A Wideband Monopole Antenna Design

A Wideband Monopole Antenna Design
by
Jako Lourens
Thesis presented in partial fulfilment of the requirements for the degree

Master of Science in Engineering
at Stellenbosch University
Supervisor: Prof. KD Palmer

Department of Electrical & Electronic Engineering
March 2013
Stellenbosch University http://scholar.sun.ac.za
Declaration
By submitting this thesis I declare that the entirety of the work contained therein is my own,
original work, that I am the sole author thereof (save to the extent explicitly otherwise
stated), that reproduction and publication thereof by Stellenbosch University will not infringe
any third party rights and that I have not previously in its entirety or in part submitted it for
obtaining any qualification.
Date: March 2013
Copyright © 2013 Stellenbosch University
All rights reserved
2
Acknowledgements
I would like to thank Electromagnetic Software and Systems (EMSS) for donating FEKO and
allowing it to be used on this antenna design. Without it‟s use, complex mathematics would
have to be performed manually – a nigh impossible task.
3
Abstract
The successful operation of a man-pack VHF jamming system requires a compact and
efficient antenna operating over a wide bandwidth. The design of such an antenna is the
focus of this thesis.
The antenna should be of a practical size for a portable system and it must radiate energy
efficiently across a frequency bandwidth in excess of a decade. A practical „target‟
specification of such an antenna has been drawn up based on the performance of a
commercially available system.
Several possible antenna topologies, each with a variety of loading section options, are
tested using “Full wave” electromagnetic modelling (FEKO). Each topology/loading-section is
numerically optimised for load element values by considering both its gain and reflection
coefficient. Results of the „optimally loaded‟ solution for each topology are then compared to
each other to arrive at the best overall design.
The best result is found to be the traditional monopole whip-type antenna, with four R-L
loading sections spread along its length. The simulated results show that the proposed
antenna can be expected to meet the target standing wave ratio (SWR) specifications while
offering a gain advantage of between 5 and 10 dBi higher than is available commercially.
The selected design is constructed and its performance measured.
Opsomming
Die suksesvolle werking van ŉ mobiele VHF „jammer‟ benodig ŉ kompakte antenna met ŉ
bruikbare benuttingsgraad wat oor ŉ wyeband funksioneer. Die ontwerp van so ŉ antenna is
die fokus van hierdie tesis.
Die antenna moet kompak genoeg wees om draagbaar te wees en moet ŉ bruikbare
benuttingsgraad hê oo r ŉ frekwensie-bandwydte van meer as 10:1. ŉ Praktiese spesifikasie
is opgestel vir die antenna deur te kyk na die sigblaaie van beskikbare stelsels.
“Volgolf” elektromagnetiese modelleringsagteware is daarna gebruik om ŉ parametriese
ondersoek te loods van verskillende antennas.
Verskillende topologieë is getoets met ʼn verskeidenheid van belaaide seksies waar die
topologieë ge-optimaliseer was vir wins en weerkaatskoëffisiënt. Die resultate vir elke
optimale oplossing is vergelyk.Opgrond van hierdie resultate is bevind dat die beste
topologie die tradisionele monopoolmas „whip-type‟ antenna is met vier RL lading afdelings
langs die lengte versprei. Analise word gebruik om te wys dat verwag kan word dat dit aan
die aanwins en staande golf verhouding (SGV) spesifikasies sal voldoen met n 10 dB
verhoging in aanwins vir n laer SGV.
Die geselekteerde ontwerp is gebou en gemeet om te verifieer dat dit aan die spesifikasies
voldoen.
4
Table of Contents
ABSTRACT 4
OPSOMMING 4
INTRODUCTION 8
1.1 BACKGROUND 8
1.2 THESIS OUTLINE 10
LITERATURE STUDY 11
2.1 MONOPOLE ANTENNA LOADED WITH NON-FOSTER CIRCUIT 11
2.2 THE SHORT LOADED MONOPOLE ANALYSED BY HANSEN, STEWART AND HARRISON 15
2.3 FREQUENCY ENHANCED MONOPOLE DESIGN 16
2.4 EFFECT OF GROUND-PLANE ON ANTENNA PERFORMANCE 16
2.5 REALISED GAIN: 17
2.6 CONCLUSION FROM LITERATURE 18
SIMULATIONS AND ANALYSIS OF ANTENNA DESIGNS 19
3.1 CONSTRUCTION OF INDUCTORS 23
3.2 TOPOLOGIES CONTAINING BEAD INDUCTORS 24
3.2.1 Monopole antenna with RL beads at fixed and various locations 24
3.2.2 Top-loaded antenna with RL bead and RL chip sections at various locations 26
3.3 TOPOLOGIES CONTAINING CHIP INDUCTORS 27
3.3.1 Top-loaded antenna with RL chip sections at various locations 27
3.3.2 Monopole antenna with a helix structure and RL chip sections at fixed locations 29
3.3.3 Monopole antenna with a folded structure and RL chip sections at fixed locations 31
3.3.4 Monopole antenna with RL chip sections at fixed locations 32
3.4 FINAL MONOPOLE ANTENNA CONTAINING CHIP INDUCTORS 34
3.4.1 Choice of inductors 34
3.4.2 Inductor’s frequency characteristics added to antenna model 36
3.4.3 Final antenna design compared to reference antenna 39
3.4.4 The influence of a human body on the antenna 40
4.1 PRACTICAL MEASUREMENTS 42

4.2 FINAL ANTENNA CONFIGURATION 46
4.3 MEASUREMENT EQUIPMENT 47
4.4 MEASUREMENTS PERFORMED 47
4.4.1 The reflection coefficient (S₁₁) 47
4.4.2 The forward transmission coefficient (S₂₁) - near-field 54
4.4.3 Power handling 59
CONCLUSION 60
BIBLIOGRAPHY 62
APPENDIX A: PHOTOGRAPHS OF MEASUREMENTS AND SCHEMATICS 63
APPENDIX B: RESULTS OF SWR MEASUREMENTS 66
5
List of Figures
FIGURE 1: GMJ9000V PORTABLE JAMMER WITH REFERENCE ANTENNA [8].................................................................... 8

