. 725–740, Nov./Dec. 1991
[53] Poković, K., Schmid, T., and Kuster, N., “Robust setup for precise calibration of E-field
probes in tissue simulating liquids at mobile communication frequencies,” in
Proceedings ICECOM ’97. Dubrovnik, Croatia, Oct. 12–17, 1997, pp. 120–124
[54] EIA RS-261-B: 1979, Rectangular Waveguides (WR3 to WR2300)
[55] Person, C., Ahlonsou, L.N, and Grangeat, C., “New test bench for the characterization
of SAR measurement probes used in tissue equivalent liquids,” Bioelectromagnetics
Society 22 nd Annual Meeting, Munich Germany, June 11–16, 2000
[56] Ishii, N., Sato, K., Hamada, L., Watanabe, S., “Gain calibration in near-field region of
antenna in tissue-equivalent liquid for SAR assessment,” 2008 Asia-Pacific Symposium
on Electromagnetic Compatibility, Singapore, pp.112–115, May, 2008
[57] Loader, B. G, Gregory, A.P., and Bownds, D., “Coaxial artefact standard for specific
absorption rate 100 kHz to 400 MHz,” Progress in Electromagnetic Research
Symposium (PIERS), 23–27 March, 2009, Beijing, China
IEC/IEEE 62209-1528:2020 – 275 –
© IEC/IEEE 2020
[58] Faraone, A., McCoy, D.O., Chou, C.K., and Balzano, Q., “Characterization of
miniaturized E-field probes for SAR measurements,” IEEE Intl. Symp. Electromag.
Compat., Washington, DC, 2000, pp. 749–754
[59] Fukunaga, K., Watanabe, S., Wake, K. and Yamanaka, Y., "Time dependence of
tissue-equivalent dielectric liquid materials and its effect on SAR," EMC Europe Symp.,
Sorrento, Italy, Sep. 2002
[60] Loader, B. G., Gregory A.P., Mouthaan, R., “Formulation and properties of liquid
phantoms 1 MHz to 10 GHz,” NPL Report TQE 9, June 2017
[61] Gordon, C.C., Churchill, T., Clauser, C.E., Bradtmiller, B., McConville, J.T., Tebbetts,
I., and Walker, R.A., “1988 Anthropometric Survey of U.S. Army Personnel: Methods
and Summary Statistics,” Technical Report NATICK/TR-89/044, U.S. Army Natick
Research, Development and Engineering Center, Massachusetts: Natick, Sep. 1989
[62] Lee, A.K., and Pack, J.K., “Effect of head size for cellular telephone exposure on EM
absorption,” IEICE Trans. Commun., Mar. 2002, vol. E85-B, no. 3, pp. 698–701
[63] Lee, A.K., Choi, H.D., Lee, H.S., and Pack, J.K., “Human head size and SAR
characteristics for handset exposure,” ETRI J., Apr. 2002, vol. 24, no. 2, pp. 176–179
[64] Leisten, O., Vardaxaglou, Y., Schmid, T., Rosenberger, B., Agboraw, E., Kuster, N.
and Nicolaidis, G., “Miniature dielectric-loaded personal telephone antennas with low
user exposure,” Electron. Lett., Aug. 20, 1998, vol. 34, no. 17, pp. 1628–1629
[65] Schönborn, F., Burkhardt, M., and Kuster, N., “The difference of EM energy absorption
between adults and children,” Health Phys., Feb. 1998, vol. 74, no. 2, pp. 160–168
[66] Keshvari, J., Kivento, M., Christ, A., and Bit-Babik, G., “Large scale study on the
variation of RF energy absorption in the head & brain regions of adults and children
and evaluation of the SAM phantom conservativeness,” Physics in Medicine &
Biology, 61(8), 2991, 2016
[67] Hirata, A., Ito, F., and Laakso, I., "Confirmation of quasi-static approximation in SAR
evaluation for a wireless power transfer system," Physics in Medicine and Biology, vol.
58, p. N241, 2013
[68] Laakso, I., Shimamoto, T., Hirata, A., and Feliziani, M., "Quasistatic Approximation for
Exposure Assessment of Wireless Power Transfer," IEICE Transactions on
Communications, vol. E98.B, pp. 1156–1163, 2015
[69] Gabriel, S., Lau, R.W., and Gabriel, C., “The dielectric properties of biological tissues:
3. Parametric models for the dielectric spectrum of tissues,” Phys. Med. Bio., 1996,
vol. 41, no. 11, pp. 2271–2293
[70] Christ, A., Klingenböck, A., Samaras, T., Goiceanu, C., and Kuster, N., “The
dependence of electromagnetic far-field absorption on body tissue composition in the
frequency range from 300 MHz to 6 GHz,” IEEE Transactions on Microwave Theory
and Techniques, vol. 54, no. 5, pp. 2188–2195, May 2006
[71] Christ, A., Samaras, T., Klingenböck, A., and Kuster, N., “Characterization of the
electromagnetic near-field absorption in layered biological tissue in the frequency
range from 30 MHz to 6000 MHz,” Physics in Medicine and Biology, vol. 51, no. 19,
October 2006
– 276 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
[72] Kangas, I., and Hyttinen, J., “Thermal Simulations for Anatomical Human Models, Part
III of Dielectric properties for Tissue Simulating Liquids Study: Extension and
Validation Covering a Frequency Range from 30 MHz to 6 GHz,” Technical Report,
Tampere University of Technology, April 22, 2016
[73] Christ, A., et al. “The Virtual Family—development of surface-based anatomical