FIGURE 2: MEASURED OMNI-A0124-01 ANTENNA GAIN FROM LITERATURE [8]............................................................ 9
FIGURE 3: MEASURED OMNI-A0124-01 ANTENNA VSWR FROM LITERATURE [8] ......................................................... 9
FIGURE 4: NON-FOSTER CIRCUIT SCHEMATIC ............................................................................................................ 12
FIGURE 5: REACTIVE VALUE REQUIRED FOR AN ANTENNA LENGTH OF Λ/4 TO ACHIEVE RESONANCE AT A FREQUENCY [1] ......... 12
FIGURE 6: THE VALUE OF THE INPUT REACTANCE WITH AND WITHOUT NON-FOSTER CIRCUIT LOADING [1] ........................... 13
FIGURE 7: THE VALUE OF THE INPUT RESISTANCE WITH AND WITHOUT NON-FOSTER CIRCUIT LOADING [1] ........................... 13
FIGURE 8: THE VALUES OF THE VSWR WITH AND WITHOUT NON-FOSTER CIRCUIT LOADING [1] ........................................ 14
FIGURE 9: RADIATION PATTERN OF LOADED (LEFT) AND UNLOADED (RIGHT) MONOPOLE ANTENNAS. [1] ............................. 14
FIGURE 10: DESIGN TOPOLOGIES INVESTIGATED ....................................................................................................... 22
FIGURE 11: RL CHIP SECTION ............................................................................................................................... 24
FIGURE 12: RL BEAD SECTION .............................................................................................................................. 24
FIGURE 13: RL BEAD AND CHIP DIAGRAM .............................................................................................................. 24
FIGURE 14: MONOPOLE ANTENNA WITH RL BEAD SECTIONS ...................................................................................... 25
FIGURE 15: THE GAIN AND SWR OF MONOPOLE ANTENNA WITH FOUR, FIVE, SEVEN AND NINE RL BEADS AT FIXED LOCATIONS
(FIX STEP) AND VARIOUS LOCATIONS (VAR STEP) ...................................................................................................... 25
FIGURE 16: TOP-LOADED ANTENNA WITH RL BEADS SECTIONS ................................................................................... 26
FIGURE 17: THE GAIN AND SWR OF TOP-LOADED ANTENNA WITH FOUR, FIVE AND SEVEN RL BEAD SECTIONS ...................... 27
FIGURE 18: TOP-LOADED ANTENNA WITH RL CHIP SECTIONS...................................................................................... 28
FIGURE 19: THE GAIN AND SWR OF TOP-LOADED ANTENNA WITH FOUR, FIVE AND SEVEN RL CHIP SECTIONS ....................... 28
FIGURE 20: HELIX ANTENNA LOADING DIMENSIONS .................................................................................................. 29
FIGURE 21: HELIX MONOPOLE ANTENNA WITH RL CHIP SECTIONS ................................................................................ 30
FIGURE 22: THE GAIN AND SWR OF HELIX MONOPOLE ANTENNA WITH SMALL AND LARGE PITCH LENGTH AND FOUR 1, 2 AND
18UH RL CHIP SECTIONS...................................................................................................................................... 30
FIGURE 23: FOLDED MONOPOLE ANTENNA WITH RL CHIP SECTIONS ............................................................................. 31
FIGURE 24: FOLDED MONOPOLE ANTENNA DIMENSIONS ............................................................................................ 31
FIGURE 25: THE GAIN AND SWR OF FOLDED MONOPOLE ANTENNA WITH FOUR 1, 2 AND 18UH RL CHIP SECTIONS .............. 32
FIGURE 26: MONOPOLE ANTENNA WITH RL CHIP SECTIONS ........................................................................................ 33
FIGURE 27: THE GAIN AND SWR OF MONOPOLE ANTENNA WITH FOUR 1, 2 AND 18UH RL CHIP SECTIONS ......................... 33
FIGURE 28: INDUCTANCE OF 0805HTR50 (500NH) AND 1008HTR56 (560NH) INDUCTOR FROM COILCRAFTS ................ 35
FIGURE 29: RESISTANCE OF 0805HTR50 (500NH) AND 1008HTR56 (560NH) INDUCTOR FROM COILCRAFTS ................. 35
FIGURE 30: RL CHIP SECTION SCHEMATIC ................................................................................................................ 36
FIGURE 31: RL CHIP SECTION IMPLEMENTATION IN FEKO .......................................................................................... 36
FIGURE 32: S-PARAMETER IMPORTED INTO FEKO .................................................................................................... 37
FIGURE 33: ANTENNA SCHEMATIC IMPLEMENTED IN FEKO ........................................................................................ 37
FIGURE 34: THE GAIN AND SWR OF FINALE ANTENNA DESIGN .................................................................................... 39
FIGURE 35: THE GAIN AND SWR OF FINAL ANTENNA DESIGN COMPARED TO REFERENCE ANTENNA [8]............................... 40
FIGURE 36: RL CHIP LOADED MONOPOLE ANTENNA WITH STRAIGHT STRUCTURE ............................................................. 41
FIGURE 37: THE GAIN AND SWR OF FINAL ANTENNA DESIGN AND FINAL DESIGN WITH A HUMAN BODY IN CLOSE PROXIMITY... 41
FIGURE 38: SIMULATED (RED) VSWR OF ANTENNA MOUNTED ON CORNER (BLUE) ......................................................... 42
FIGURE 39: MEASURED AND SIMULATED VSWR RESULTS OF ANTENNA ON A GROUND PLANE WITH (BLUE) AND WITHOUT
(GREEN) HUMAN BODY IN CLOSE PROXIMITY, ANTENNA RAISED ABOVE GROUND PLANE WITH (CYAN) AND WITHOUT (RED) A DRAG
WIRE AND THE SIMULATION OF AN ANTENNA ON A GROUND PLAIN ............................................................................... 43
FIGURE 40: (A) NEAR-FIELD MEASUREMENT PROBE AND (B) MEASUREMENTS AND SIMULATION SETUP DIAGRAM. ................ 44
FIGURE 41: NEAR-FIELD MAGNITUDE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 300MM ABOVE (BLUE), BELOW
(GREEN) AND ADJACENT (BLACK) TO THE FEED OF ANTENNA. ...................................................................................... 45
6
FIGURE 42: NEAR-FIELD PHASE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 300MM ABOVE (BLUE), BELOW (GREEN)
AND ADJACENT (BLACK) TO THE FEED OF THE ANTENNA. ............................................................................................. 45
FIGURE 43: BUILT LOADED MONOPOLE ON A GROUND PLANE...................................................................................... 46
FIGURE 44: ADT4-6WT TRANSFORMER AND ANTENNA MOUNTED ON CASE ................................................................. 47
FIGURE 45: MEASURED IMPEDANCE OF ANTENNA ON A GROUND PLANE, REAL PART (RED) IMAGINARY PART (BLUE) ............. 49
FIGURE 46: MEASURED VSWR OF ANTENNA ON A GROUND PLANE (BLUE) COMPARED TO SIMULATION (GREEN).................. 49
FIGURE 47: MEASURED (BLUE) AND SIMULATED RESULTS (GREEN) OF THE VSWR OF ANTENNA ON A GROUND PLANE WITH A
HUMAN BODY IN CLOSE PROXIMITY. ....................................................................................................................... 50
FIGURE 48: MEASURED VSWR OF ANTENNA ON A GROUND PLANE WITH (BLUE) AND WITHOUT (GREEN) HUMAN BODY IN CLOSE
PROXIMITY ........................................................................................................................................................ 50
FIGURE 49: MEASURED VSWR OF AN ANTENNA RAISED ABOVE A GROUND PLANE WITHOUT A DRAG WIRE (BLUE) COMPARED TO
THE SIMULATION OF THE ANTENNA MOUNTED AT AN OFFSET ON THE CASE (GREEN) ......................................................... 51
FIGURE 50: MEASURED VSWR OF ANTENNA RAISED ABOVE A GROUND PLANE (GREEN) AND AN ANTENNA ON A GROUND PLANE
(BLUE) .............................................................................................................................................................. 52
FIGURE 51: MEASURED VSWR OF RAISED ANTENNAS WITH (GREEN) AND WITHOUT (BLUE) DRAG WIRE ............................. 53
FIGURE 52: MEASURED VSWR OF A RAISED ANTENNA WITH A DRAG WIRE (BLUE) COMPARED TO THE SIMULATION OF THE
ANTENNA MOUNTED OFFSET ON THE CASE (GREEN) ................................................................................................... 53
FIGURE 53: MEASURED IMPEDANCE OF RAISED ANTENNAS WITH (BLUE) AND WITHOUT (RED) DRAG WIRE REAL PART (SOLID
LINE), IMAGINARY PART (STIPPLE LINE).................................................................................................................... 54
FIGURE 54: NEAR-FIELD MAGNITUDE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 0MM ABOVE ........................ 55
FIGURE 55: NEAR-FIELD MAGNITUDE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 300MM ABOVE .................... 56
FIGURE 56: NEAR-FIELD MAGNITUDE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 300MM BELOW .................... 57
FIGURE 57: NEAR-FIELD PHASE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 0MM .......................................... 57
FIGURE 58: NEAR-FIELD PHASE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 300MM ABOVE............................. 58
FIGURE 59: NEAR-FIELD PHASE OF MEASUREMENT AND SIMULATION WITH X EQUAL TO 300MM BELOW ............................ 59
7
Chapter 1
Introduction
1.1 Background
In today‟s digital age, protecting soldiers from their enemy involves having better
weapons, more accurate information and better electronic shielding. Of growing
concern are radio-controlled improvised explosive devices (RCIEDs) which are used
in traps in urban environments where such devices are triggered by a radio signal to
attain the highest impact. The best protection against such RCIEDs is to avoid
detonation by jamming the enemy‟s RCIEDs communications link to allow the device
to be safely defused or removal.
A typical wideband man-pack currently used to deactivate RCIEDs

is displayed in figure 1. The jammer is mounted on a soldier‟s back
in a metal case, 270 X 230 X 90 mm. From the case a 1.5m long
whip antenna spans upward. This antenna is both flexible and
foldable, allowing movement through doorways or under other
obstacles.
Since the jammer is mounted on a soldier‟s back, it shouldn‟t

interfere with the soldier‟s performance. The weight should also be
kept to a minimum and should be easy to manipulate if needed. To
ensure all round coverage he antenna must have an omni-
directional azimuthal pattern with no pattern breakup. This
transmitter is rated for 20 Watt continuous duty power handling and
has a bandwidth stretching from 20MHz to 500MHz. The
transmitter requires an antenna with a voltage standing wave ratio
(VSWR) below 3:1, but typically 2.5:1, throughout the specified
frequency band. Figure 3 shows the VSWR performance of an
available antenna designed to have an input impedance of 50
Ohm.
When this 0.9kg antenna is folded to improve mobility it has a