models of two adults and two children for dosimetric simulations,” Physics in Medicine
& Biology, vol. 55, no. 2, p. N23, 2009
[74] Poković, K., Christ, A., Samaras, T., Douglas, M., Kuster, N., “Methods and
Instrumentation for Reliable Experimental SAR Assessment at 6 – 10 GHz,” The joint
annual meeting of the Bioelectromagnetics Society and the Eurpean
BioElectromagnetics Association, Hangzhou, China, June 5–9, 2017
[75] Chou, C.-K., Chen, G.W., Guy, A.W. and Luk, K.H., “Formulas for preparing phantom
muscle tissue at various radiofrequencies,” Bioelectromag., 1984, vol.5, pp. 435–441
[76] Von Hippel, A., Dielectric Materials and Applications. Cambridge: MA: MIT Press, 1954
[77] Blackham, D.V., and Pollard, R.D., “An improved technique for permittivity
measurements using a coaxial probe,” IEEE Trans. Instrumen. Meas., Oct. 1997, vol.
46, no. 5, pp. 1093–1099
[78] Gabriel, C., Chan, T.Y.A, and Grant, E.H., “Admittance models for open ended coaxial
probes and their place in dielectric spectroscopy,” Phys. Med. Biol., 1994. vol. 39,
no.12, pp. 2183–2200
[79] Pournaropoulos, C.L., and Misra, D.K., “The co-axial aperture electromagnetic sensor
and its application in material characterization,” Meas. Sci. Tech., 1997, vol. 8,
pp. 1191–1202
[80] Arai, M.J., Binner, G.P., and Cross, T.E., “Estimating errors due to sample surface
roughness in microwave complex permittivity measurements obtained using a coaxial
probe,” Electron. Lett., Jan. 19, 1995, vol. 32, no. 2, pp. 115–117
[81] Bao, J.Z, Swicord, M.L., and Davis, C.C., “Microwave dielectric characterization of
binary mixtures of water, methanol, and ethanol,” J. Chem. Phys., Mar. 12, 1996,
vol. 104, no. 12, pp. 4441–4450
[82] Clarke, R.N., Gregory, A.P., Hodgetts, T.E., and Symm, G.T., “Improvements in coaxial
sensor dielectric measurement: relevance to aqueous dielectrics and biological tissue,”
in Microwave Aquametry: Electromagnetic Wave Interaction With Water-containing
Materials, A. Kraszewski, ed., New York: IEEE Press, 1996, pp. 279–297
[83] Mosig, J.R., Besson, J.C.E, Gex-Fabry, M., and Gardiol, F.E., “Reflection of an open-
ended coaxial line and application to non-destructive measurement of materials,” IEEE
Trans. Instrumen. Meas., 1981, IM-30, pp 46–51
[84] Nevels, R.D., Butler, C.M., and Yablon, W., “The annular slot antenna in a lossy
biological medium,” IEEE Trans. Microwave Theory Tech., 1985, MTT-30, pp 314–319
[85] Toropainen, A., Vainikainen, P., and Drossos, A., “Method for accurate measurement
of complex permittivity of tissue equivalent liquids,” Electron. Lett., 2000, vol. 36, no.
1, pp. 32–34
[86] Hill, N.E., Vaughan, W.E., Price, A.H., and Davies, M., Dielectric Properties and
Molecular Behaviour, London: Van Nostrand Reinhold, 1969
IEC/IEEE 62209-1528:2020 – 277 –
© IEC/IEEE 2020
[87] Gabriel, S., Lau, R. W., and Gabriel, C., “The dielectric properties of biological tissues:
II. Measurement in the frequency range 10 Hz to 20 GHz,” Physics in Medicine and
Biology, Vol. 41, No. 11, pp. 2251–2269, 1996
[88] Kaatze, U., “Complex permittivity of water as function of frequency and temperature,”
J. Chem. Engin. Data, 1989, vol. 34, no. 4, pp. 371–374
[89] Gregory, A.P. and Clarke, R.N., “Tables of the Complex Permittivity of Dielectric
Reference Liquids at Frequencies up to 5 GHz,” NPL Report MAT 23, Materials
Division, Teddington, England: National Physical Laboratory, Revised Jan 2012, ISSN
1754-2979
[90] Rocher, F., Julien, T., Derat, B., "Impact of a Hand Phantom on the Measured Average
SAR," presentation to ICES TC-34 committee, December 11, 2008
[91] Bit-Babik, G., Summary and Review of the SAM Study Phase III, April 2012
[92] Derat, B., Gabriel, S., Faraone, A., “Hand effect measurement interlab: final report,”
presentation to IEC MT 62209, August 2011
[93] Derat, B., Gabriel, S., Faraone, A., “Hand effect measurement interlab: Conclusions of
the study,” presentation to IEC MT 62209, Dec 2011
[94] Christ, A., “Analysis of the exposure of the hand and its impact on the absorption in the
heads of cell phone users,” presentation to IEC MT 62209, January 2013
[95] Kuster, N., Douglas, M., "Proposal for inclusion of the hand for IEC 62209," IEC MT
62209 meeting, Thessaloniki Greece, June 2013
[96] Douglas, M., Bucher, C., Ofli, E., Derat, B., Gabriel, S., Kuster, N., "Investigation of the
influence of the Hand on Head SAR," IEC MT 62209 meeting, Newbury UK, March
2010
[97] Douglas, M., Kuster, N., "Summary of two studies of hand effect on SAR in head," IEC
MT 62209 meeting, Brisbane Australia, June 17, 2012
[98] Li, C-H., Douglas, M., Ofli, E. Derat, B. and Kuster, N., "Investigation of the influence
of the Hand on Head SAR," report to IEC PT 62209 MT-1, April, 2010
[99] Nesterova, M., Review of Head-Hand Studies between 1997-1999, April 2012
[100] Kivento, M., Keshvari, J., “Hand effect during the use of real mobile phones,” Jan 2013
[101] Park, G. B., Joyner, K., Jang, J. D. ”Modified DUT Holder for SAR Measurement,”
IEC 62209 MT1 Meeting Turin, 11–13 Sept. 2012
[102] Joyner, K., Jang, J.D., Park, G.B., Park, Y. H., “2nd Modified DUT Holder for SAR
measurement,” IEC 62209 MT1 Meeting, Ft Lauderdale, 22–24 Jan. 2013
[103] Joyner K., “The Case to Exclude the Hand,” IEC 62209 MT1 Meeting, Newbury, 7–9
May 2013
[104] Monebhurrun, V., Wong, M. F., Gati, A., Wiart, J., “Study of the influence of the hand
on the specific absorption rate evaluation of mobile phones,” in Proc. URSI, Istanbul,
Aug. 2011
– 278 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
[105] Monebhurrun, V., Wong, M. F., Gati, A., Wiart, J., “Numerical and experimental
investigations of the influence of the hand on the specific absorption rate evaluations
of mobile phones,” in Proc. 33rd Annual Meeting of the Bioelectromagnetics Society,
Halifax, June 2011
[106] Monebhurrun, V., “Influence of the hand on the specific absorption rate assessments