height of 0.85m. The antenna‟s gain variation over its frequency
bandwidth has the lowest gain of -15 dBi at 20MHz which rises to a
maximum of -4 dBi at 300MHz (refer to Figure 2).
Figure 1: GMJ9000V
Portable Jammer with This thesis discusses the design of an antenna with improved
reference antenna [8] performance above that of this current antenna which is taken as
the performance reference.
8
Figure 2: Measured OMNI-A0124-01 Antenna Gain from literature [8]
Figure 3: Measured OMNI-A0124-01 Antenna VSWR from literature [8]
9
1.2 Thesis outline
The omni-a0124 antenna in figure 1 is taken as the reference antenna design. As this is an
electrically short antenna, no spatial gain increase can be achieved. So, if the power
radiated is to be increased through “effective” gain increase the focus must be on increasing
the antenna‟s efficiency. As the power reflected obviously depends on the VSWR, the two
electrical parameters of importance in the design are the Gain (or more accurately the
efficiency) and the VSWR.
The antenna design specification, targets based on the reference antenna design, can be
summarised as follows:
 Bandwidth from 20 MHz to 500 MHz
 The lower frequency gain should be higher than -14dBi and at the higher frequency
gain should exceed -4dBi.
 The voltage standing wave ratio should be less than 3:1.
 It must be foldable to a length equal to or less than 0.85 m
 The unfolded whip should not exceed a length of 1.6 m and a diameter of 25 mm
 The feed power handling should allow for 30 W
 It must weigh less than 1 kg
10
Chapter 2
Literature study
As stated in Paragraph 1.2, for an electrically short antenna no significant spatial gain
increase can be achieved by changing its shape. One standard way to improve
bandwidth performance of such an antenna is by loading the antenna with various
combinations of R, L and C circuits. Some aspects of „loading‟ will be introduced as
part of a literature discussion.
2.1 Monopole antenna loaded with non-Foster circuit
Zhang, Sun, Li, Wang and Xue [1] worked on the design of an antenna with a stable
impedance and radiation pattern over a wide frequency range. Methods to reduce the
dimensions of antennas were also investigated and a monopole antenna with a
length less than λ/4 was introduced in their paper. The antenna becomes capacitive
at low frequencies; it requires the use of a compensatory inductor to obtain real input
impedance. The term adopted where the impedance has only a small imaginary part
is “resonance”.
The poor impedance matching of the antenna‟s input impedance due to the small
radiation resistance at the lower frequencies will also hinder the antenna‟s abilities to
radiate energy due to the high reflection coefficient. By adding inductive series
elements and changing the position and values of the lumped elements the antennas
current distribution and effective radiation pattern amplitude is changed, allowing
wideband behaviour to be obtained.
The monopole antenna‟s bandwidth was extended and the dimension was reduced
when it was loaded with several lumped elements composed of resistors, inductors
and capacitors in different positions along a radiator. The antenna‟s efficiency was
however low over the whole frequency range as its Q factor could not break through
the Chu-Harrington limit where the Q value of a small antenna is relatively equal to
the volume of a sphere that encloses it, thus the antennas bandwidth will shrink.
They used a technique presented by Sussman-Fort and Rudish regarding non-Foster

impedance matching. Networks of negative inductors and capacitors were used to
produce a non-Foster matching network where the network was used to bypass the
restrictions of gain-bandwidth theory.
The concept adopted here is the addition of a variable inductor (non-Foster circuit) to
try extending the “resonant” behaviour. The non-Foster circuit when compared to
passive matching achieved higher efficiency over a wider bandwidth. The antenna‟s
current distribution and radiation pattern stays unaffected by the use of a non-Foster
circuit.
11
They present a monopole antenna with a length less than λ/4. Due to its length, the
monopole antenna is mostly capacitive; this can be compensated for with the use of
an inductor. The antenna can “resonate” at different frequencies if the value of the
inductor is changed. It is considered, from these results, that a monopole antenna
should be loaded with a non-Foster circuit to obtain resonance over a broad
frequency range.
The bandwidth could be increased if the inductor‟s inductance could be changed to

the reactive values needed to obtain the required impedance properties as given in
Paragraph 1.2, over a wide frequency range. Jjj you should automate your fig
numbers – it is an easy skill worth knowing
Zhang et al proposed a non-Foster circuit constructed from inductors and capacitors

with positive and negative element values, used as illustrated in Figure 4. This circuit
will act as an element with a negative reactance slope, which will allow resonance at
the frequency depending what reactance is required (refer to Figure 5). Thus if the
wire, with a length smaller than λ/4, impedance is transformed to the required
reactance, shown on figure 5, the wire will resonate at that depended frequency. The
admittance of a network will increase when Foster elements, L and C, are applied to
the network.
Figure 4: non-Foster circuit schematic
Figure 5: Reactive value required for an antenna length of

λ/4 to achieve resonance at a frequency [1]
12
The input reactance and resistance with and without the addition of a non-Foster
circuit are illustrated in Figures 6 and 7. As shown in these figures, without the Foster
circuit the antenna‟s impedance, especially at the low frequencies, is capacitive and
the radiation resistance is so small that the energy concentrating near the antenna
could not be radiated.
Figure 6: The value of the input reactance with and without non-Foster circuit
loading [1]
Figure 7: The value of the input resistance with and without non-Foster
circuit loading [1]
13
With the non-Foster circuit loaded, the reactance of the antenna is smaller than the
unloaded antenna over the wide frequency range with slight variances over
frequency making the non-Foster loaded monopole antenna better matched over the
frequency range..
The unloaded monopole antenna bandwidth with VSWR < 3 is between 85 MHz and
130 MHz, where the bandwidth of the loaded antenna is between 65 MHz and 150
MHz, the non-Foster circuit has increased the bandwidth by 90%. An optimization
algorithm based software designed to calculate matching networks was used to
obtain a passive matching network to further increase the bandwidth, which is also
illustrated in Figure 8.
Figure 8: The values of the VSWR with and without non-Foster circuit loading [1]
The radiation patterns of the loaded and unloaded monopole antenna are depicted in
Figure 9. The radiation pattern of the loaded antenna is more stable over its
bandwidth, where the unloaded monopole antenna‟s pattern deteriorates. The
bandwidth obtained using this technique is too narrow to meet the specifications.
Figure 9: Radiation pattern of loaded (left) and unloaded (right) monopole antennas. [1]
14
2.2 The Short Loaded Monopole analysed by Hansen, Stewart and Harrison
Hansen [2], in a paper that relates to the current discussion on short loaded
monopole antennas, mentions the following with regard to the principle of antenna
loading:
"In principle the loading inductor functions by keeping the current distribution nearly
(emphasis added) constant from the feed to the load point, thereby increasing the
current moment. Since the transmitting parameter, radiation resistance, varies as
current moment squared, and since the receiving parameter, effective length, varies
with the current moment, it is clear that inductive loading will improve short monopole
performance. There is a value of loading reactance which allows the current to
approximate a constant value out to g·h (the fractional distance of the coil position
from the feed) with a linear drop-off beyond. This value of inductance is, however,
insufficient to produce input impedance resonance. The resonant value of load
produces a modest current peak (emphasis added) just beyond g·h so that the
current moment is increased by an additional amount over that expected from the
constant current model."
Stewart [3] pointed out that Devoldere‟s paper [10] about Low-Band DXing (receiving
and identifying distant radio or television signals) and Brown‟s paper [11] on mobile
antennas, fail to show the current peak in shorter (< 0.1 lambda) loaded antennas.
He tried duplicating the Hansen and Cebik [2] findings that the current distribution
along a radiator has a direct impact on the efficiency, but that didn‟t appear to be the
case. Referring to Hansen: "radiation resistance varies as current moment squared”.
The change in the radiation resistance impacts the efficiency, thus the gain of the
radiator. According to Stewart, an unloaded, no-resonating monopole has the same
efficiency as a resonating monopole of the same size. The current distribution will
only have an effect on the radiation resistance when loss is introduced.
Harrison [4] analysed the loaded monopole using superposition of asymmetrically

excited dipoles. He calculated the input impedance, current and loading inductor
currents for several lengths, with various loading coil Q values.
There seems to be a gradual increase in efficiency as the load point moves closer to
the ends of the antenna. His data stopped with the loading point‟s position 2/3 from
the antenna‟s feed. Method of Moments (MoM) calculations were made by Hansen
over several years and indicated that the maximum efficiency point for the loading
occurs closer to the feed.
The current should rise from the feed to a modest peak near the load point and then
decay. This will improve radiation resistance over an unloaded monopole. A point
can then be chosen to give input impedance closer to 50 Ω with some loss in
efficiency.
In conclusion, the previous authors‟ theories were combined to design an improved