of mobile phones,” Microwave Optical Technology Letters, 54, 3, 2012, pp. 654–656
[107] Monebhurrun, V., “Study of the influence of a CTIA hand phantom on the specific
absorption rate measurements of mobile phones,” in Proc. BioEM 2013, Thessaloniki,
June 2013
[108] Francavilla, M. and Schiavoni, A., "Effect of the Hand in SAR Compliance Tests of
Body Worn Devices," Applied Computational Electromagnetic Society Conference,
Verona, Italy, March 21, 2007
[109] Li, C.H., Douglas. M., Ofli, E., Chavannes, N., Balzano, Q., Kuster, N., "Mechanisms of
RF Electromagnetic Field Absorption in Human Hands and Fingers," IEEE
Transactions on Microwave Theory and Techniques, vol. 60, no. 7, pp. 2267–2276,
July 2012
[110] Li, C.H., Douglas, M., Ofli, E., Derat, B., Gabriel, S., Chavannes, N., Kuster, N.,
"Influence of the Hand on the Specific Absorption Rate in the Head from a Mobile
Phone," IEEE Transactions on Antennas and Propagation, vol. 60, no. 2, pp. 1066–
1074, February 2012
[111] Li, C.-H., Ofli, E., Chavannes, N., and Kuster, N., “Effects of hand phantom on mobile
phone antenna performance,” IEEE Transactions on Antennas and Propagation, vol.
57, no. 9, pp. 2763–2770, Sep. 2009
[112] CTIA, “CTIA test plan for wireless device over-the-air performance, revision 3.7.1,”
CTIA Wireless Association, Feb. 2018
[113] Greiner, T.M., “Hand Anthropometry of US Army Personal,” Army Natick Research
Development and Engineering Center, Technical Report Natick/TR-92/011, Dec. 1991
[114] Tilley, A.R., and Henry Dreyfuss Associates, “The Measure of Man and Woman:
Human Factors in Design,” Wiley, Dec. 31, 2001
[115] Buchholz, B., Armstrong, T.J. and Goldstein, S.A., “Anthropometric data for describing
the kinematics of the human hand,” Ergonomics vol. 35, no. 3, pp. 261–273, 1992
[116] Bugbee, W.D., and Botte, M.J, “Surface Anatomy of the Hand: The Relationships
Between Palmar Skin Creases and Osseous Anatomy,” Clinical Orthopaedics and
Related Research, 296, 122–126, 1993
[117] Derat, B., "Impact of the hand on the averaged SAR in the head: simulation of a CAD
phone model," presentation to IEC PT62209 MT-1, April 22, 2009
[118] Pelosi, M., Franek, O., Knudsen, M. B., Christensen, M., and Pedersen, G. F., “A grip
study for talk and data modes in mobile phones,” IEEE Transactions on Antennas and
Propagation, vol. 57, no. 4, pp. 856–865, Apr. 2009
[119] Pelosi, M., Franek, O., Pedersen, G.F., Knudsen, M., "User's Impact on PIFA Antennas
in Mobile Phones," IEEE 69th Vehicular Technology Conference, Barcelona,
Spain, 26–29 April 2009
IEC/IEEE 62209-1528:2020 – 279 –
© IEC/IEEE 2020
[120] Wu, T., Lin, X., Yang, J., Zhao, C., Zhang, C. and Shao, Q., "TD-SCDMA Mobile
Phone’s SAR Measurement about OTA Hand Phantom," Progress in Electromagnetics
Research Symposium, Xi’an, China, March 22–26, 2010
[121] Pradier, A., Colas, O., Celin, P., Sarrebourse, T., Laudru, D., Wong, M.-F., Fouad