antenna. Zhang et al showed that a change in input reactance would be needed to
15
resonate over a wide frequency range. This will make the impedance stable for
matching. Hansen stated that the current should rise from the feed to the first RL
loading point and decay for the best efficiency. Stewart added that the current
distribution will only have an effect when loss is introduced. Loading points towards
the end of the antenna were investigated with MoM calculations, as indicated by
Harrison and pointed out in his paper.
2.3 Frequency enhanced monopole design
Richard E Deasy and Cedar Rapids [7] designed an antenna with a length
substantially less than one quarter wavelength at the operational frequency. A large
number of inductors, installed in series with the wire, were used to enhance the
antenna‟s bandwidth. These inductors were isolated from one another so as to have
negligible mutual inductive coupling. The inductors‟ size must be kept to a minimum
to avoid individual resonance within the operating frequency range. Inductors with
progressively greater inductance values are spaced closely together towards the end
of the antenna.
They believed the best performance might be achieved when the distance between
the inductors decreased logarithmically as the inductance increased towards the end
of the antenna. Any positioning pattern with an increasing inductive loading in a linear
fashion towards the end will be beneficial.
2.4 Effect of Ground-Plane on antenna performance
As the antenna to be designed is operated close to the earth‟s surface, the issue of
whether the earth should be considered as a ground plane was addressed. Antenna
characteristics are generally derived by assuming the antenna to be in free-space. A
monopole antenna can be realised and analysed as a dipole when an infinite, perfect
conducting ground plane is introduced. This is known as the method of images, were
the ground plane is removed and an image antenna is supplemented to mimic the
signal reflection associated with the ground plane. An antenna‟s most important
characteristics, antenna pattern and terminal impedance, will be altered with the
addition of a ground plane in the vicinity of the antenna.
An antenna, will in general, radiate without a ground plane, but with a pattern and
impedance different to those of an antenna above a ground plane. A ground plane
can be seen as a reflector of energy from the antenna itself, which sets up
constructive and destructive interference of signals in space which, in turn, alters the
antenna pattern. The parasitic capacitance from antenna to ground plane alters the
antenna‟s terminal impedance.
16
Tom Yestrebsky‟s [5] paper that relates to the design of a wireless system and the
effect a ground plane has on an antenna system, mentions the following:
For applications where one has the luxury to use or not use a ground plane, the
choice is not particularly clear. If, by using a ground plane, the modified antenna
pattern, directionality, and terminal impedance yield the best system performance,
then it should be used. Otherwise it is not advisable. For applications where a ground
plane must exist, or where no good ground plane can be allowed, the antenna should
be optimized for that particular condition. Finally, there is no reason an adequate
antenna cannot be constructed, even if there is no good ground plane to work
against.
In determining a design for this wideband monopole antenna the advice of Tom
Yestrebsky is followed - all electromagnetic modelling will assume the addition of a
perfect electrically conducted ground plane.
2.5 Realised Gain:
The term realised or apparent gain is used in this work when the mismatch loss is
added to the strict definition of gain. The apparent gain is an important parameter in
this work as it includes the effect of the antenna‟s standing wave ratio (SWR) and
efficiency in one parameter. This may then be taken as the figure-of-merit.
Apparent/realized gain can be defined as:
Realized_gain = Directivity X Radiation_efficiency [Def1]
Realised_gain =|s21|²X directivity
The forward transmission coefficient (S21) can be calculated using the antenna‟s
reflection coefficient (S11) with the following equation
| [1]
This will determine the gain of the antenna with the power decrease due to reflected
power taken into account.
The radiated power, , of the antenna would therefore be:
.η [2]
η - radiated efficiency
Pᵢ - incident power
Clearly, realised gain is always <=true gain.
17
2.6 Conclusion from Literature
After examining the above mentioned papers we can conclude the following:
 The current distribution on the antenna may be changed from the triangular unloaded
case to yield an improved gain and radiation power.
 The antenna‟s impedance must be managed so as to meet the standing wave ratio
requirement.
 The antenna‟s current distribution and impedance can both be changed through the
use of inductive and resistive loads located at different positions on the antenna.
Zhang et al advised that to achieve an improvement in the antenna‟s impedance and

efficiency; the load‟s impedance should change with a certain gradient by using negative
inductors and capacitors.
To achieve the goals set out in Paragraph 1.2, the work of Richard E Deasy, Cedar Rapids
and Robert Hansen will be used by positioning inductors on an antenna to reduce its
capacitance and alter the antenna‟s current distribution.
18
Chapter 3
Simulations and Analysis of Antenna Designs
This paragraph presents a brief synopsis of the antenna development and results.
The design process normally starts by considering what building blocks are available to meet
the requirements, and in this case an inductor is obviously the element of choice with a high
inductance being needed over a wide bandwidth. As available commercial inductor‟s
inductance value decreases with an increase in frequency and bandwidth, it will be shown
that the best results are obtained when using a number of series inductors for realisation.
The required inductance can be realised in two ways, ferrite beads or chip inductors. The
term ferrite “bead” is used to denote a toroidal type ferrite core which is slid over the metallic
rod/whip of the antenna, while “chip” refers to a conventional inductor inserted in series
between rod/whip sections. The bead is considered first.
The ferrite bead is placed over the antenna with, the antenna passing through the bead.
Refer to Figure 12 in Paragraph 3.1. Although it is at first counter-intuitive to see this bead as
adding a series inductor, an imaginary wire can be considered as closing over the bead to
form the single turn. The available ferrite material found to use for this design was Perminvar
with a working frequency range between 10MHz and 200MHz with one turn around this
ferrite bead providing an inductance of 330nH. Practically the imaginary wire is replaced with
a resistor to add resistivity in parallel to the inductance so as to reduce the reflected current‟s
amplitude.
The second inductance realisation option investigated is the use of “chip” inductors with the
RL chip sections using one or a number of series inductors connected over a parallel
resistor. Refer to Figure 11 in Paragraph 3.1 for a diagram of the RL chip section.
With two inductance realisation options available the first topology investigated was that of
the monopole antenna loaded with the RL bead sections, refer to Figure 10(a). The
inductance that the bead could provide was fixed at 330nH as it can only have one turn. The
number of RL bead sections loaded on the antenna was varied to change the inductance
added and the use of four, five, seven and even nine RL bead sections did not provide
significant gain and impedance improvement.
After establishing that bead sections do not offer sufficient inductance to improve the
monopole antenna‟s performance, a topology change was investigated. The topology used
first is popular in low frequency communications: The top-loaded antenna (refer to Figure
10(b)). This antenna is again loaded with various RL bead sections positioned at various
positions during the optimisation process to determine the optimal resistance and position.
The number of RL bead sections used in the simulation was four, five and seven and as with
the monopole even using seven bead sections it is again shown that the low bead
inductance cannot significantly improve the antenna‟s performance, making it clear that
higher inductance RL chip sections should be used.
19
The next topology focussed on increasing the antenna‟s length using RL chip sections at
four locations on the antenna at fixed locations as the position did not make a significant
difference in previous simulations. The antennas length was increased by “winding” it into a
helix structure that would allow a longer wire within the 1.6 m height restriction (refer to
Figure 10(c). The inductance required was investigated using three inductance values
starting at 1 µH. This was increased to an inductor of 2 µH which could be obtained with two
of the previous inductor in series etc. Finally the optimisation yielded an inductance of 18 µH
that cannot be obtained for this bandwidth.
After exhausting the loaded helix structure a different topology was attempted to again
increase the antenna‟s wire length. By simply folding the antennas at the top by 180° the
length can be increased by 800mm (refer to Figure 10(d) because the antenna should be
able to fold halfway to avoid obstructions. This folded antenna is loaded with RL chip
sections between 1 µH and 18 µH.
All the above simulations were optimised with the required goals being
1. a reflection coefficient less than 0.5 to give a VSWR less than 3 and
2. a gain higher than -10 dB across the bandwidth.
As none of the above yielded a satisfactory result the last topology chosen was again the
simple monopole antenna, but loaded with RL chip sections (refer to Figure 10(e)).
Optimisations were performed on the antenna loaded with four RL chip sections with similar
inductance values of 1 µH, 2 µH and 18 µH used and with resistor values ranging between
0 Ω and 1 kΩ.
The optimisation showed that an inductive load of 2 µH could potentially improve the
antenna‟s impedance and the performance to meet the specifications across the required
bandwidth.
The next step was to find suitable commercial inductors. Datasheets showed that the only
way to realise a 2 µH inductor for the frequency range was to make use of inductors in
series. The inductor selected is provided by Coilcraft which has a 560 nH inductor capable of
operating between 1 MHz and 900 MHz. The supplied inductor‟s measured S-parameters
were imported into FEKO in order to model the actual non-ideal inductance and resistance. It
was found that the number practical inductors used in the RL chip section needed to be
increased to six (theoretically 6*0.56=3.4 µH) to have a similar performance as the 2 µH
ideal case.
The final antenna design was compared to the reference antenna‟s VSWR and gain to
determine the improvement in performance. The new antenna design had a similar VSWR to
that of the reference antenna, keeping below 2.5 where the reference antenna reaches 2.7
at the high frequency. When comparing the gain of the final antenna and the reference
antenna to each other a satisfactory improvement can be observed in the antennas
performance having a gain of 5 to 10 dB higher across the bandwidth.
The following sections will describe the results mentioned above in detail.
20
Table 1 provides an overview of the results. Each of the rows represents a topology/loading-
section combination with the values of the loads having been optimised. Entries in the row
show, as a percentage, over how much of the frequency band the antenna meets the
specifications at the top of VSWR, gain and efficiency, with the last column showing when
both gain and VSWR are met simultaneously. The cells of the table are coloured with green-
yellow--orange-red as 100% to 0% to allow visual selection.
A good design should thus

 meet the VSWR over the whole band (first column 100% green), and
 have as much gain as possible (second and last column as far away from red as
possible).
Taking the first row as an example: This antenna has satisfactory S11 over only 17% off the
band; and high gain over 84%; and high efficiency over the whole band. However the last
and most important column shows that this antenna works well only over 2% of the band.
Simulation 22 is the final antenna design using ideal inductors of 2 µH.
The results for the monopole using six practical series 1008HT-R56T inductors with imported
S-parameters are simulation 24
Table 1: Percentage simulation met the VSWR and Gain specification
21
Figure 10: Design topologies investigated
Overview of simulations performed:
Optimisation for the best reflection coefficient and gain was performed on the following
antenna designs. Antenna designs containing RL bead sections are discussed in Paragraph
3.2. The use of a load containing a higher inductance is investigated on various topologies in
Paragraph 3.3. The final antenna design is discussed in Paragraph 3.4.
Paragraph 3.2
 Monopole with four resonant RL bead sections at fixed locations failed to
meet specification.
 Monopole with five resonant RL bead sections at fixed locations failed to
meet specification.
 Monopole with seven resonant RL bead sections at fixed locations failed to
meet specification.
 Monopole with nine resonant RL bead sections at fixed locations failed to
meet specification.
 Monopole with five resonant RL bead sections at varying locations failed to
meet specification.
22
 Top-loaded antenna with four resonant RL bead sections at varying locations

failed to meet specification.
 Top-loaded antenna with five resonant RL bead sections at varying locations
failed to meet specification.
 Top-loaded antenna with seven resonant RL bead sections at varying
locations failed to meet specification.
Paragraph 3.3
 Top-loaded antenna with four RL chip sections at varying locations failed to

meet specification.
 Top-loaded antenna with five RL chip sections at varying locations failed to
meet specification.
 Top-loaded antenna with seven RL chip sections at varying locations failed to
meet specification.
 RL chip loaded monopole with helix structure failed to meet specification.
 RL chip loaded monopole with folded structure failed to meet specifications.
 RL chip loaded straight monopole with 2 μH inductors met the required
specifications.
Paragraph 3.4
 Including the 1008HT-R56TInductor S-parameter model into the FEKO

loaded monopole simulation, simulations showed that six series inductors
would be needed, giving an inductance of 3.36 μH.
3.1 Construction of Inductors
In this chapter a „bead‟ section refers to an inductor formed by sliding a ferrite bead over the
continuous monopole conductor (Figure 12), while a „chip‟ inductor refers to a normal
commercially available chip inductor which, of course, requires that the monopole be cut so
that it can be inserted in series (Figure 11).
A parallel resistor will be attached to the bead and chip inductor, producing the diagram
Figure 13, in order to change the antenna‟s impedance. The bead and chip sections will
induce resonance over the required operating frequency range.
The parallel bead inductor will have an inductance of 330 nH between the operating
frequency (10 MHz to 200 MHz) of the ferrite material. Inductors with different operating
frequencies are commercially available. The required size of inductance can be achieved by
placing smaller inductors with the desired operating frequency in series with each other.
23
Figure 11: RL Chip section

Figure 12: RL Bead section
Figure 13: RL Bead and Chip Diagram
3.2 Topologies containing Bead inductors
3.2.1 Monopole antenna with RL beads at fixed and various locations
The first topology investigated to meet the required specification, was the monopole antenna
loaded with RL bead sections. The monopole antenna was loaded separately with: four,
five, seven and nine RL bead sections at fixed locations (refer to Figure 14). The monopole
antenna was also loaded with five RL bead sections at various locations to investigate the
effect of the beads location on the antenna‟s performance.
The RL beads of the first monopole antenna were spaced 300mm from one another starting
300mm from the feed. The RL beads of the second antenna model were located at the
following distances from the feed:
200 mm, 500 mm, 800 mm, 1000 mm and 1400 mm. The RL beads for the third and fourth
model were both spaced 200mm from each other starting 200mm from the feed. The last
model of this section had RL bead sections at various positions.
The model was optimised with the goal of having a reflection coefficient lower than 0.5 while
having a gain higher than -10 dB. The antenna with five RL beads at fixed and various
24
locations have equal gain and SWR therefore the position of RL beads will not have an
effect on the antenna‟s characteristics as Figure 15 illustrates. Neither one of the simulations
meets the specification of the SWR lower than three nor a gain higher than -14 dB across its
bandwidth
This topology will not meet the specifications as required and therefore the top-loaded
antenna with RL bead and chip sections must be investigated.
Figure 14: Monopole antenna

with RL Bead sections
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
4
VSWR
Straight Mono with 4 Bead loads at Fix Step

2 Straight Mono with 5 Bead loads at Fix Step
Straight Mono with 5 Bead loads at Var Step
1
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 15: The gain and SWR of monopole antenna with four, five, seven and nine RL Beads at fixed
locations (Fix Step) and various locations (Var Step)
25
3.2.2 Top-loaded antenna with RL bead and RL chip sections at various locations
The top-loaded antenna was considered because of its common use for low frequency
communications. Antenna models were optimised with four, five and seven RL beads at
various locations (Figure 16), by assigning an area to each inductor.
Boundaries were set for each bead from which a position was chosen during the
optimisation process. The resistance value had n boundary between 0 Ω and 1 kΩ.
The following boundaries conditions were set for the antenna with four, five and seven RL
beads:
 Antenna with four RL beads: 0mm, 800mm and 1600mm

 Antenna with five RL beads: 0mm, 400mm, 900mm and 1600mm
 Antenna with seven RL beads: 0mm, 400mm, 700mm, 1000mm, 1300mm
and 1600mm
Two RL beads were located on the horizontal wire at the top of the antenna. RL bead was
located in the middle of each horizontal arm extending between 50mm and 300mm. The
antennas gain, SWR and efficiency graphs are plotted in figure 17. The antennas with four
and five RL bead sections have a gain higher than -15 dB, with a VSWR higher than three
across most of the bandwidth. The antennae containing the four and five bead sections have
100% efficiency across the bandwidth and aren‟t displayed on the graph. The beads contain
a low inductance which short-circuits the parallel resistor allowing for a higher current and
thus a high efficiency.
Only when seven RL Bead sections are used the VSWR decreases to lower than three
across most of the bandwidth, excluding between 30MHz and 60 MHz. The addition of more
loads on the antenna started to affect the antenna‟s efficiency, having a maximum efficiency
of 50% at 500MHz.
It was clear the inductance must be increased to obtain a VSWR less than three across the
bandwidth. The use of RL chip sections must be investigated.
Figure 16: Top-loaded Antenna with

RL Beads sections
26
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Toploaded Mono with 4 Bead loads at Var Step

60
Efficiency %)

40
20
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 17: The gain and SWR of Top-loaded antenna with four, five and seven RL bead sections
3.3 Topologies containing chip inductors
3.3.1 Top-loaded antenna with RL chip sections at various locations
It was clear that more inductance was required to reduce the antennas SWR as the gain
specification was met when RL beads were used but not the standing wave ratio. The
inductance was increased using RL chip sections on the top-loaded antenna (Figure 18).
RL chip sections consist out of a 1µH inductor and a parallel resistor between 0Ω and 1kΩ.
Similar boundaries were used on the RL chip loads than on the RL beads as presented
below:
 Antenna with four RL chips: 0mm, 800mm and 1600mm

 Antenna with five RL chips: 0mm, 400mm, 900mm and 1600mm
 Antenna with seven RL chips: 0mm, 400mm, 700mm, 1000mm, 1300mm
and 1600mm
Figure 19 indicates that the use of RL chip loads decrease the antenna‟s capacitance
reducing the reflection coefficient decreasing the SWR. The addition of more RL chip loads
to the antenna will reduce the SWR at 20MHz but will increase at higher frequencies as seen
in Figure 19. All the RL chip loaded antennas will meet the gain specification for they are
higher than -14dB across the bandwidth.
27
The top-loaded antenna will not be a usable solution because of fluctuations SWR across
the bandwidth. By reducing SWR at 20MHz and 275MHz by adding more inductance
increases SWR at 50MHz and 350MHz. Thus the antennas SWR won‟t be less than three
across the bandwidth by increasing the inductance added.
The next topology investigate will be a monopole antenna with a helix structure loaded with
RL chip sections.
Figure 18: Top-loaded Antenna with

RL Chip sections
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Toploaded Mono with 4 Chip loads at Var Step