Hanna, V., Wiart, J. "Evaluation of the Specific Absorption Rate Induced by the
Handset Close to the Body," in BEMS 28th Annual Meeting, 2006
[122] Diller, K.R., Adapting adult scald safety standards to children (2006) J Burn Care Res
27: 314–22
[123] Gemperle, F., Kasabach, C., Stivoric, J., Bauer, M., and Martin, R.,
Design for Wearability, Techn. Rep. [online]. Carnegie Mellon University, PA (USA)
[viewed 2019-07-16], Available at https://www.cs.cmu.edu/~15-
821/READINGS/PAPERS/gemperle1998.pdf
[124] Douglas, M., Bit-Babik, G., Nadakuduti, J., Faraone, A., and Chou, C-K., “Modeling of
SAR in the User for Body-Worn Wireless communication devices, ” Proc. 29 th Annual
Meeting of the Bioelectromagnetics Society (BEMS 2007), Kanazawa, Japan, June 11–
15, 2007
[125] Taylor, J.R., An Introduction to Error Analysis: The Study of Uncertainties In Physical
Measurements. 2nd ed. Sausalito, CA: University Science Books, 1997
[126] Evans, S., and Michelson, S.C., “Intercomparison of dielectric reference materials
available for the calibration of an open-ended probe at different temperatures,” Meas.
Sci. Tech., Dec. 1995, vol. 6, no. 12, pp. 1721–1732
[127] Jenkins, S., Hodgetts, T.E., Clarke, R.N., and Preece, A.W., “Dielectric measurements
on reference liquids using automatic network analysers and calculable geometries,”
Meas. Sci. Tech., July 1990, vol. 1, no. 7, pp. 691–702
[128] Migliore, M.D., “Partial self-calibration method for permittivity measurement using
truncated coaxial cable,” Electron. Lett., July 20, 2000, vol. 36, no. 15, pp. 1275–1277
[129] Nyshadam, A., Sibbald, C.L, and Stuchly, S.S., “Permittivity measurements using
open-ended sensors and reference liquid calibration – an uncertainty analysis,” IEEE
Trans. Microwave Theory Tech., Feb. 1992, vol. 40, no. 2, pp. 305–314
[130] Press, W.H., Flannery, B.P., Teukolsky, S.A., and Vetterling, W.T., Numerical Recipes
in FORTRAN 77: The Art of Scientific Computing, New York: Cambridge University
Press, 1992
[131] Fieguth, P.W., Karl, W.C., Willsky, A.S., and Wunsch, C., “Multi-resolution optimal
interpolation and statistical analysis of TOPEX/POSEIDON satellite altimetry,” IEEE
Trans. Geosci. Remote Sens., Mar. 1995, vol. 33, pp. 280–292
[132] Lancaster, P., and Salkauska, K., Curve and Surface Fitting: An Introduction. New
York: Academic Press, 1986
[133] Ferreira, P.J.S.G., “Non-iterative and fast iterative methods for interpolation and
extrapolation,” IEEE Trans. Sig. Proc., Nov 1994, vol. 41, pp. 3278–3282
[134] Ford, C., and Etter, D.M., “Wavelet basis reconstruction of nonuniformly sampled
data,” IEEE Trans. Circuits Sys. II: Analog Dig. Sig. Proc., Aug. 1998, vol. 45, no. 8,
pp. 1165–1168
– 280 – IEC/IEEE 62209-1528:2020
© IEC/IEEE 2020
[135] Ustuner, K.F., and Ferrai, L.A., “Discrete splines and spline filters,” IEEE Trans.
Circuits Sys., July 1991, vol. 39, no. 7, pp 417–422
[136] ISO 10012:2003, Measurement management systems – Requirements for