60
Efficiency %)

40
20
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 19: The gain and SWR of Top-loaded antenna with four, five and seven RL chip sections
28
3.3.2 Monopole antenna with a helix structure and RL chip sections at fixed
locations
Another design option would be to curl the antenna into a helix, thereby increasing the
length of the antenna but keeping to the maximum height of 1.6 meters (Figure 21).
Capacitance will be added to the antenna‟s impedance if the distance from curve to curve in
the helix structure becomes too short. The antenna‟s length will increase if a shorter distance
from curve to curve (Dimension C in Figure 20) is used. This will allow more wire in the
specified allowed height.
The number of loads used must be kept to a minimum to prevent effecting the antenna‟s
SWR and gain at the high frequency. A simulation was performed where the pitch length
was set to a minimum of 10mm to increase the total length of the wire.
Four RL chip loads, with an inductance of 1 µH, were spaced 300mm from each other
starting 300mm from the feed. The antenna model with the small pitch length was optimised
for a SWR less than 3 and gain higher than 10dB. The SWR is suitable at 20MHz but
exceeds 3 at the higher frequencies as a result of the capacitance added to antenna‟s
impedance by the closely positioned curves. This is the result of the small pitch length
(Dimension C in Figure 20) used to increase the antenna‟s length.
Figure 20: Helix antenna loading dimensions
It was required to use a large pitch length of 200mm to reduce the capacitance added to the
antenna‟s impedance. Three inductance loads were used, 1 µH, 2 µH and 18 µH to
investigate the inductance required to counteract the antenna‟s capacitance. The antenna
with a 1 µH inductor meets the SWR specification over its bandwidth except at 20MHz but its
gain drops far below -14dB at 350MHz.
The antenna does not meet the gain and SWR specifications when the inductance is
increased to 2 µH. Then antenna‟s gain is increased and SWR is improved (except at
20MHz) when an inductance of 18 µH is used, refer to Figure 22.
The addition of more inductance will allow the antenna to meet the SWR specification but
will not allow the antenna to meet the gain specification. This is made visible in Figure 22
when the magnitude of the antenna‟s gain experiences large drops at different frequencies.
29
The monopole antenna with helix structure will not be a valid solution for the required
specifications. Another topology where the antenna‟s length can be increased must be
investigated.
Figure 21: Helix monopole antenna

with RL chip sections
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Helix Mono small pitch lenght with 4 Chip loads at Fix Step
60
Efficiency %)
Helix Mono large pitch length with 4 1uH Chip loads at Fix Step
Helix Mono large pitch length with 4 2uH Chip loads at Fix Step
40 Helix Mono large pitch length with 4 18uH Chip loads at Fix Step
20
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 22: The gain and SWR of Helix monopole antenna with small and large pitch length and four 1, 2
and 18uH RL chip sections
30
3.3.3 Monopole antenna with a folded structure and RL chip sections at fixed
locations
In this design the antenna is lengthened by another half of its maximum length by folding the
antenna at the top (Figure 23). The antenna is still loaded with parallel inductors and
resistors (Figure 24). Simulations were also performed using the same inductive loading
used in the helix structured antenna simulations. Dimensions are: A = 300 mm, B = 600 mm,
C = 900 mm and D = 1200 mm. The parallel resistor‟s resistance values were changed to
values between 0 Ω and 1 kΩ to find the loading needed for a reflection coefficient lower
than 0.48 and a gain higher than -10 dB.
The antenna‟s SWR was lower than 3 when 1 and 2 µH inductors were used in the RL chip
sections, refer to Figure 25. The antenna‟s SWR increases above 3 at 20 MHz, reaching 2.8
at 175 MHz when an 18 µH inductor is used in the RL chip sections.
The folded antenna designs will not meet the specifications. Reason being that the antenna
loaded with an 18 µH inductor will not meet the SWR specifications and the other two
antennas do not meet gain specification. The antenna loaded with 1 and 2 µH inductors has
a drop in gain at 400 MHz.
The folded antenna with 1 or 2 µH inductors would have been the ideal solution for it has
stable SWR less than 3 across the bandwidth but does not contain a constant gain over its
bandwidth.
The last topology to investigate is the monopole antenna loaded with RL chip sections.
Figure 24: Folded monopole antenna dimensions

Figure 23: Folded monopole
antenna with RL chip sections
31
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Folded Mono with 4 1uH Chip loads at Fix Step

60
Efficiency %)

40
20
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 25: The gain and SWR of Folded monopole antenna with four 1, 2 and 18uH RL chip sections
3.3.4 Monopole antenna with RL chip sections at fixed locations
In this design the amount lumped elements were reduced from five (current amount on
reference antenna) to four RL chip sections. The four RL chip sections were positioned
300mm from each other starting 300mm from the feed (refer Figure 26)
The RL chip sections were comprised of 1, 2 or 18 µH inductors in parallel with a resistor

between 0Ω and 1kΩ. The parallel resistors‟ resistance values were changed to values
between 0 Ω and 1 kΩ to find the loading needed for a reflection coefficient lower than 0.48
and a gain higher than -10 dB.
The monopole antenna with 18 µH inductors does not meet the SWR specifications between
20 MHz and 30 MHz for it still has a large capacitance at the low frequency. The antenna
also does not meet the gain specification, being less than-14 dB, at 20 MHz and 500 MHz
(refer to Figure 27).
The antenna meets the SWR and gain specifications when an inductor between 1 and 2 µH
is used. The antenna with the 1 µH inductor has a 2 dB higher gain between 20 MHz and
250 MHz. The antenna‟s gain between 250 MHz and 500 MHz has a maximum difference of
2 dB lower than the antenna with the 2 µH inductor. The deciding factor of the inductance
required lies in the SWR. When a 1 µH inductor is used the SWR of the antenna will
increase to 3 at 20 MHz revealing that a 2 µH inductor is required.
32
The next step in the design proses would be to confirm the use of the right inductor by
incorporating the inductor‟s S-parameters into the simulation and obtain the required resistor
values.
Figure 26: Monopole antenna

with RL chip sections
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
60 Straight Mono with 4 1uH Chip loads at Fix Step

Efficiency %)
Straight Mono with 4 2uH Chip loads at Fix Step

Straight Mono with 4 18uH Chip loads at Fix Step
40
20
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 27: The gain and SWR of Monopole antenna with four 1, 2 and 18uH RL chip sections
33
3.4 Final monopole antenna containing chip inductors
3.4.1 Choice of inductors
The inductance values required can be obtained from a series combination of fixed ferrite
inductors. A suitable ferrite material for the low frequency range is Perminvar, since its
working frequency range is from 10 MHz to 200 MHz with a μi equal to 12. The problem with
Perminvar would be the reduction in its permeability from 200 MHz upwards. This will cause
the inductance to decrease from 200 MHz because of a change in permeability. The
inductance is depend on the permeability, thus in this case adding less inductance to the
antenna from 200 MHz upwards. Richard E Deasy and Cedar Rapids [7] advise the use of a
chip inductor with a small cross-section for the upper frequency range.
The catalogues of a large inductor supplier, Coilcraft, were studied in the search for
inductors which will work over the required frequency range, but the available 1 to 8 µH
inductors mostly supply the inductance at frequencies lower than required. A compromise
can be reached by adopting stacked combinations of either the 500 nH or the 560 nH
inductors, which both work over the required bandwidth with acceptable resistance.
Referring to Figure 28 and Figure 29, the1008HT-R56T inductor with a 560 nH inductance
was designed to be used for a bandwidth between 1 and 900 MHz, while the 10HTR56 has
a higher than 590 nH inductance at 20 MHz, which increases to 1 µH at 475 MHz. The
08HTR50 model has a 490nH inductance, reaching 620 nH at 500 MHz and both inductors
are experimented with numerically in FEKO.
Due to the strong deviation from ideal performance associated with ferrite inductors, it was
imperative to optimize the design with the actual inductor‟s frequency characteristics. The
inductors S-parameters were loaded from the supplier‟s web-page and added into the FEKO
simulations in order to optimise the resistance that needs to be added. The addition of the
supplied two-port inductance parameter S-parameters proved complicated in FEKO, but this
was resolved by defining the second port as „grounded‟.
34
0.8
Inductance (uH)
0.6
0.4
0.2
1008HTR56
0805HTR50
0
0 100 200 300 400 500
Frequency (MHz)
Figure 28: Inductance of 0805HTR50 (500nH) and 1008HTR56 (560nH) inductor from Coilcrafts
90
80
70
Resistance (ohm)
60
50
40
30
20
10
1008HTR56
0805HTR50
0
0 100 200 300 400 500
Frequency (MHz)
Figure 29: Resistance of 0805HTR50 (500nH) and 1008HTR56 (560nH) inductor from Coilcrafts
35
3.4.2 Inductor’s frequency characteristics added to antenna model
The frequency characteristic data of the component which were going to be used, the
10HTR56 inductor, must be added in the form of S-parameters to the antenna model. The
ideal antenna design, which fulfils the necessary specifications, was optimised with the
component data to tune the necessary resistance which should be added. Any negative
effects caused by the inductor can be tuned out when the S-parameters are introduced to
the model.
The RL chip sections were constructed as shown in the schematic (Figure 30). Six inductors
were placed in series with each other, with a parallel resistor. These RL chip sections were
implemented in FEKO by placing six ports in series with each other on n wire segment and
connecting a wire segment with a resistive load (Figure 31).
The measured S-parameters of the inductors for different frequencies are stored in a
touchstone file and can be imported into FEKO as a general network. The measurements
are those of a two port network and FEKO requires a one port network. The general
network‟s first port is connected to the port which is located at the intended location (each of
the six series ports) and the second port is short-circuited (Figure 32).
These RL chip sections are positioned at the locations shown in the antenna schematic
(Figure 33). The only variable would be the magnitude of the parallel resistor connected to
the six series inductors.
Figure 30: RL chip section schematic
Figure 31: RL chip section implementation in FEKO
36
Figure 32: S-parameter imported into FEKO
Figure 33: Antenna schematic implemented in FEKO
37
The FEKO model could be optimised after adding the 10HTR56 inductor‟s S-parameters.
The required resistor values could be obtained for using atypical inductor. The optimisation
showed that the following resistor values would allow the antenna to meet the SWR and gain
specifications
Resistors in Figure 33:
R1. 270 Ω
R2. 200 Ω
R3. 560 Ω
R4. 680 Ω
Figure 32 depicts the SWR and gain of the antenna with the 10HTR56 inductors.
The 2 µH inductor was obtained by pressing four 10HTR56 inductors (560 nH) in series at
each of the four RL chip sections. Figure 34 illustrates that the four series inductors,
providing 2.2 µH, will still have a high SWR at 20 MHz and 500 MHz.
The inductance was increased by adding two additional inductors, forming six series
10HTR56 inductors. The antenna‟s gain is higher than -11 dB across the entire bandwidth
reaching a maximum of 4 dB at 210 MHz (refer to Figure 34). The antenna has a standard
wave ratio less than 2.5 over the bandwidth (20 MHz – 500 MHz).
The antenna‟s SWR boundary is reached at 20 MHz and 500 MHz when four series
10HTR56 (560 nH) inductors are placed in series to construct a 2 µH inductor. This problem
is solved by using six series 10HTR56 inductors, raising the parallel inductance to 3.3 µH.
This will deliver an antenna with a gain 3 dB higher with a similar SWR than the reference
antenna.
38
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
60
Efficiency %)
40
20
Straight Mono with 4 4x 1008HT-R56T loads at Fix Step
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 34: The gain and SWR of Finale antenna design
3.4.3 Final antenna design compared to reference antenna
The next phase will be to compare the new monopole antenna design with the reference
antenna. The reference antenna‟s gain in SWR measurement data was extracted from its
technical reference manual provided by Poynting. The gain and SWR of the two antennas
are depicted in Figure 35. The gain of the new monopole antenna design with four RL chip
sections, consisting of six series 560 nH inductors with a parallel resistor, exceeds the
reference antenna‟s gain by 10 dB.
The monopole antenna design has a higher SWR, below three, than the reference antenna
between 20 MHz and 275 MHz. The new design has lower SWR between 275 MHz and
500 MHz.
39
The new monopole antenna design will perform better than the reference antenna due to its
higher gain. The next phase of the design process would be to build the new monopole
antenna design containing the four RL chip sections with the six series 10HTR56 inductors.
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4.5 Reference antenna
3.5
VSWR
2.5
1.5
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 35: The gain and SWR of Final antenna design compared to Reference antenna [8]
3.4.4 The influence of a human body on the antenna
The next step is to analyse the effect a human body will have on the antenna since they
would be in close proximity. This is done by adding a body, modelled as a box, with roughly
the same dimensions of an average human body, 1.7m tall by 500 mm wide by 50mm deep,
into the antenna model, Figure 36. The entire human body was set to the relative permittivity
of 30.
Figure 37 shows that the body worsens the antenna‟s SWR at the low frequencies by adding
extra capacitance to the impedance. The SWR has a value lower than 3 from 62 MHz
upwards to 500 MHz.
The body‟s effect on the gain is partly advantageous: at 20 MHz it has a gain of -10.5 which
rises to a maximum of 5 dB at 320 MHz and 7 dB at 450 MHz ending at 500 MHz at -2.5 dB,
see Figure 35.
It is important to optimise the antenna with the body incorporated in the model. Techniques
to reduce the body‟s influence on the antenna are left for a next phase of the work.
40
Figure 36: RL chip loaded monopole antenna with straight structure
10
5
Gain (dB)
-5
-10
-15
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
5
4
VSWR
1
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
60
Efficiency %)
40
20
Straight Mono with 4 6 series 1008HT-R56T loads at Fix Step with a body
0
0 50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 37: The gain and SWR of Final antenna design and Final design with a human body in close proximity
41
Chapter 4
4.1 Practical Measurements
Overview of measurements performed:
The antenna‟s reflection coefficient (S11) was measured and the near-fields were probed
using a Rohde & Schwarz FSH6 network analyser (Appendix A.3). The network analyser
was first calibrated using an open, short and load circuit to give an accurate measurement.
The antenna was connected to the network analyser at its RF input with a coaxial cable. The
antenna‟s reflection coefficient was measured for the following cases:
 The antenna position on a ground plane with and without a human body in close
proximity (closer than a meter) to the antenna
 The antenna above the ground with a block of polystyrene with and without a wire
extending two meters from the case (drag wire).
The influence of a human body on the antenna was tested by measuring the antenna‟s
reflection coefficient with and without a body present whilst positioned on a ground plane.
There were small discrepancies over the frequency range, with the most between 20 MHz
and 300 MHz. All the measurements were compared to the simulation with the antenna
mounted on the case at an offset, positioned above a ground plane.
The effect of the position at which the antenna is mounted on the case should be brought
into account in the antenna model and simulated to compare with the measured results.
Figure 38 shows the difference in the antenna‟s VSWR caused by the position at which the
antenna is attached to the case. Simulations were performed with the antenna attached to
the corner of the case and attached to the middle of a side top edge (at an offset).
3.5
2.5
SWR
1.5
Simulation of corner mounted antenna
1 Simulation of offset mounted antenna
0 100 200 300 400 500

Frequency [MHz]
Figure 38: Simulated (red) VSWR of antenna mounted on corner (blue)
42
The antenna has a VSWR lower than 3:1 in all four cases, satisfying the required VSWR
specification (Figure 39). The measurement with the antenna on the ground plane (green)
should agree with that of the simulation (purple).
Measured and Computed SWR

3.5
2.5
SWR
1.5
Measurement of antenna on groundplane with human body
Measurement of antenna on ground plane
Measurement of antenna raised without a dragwire
Measurement of antenna raised with a dragwire
Simulation of antenna on ground plane
1
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 39: Measured and simulated VSWR results of antenna on a ground plane with (blue)
and without (green) human body in close proximity, antenna raised above ground plane with
(cyan) and without (red) a drag wire and the simulation of an antenna on a ground plain
Near-field measurements:
The near-field probe was positioned at three different heights which were set to a distance X
equal to 0 and 300 mm above and below the antenna‟s connection point. The antenna‟s
near-field was measured at a fixed distance X on the horizontal axis equal to 300 mm, refer
to Figure 40 (b).
The Rohde & Schwarz FSH6 network analyser was calibrated with its open, short and load
circuit in order to elude the effect of the transmission line (coaxial cable) on the phase
measurement. The near-field probe has to be shaped as illustrated in figure 40 (a) in order to
obtain an accurate reading. Extra currents were induced onto the wire affecting the
measurements accuracy.
Each measurement is compared with its counterpart simulation. The near-field magnitude
and phase was extracted and depicted on figure 41 and 42 below.
The measured near-field magnitude with the near-field probe at a height (X) equal to zero
mm is similar to that of the simulation containing a small offset equal to 1 dB at the high
frequency and 6 dB at 20MHz.
When the height (X) of the near-field probe is increased to 300 mm the offset increases to 2
dB and stays constant across the bandwidth. Only at the low frequency the measurement is
6 dB lower than the simulation
43
The offset between the near-field measurements and simulation fluctuate from positive to
negative when the near-field probe height is decreased to 300mm below the antenna feed
point. The offset fluctuates between 0 and 2 dB from 44 MHz towards 490 MHz with a larger
near-field offset equal to 9 dB at 20 MHz
The phase component must also be compared to the simulations to verify the antenna‟s
gain. The difference between the measured and simulated phase of the antenna decreases
with an increase in height (X). The phase offset for the near-field probe 300 mm below the
feed point is equal to 150° that decreases to 80° towards the feed point. The phase offset
finally decreases to 40° when the near-field probe is raised to 300 mm above the feed point.
The phase difference decreases from the maximum amount mentioned above to lower than
10°. The simulated and measured graphs follow a similar gradient where the main difference
in the three measurements is an offset that decreases with an increase in frequency from
20 MHz.
The antennas far-field phase will be similar to the simulation‟s far-field calculations, with the
only difference being an offset. It can be accepted that the antenna will have a gain equal to
the simulated results.
(a) (b)
Figure 40: (a) Near-field measurement probe and (b) Measurements and
simulation setup diagram.
44
Nearfield Magnitude Comparison

-30
FEKO X=-300
CST X=-300
Measured X=-300
FEKO X=0
-35
CST X=0
Measured X=0
FEKO X=300
CST X=300
-40 Measured X=300
Nearfield Mag. (dB)
-45
-50
-55
-60
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 41: Near-field magnitude of measurement and simulation with X equal to 300mm
above (Blue), below (Green) and adjacent (Black) to the feed of antenna.
Nearfield Phase Comparison