measurement processes and measuring equipment
[137] EA-2/07 (rev.01), EAL Strategy to Achieve Comparability of Results in Calibration and
Testing, Mar. 1997
[138] ISO/IEC 17043:2010, Conformity assessment – General requirements for proficiency

testing
[139] Onishi, T., et. al, “LTE SAR measurement consideration,” IEC TC106 MT1 meeting in
Turin, Sept. 2012
[140] Onishi, T., “Considerations on SAR measurement for recent trend of the usages and
new technologies,” 7 th European Conference on Antennas and Propagation (EuCAP)
2013, March 2013
[141] Onishi, T., “Investigation on SAR measurement procedure using SAM and Flat
phantoms for LTE handsets,” IEC TC106 MT1 meeting in Ottawa, Sept. 2013
[142] Hinkley, D.V., “On the ratio of two correlated normal random variables,” Biometrika,
Vol. 56, No. 3, pp. 635–639, Dec. 1969
_____________
INTERNATIONAL
ELECTROTECHNICAL
COMMISSION
3, rue de Varembé
PO Box 131
CH-1211 Geneva 20
Switzerland
Tel: + 41 22 919 02 11
[email protected]
www.iec.ch

Iec - Ieee 62209-1528-2020

Uploaded by

Copyright:

Available Formats

Iec - Ieee 62209-1528-2020

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Iec - Ieee 62209-1528-2020

Uploaded by

Copyright:

Available Formats

IEC/IEEE 62209-1528

Measurement procedure for the assessment of specific absorption rate of

IEC Central Office Institute of Electrical and Electronics Engineers, Inc.

About the IEC

About the IEEE

IEC publications search - webstore.iec.ch/advsearchform Electropedia - www.electropedia.org

Measurement procedure for the assessment of specific absorption rate of

ICS 17.220.20 ,(&ISBN 978-2-8322-8533-6

® Registered trademark of the International Electrotechnical Commission

Figure 1 – Quick start guide .................................................................................................. 29

Table 1 – Evaluation plan checklist ....................................................................................... 28

MEASUREMENT PROCEDURE FOR THE ASSESSMENT OF SPECIFIC

Part 1528: Human models, instrumentation, and procedures

a) extension of the frequency range down to 4 MHz and up to 10 GHz;

This publication is published as an IEC/IEEE Dual Logo standard.

The text of this standard is based on the following IEC documents:

FDIS Report on voting

– measurement system requirements (Clause 6),

Part 1528: Human models, instrumentation, and procedures

This document does not apply to implanted medical devices.

ISO/IEC Guide 98-3:2008, Uncertainty of measurement – Part 3: Guide to the expression of

3 Terms and definitions

• IEC Electropedia: available at http://www.electropedia.org/

EXAMPLE 5 Batteries and related DC power components.

Note 1 to entry: Details are specified in IEC TR 62630 [1].

Note 1 to entry: This definition excludes hand-held devices.

σ is the electrical conductivity of the lossy medium, in siemens per metre.

EXAMPLE Two-way radio.

4 Symbols and abbreviated terms

4.1 Physical quantities

The International System of Units (SI) is used throughout this document.

Symbol Quantity Unit Dimensions

P Average (temporal) absorbed power watt W

J Current density ampere per square metre A/m 2

σ Electric conductivity siemens per metre S/m

E Electric field strength volt per metre V/m

ρ Mass density kilogram per cubic metre kg/m 3

δ Penetration depth metre m

µ Permeability henry per metre H/m

joule per kilogram per

ε Permittivity farad per metre F/m

SAR Specific absorption rate watt per kilogram W/kg

Table 1 – Evaluation plan checklist

Stage Evaluation steps Action

Preliminary – A liquid dielectric properties measurement shall be performed Should be conducted

Figure 1 – Quick start guide

6.1 General requirements for full SAR testing

6.2 Phantom specifications

6.2.2 Basic phantom parameters

• εr′ ≤ 5, for f ≤ 3 GHz

Table 2 – Dielectric properties of the tissue-equivalent medium

Real part of the

6.2.3 Head phantom

a) The psSAR shall provide conservative exposure.

Figure 2 – Dimensions of the elliptical phantom

6.2.5 Device-specific phantoms

6.3 Influence of hand on SAR in head

6.4 Scanning system requirements

6.5 Device holder specifications