0
-50
-100
-150
Nearfield Angle(deg)
-200
-250
-300
FEKO X=-300
-350 CST X=-300
Measured
-400 FEKO X=0
CST X=0
Measured
-450 FEKO X=300
CST X=300
Measured
-500
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 42: Near-field phase of measurement and simulation with X equal to 300mm above
(Blue), below (Green) and adjacent (Black) to the feed of the antenna.
45
4.2 Final Antenna Configuration
The antenna design which met all the design

specifications was the 1008HT-560T inductive and
resistive loaded monopole antenna. The next step was
to construct the antenna and compare the
measurements with the results of the simulations.
Referring to Figure A.1, the four loads on the antenna

each contain six series 570 nH inductors supplying
nominally 3420 nH per load. Each set of inductors has a
parallel resistor with different resistance values
connected to it. The distance between the feed and
each load point is 300 mm. The wire section consists of
copper foil tape with the loads connected and wrapped
against a Perspex plastic rod to keep the copper wire
vertical (shown in Figure 43).
Figure 43: Built loaded monopole on

a ground plane
A 6:1 impedance transformer was needed to transform the design reference impedance
used in all the simulations from 300 Ω down to 50 Ω. The best available impedance
transformer which will function in the working frequency range is a 4:1 Mini-Circuits
impedance transformer. A larger reflection coefficient in relation to that of the simulation is to
be expected, but this can be readily compensated mathematically from the measurements.
Another difference between the experimental implementation and the FEKO model is the
antenna‟s attachment position on the case. In the simulation model it is positioned at one
corner of the conducting case, whereas it is physically mounted in the middle of one side of
top edge of the case. This should have only a minor effect, but a similar model should be
used for the comparison.
An AN-type connector „Panel Receptacle Jack‟ is used for the connection between the
antenna and measuring equipment. The ADT4-6WT 4:1 transformer‟s primary pin is
attached to the connector and the secondary pin to the antenna, with the primary and
secondary DOT pins grounded to the case, as displayed in Figure 44.
46
Figure 44: ADT4-6WT transformer and antenna mounted on case
4.3 Measurement equipment
The main measurements which need to be performed are those of the reflection coefficient
and the gain. The reflection coefficient measurement was performed on top of the roof of a
large building which contains enough conductive material (concrete reinforcing) to
approximate a conductive ground plane.
The measuring equipment should be portable and must function in the required frequency
range. The Rohde & Schwarz FSH6 Network Analyser, shown in Figure B.3, was used both
to measure the reflection coefficient of the antenna and as a near-field probe.
4.4 Measurements performed
4.4.1 The reflection coefficient (S₁₁)
S11 measurements were performed for four different situations: with the antenna box
situated on the ground, on the ground with a human body in close proximity to it, mounted on
a block of polystyrene with and without a wire extending from the case (drag wire).
The spectrum analyser was used to measure the antenna‟s reflection coefficient with a
reference impedance of 200 Ω because of the 4:1 impedance transformer. These
measurements needed to be transformed to a reference impedance equivalent to that of the
simulation. Reference impedance equal to 300 Ω was used in the simulation. The antenna
has to have nominal input impedance equal to 50 Ω, meaning a 6:1 impedance transformer
would be needed. Since a 4:1 impedance transformer was necessary it was needed to re-
normalise the measurements to a reference impedance of 75 Ω.
47
The phase delay through the N-type connector is significant at the higher frequencies. This
was measured to be 40 mm, which will have an effect on the reflection coefficient‟s phase
with a delay (Δ) double its length.
The phase will shift with:
Δ = 0.08 m [1]
The shift must be removed to view the correct reflection coefficient, and this is done by
multiplying the measured reflection coefficient, , for the different frequencies as follow:
[2]
To subtract the phase one can multiply equation [3] by equation [2] giving one the real
reflection coefficient, , in [4]:
[3]
[4]
[5]
These mathematical instructions need to be performed to analyse and compare the

antenna‟s frequency characteristics with those of the simulation.
Measurement cases
(a) Antenna on a ground plane

The antenna‟s VSWR was calculated by applying [5] to the measured reflection
coefficient at different frequencies. The network analyser was used to measure the
antenna‟s reflection coefficient in order to calculate its VSWR and phase; measuring the
phase would be difficult using a VNA. The calculated VSWR is depicted in Figure 45 and
compared to the VSWR of the simulation of the antenna on a ground plane.
The simulation‟s reference impedance must be set to 300 Ω to have a VSWR lower
than 3, except at 475 MHz. The antenna‟s VSWR was measured at a reference
impedance of 200 Ω because a 4:1 impedance transformer is used. The VSWR is
mathematically scaled to a reference impedance of 300 Ω to compare with that of the
simulation. The antenna‟s VSWR is lower than 3 across the entire frequency range with
reference impedance equal to 300Ω.
48
The input impedance of the antenna with the 4:1 transformer is equal to 150 Ω, where
the simulation predicted an impedance of 300 Ω (Figure 46). The antenna has a
maximum VSWR equal to 2.8 at 20 MHz and which is lower than 2.5 between 450 MHz
and 500 MHz. This simulation predicted a rapid increase to higher than 3.

400
300
200
Impedance (Ohm)
100
-100
Real part
Imaginary part
-200
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 45: Measured impedance of antenna on a ground plane, Real part (red) Imaginary
part (blue)
3.5
2.5
SWR
1.5

1
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 46: Measured VSWR of antenna on a ground plane (blue) compared to Simulation
(green)
49
(b) Antenna on a ground plane with human body

A human body was placed in close proximity to the antenna when the antenna‟s
reflection coefficient was measured. Figure 47 shows that the antenna has a VSWR of
lower than 3, with a maximum of 2.8 at 20 MHz and an average of lower than 2.5.
Comparing the VSWR of the antenna with and without the presence of a human body,
(Figure 48) it seems that the body had minor effects on the antenna and less than
predicted by simulations (Figure 37). The only difference between the two cases‟ VSWR
in figure 38 appears between 100 MHz and 175 MHz, where the human body lowers the
VSWR slightly.
3.5
2.5
SWR
1.5

1
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 47: Measured (blue) and simulated results (green) of the VSWR of antenna on a
ground plane with a human body in close proximity.
2.8
2.6
2.4
SWR
2.2
1.8
1.6

1.4
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 48: Measured VSWR of antenna on a ground plane with (blue) and without (green)
human body in close proximity
50
(c) Antenna raised without a drag wire

In this measurement the antenna was placed on top of a polystyrene block to achieve a
1 m elevation. The effect of the distance between the case and the ground plane was
tested with this measurement. The maximum VSWR occurs between 50 MHz and
75 MHz, as depicted in Figure 49. The antenna‟s VSWR is mostly lower than 2.5 over
the rest of the bandwidth.
3.5
2.5
SWR
1.5

1
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 49: Measured VSWR of an antenna raised above a ground plane without a drag wire (blue) compared
to the simulation of the antenna mounted at an offset on the case (green)
The main difference between the antennas situated on the ground and those raised
1 m above the ground, occurs at 20 MHz. Referring to Figure 50, where the VSWR of
antenna on the ground plane is 2.8 and for the raised antenna is 2.2. The raised antenna
has a better VSWR over the bandwidth, except between 40 MHz and 80 MHz.
51
2.8
2.6
2.4
SWR
2.2
1.8
1.6

1.4
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 50: Measured VSWR of antenna raised above a ground plane (green) and
an antenna on a ground plane (blue)
(d) Antenna raised with a drag wire

In this measurement the antenna was place on a polystyrene block to give an elevation
of 1 m and a 2 m long wire was connected to the case. The antenna still has a high
VSWR at the low frequencies, with a maximum of 2.8 at 26 MHz, (Figure 51).
The VSWR is lowered over the bandwidth by the addition of the drag wire, except
between 20 MHz and 46 MHz and at the high frequency between 460 MHz and
500 MHz (Figure 52).
The big difference in the VSWR is between 125 MHz and 225 MHz and 400 MHz and
450 MHz. This is because the drag wire adds more resistance to the antenna‟s
impedance, (Figure 53). A small capacitance is also added but the imaginary part of the
antenna‟s impedance is virtually the same.
52
3.5
2.5
SWR
1.5

1
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 52: Measured VSWR of a raised antenna with a drag wire (blue) compared to the
simulation of the antenna mounted offset on the case (green)
2.8
2.6
2.4
2.2
SWR
1.8
1.6
1.4
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 51: Measured VSWR of raised antennas with (green) and without (blue) drag wire
53
400
300
200
Impedance (Ohm)
100
-100
Real without a dragwire
Imaginary without a dragwire
Real with a dragwire
Imaginary with a dragwire
-200
0 50 100 150 200 250 300 350 400 450 500
Frequency [MHz]
Figure 53: Measured impedance of raised antennas with (blue) and without (red) drag wire Real part
(Solid line), Imaginary part (stipple line)
4.4.2 The forward transmission coefficient (S₂₁) - near-field
This measurement was performed using the same spectrum analyser used in the previous
section but with a different setup. A near-field probe was constructed to the form depicted in
Figure 36 in order to avoid interference from other induced currents on the wire connected to
the coaxial cable. The probe has a diameter equal to 90 mm and the length of the
connecting wire is 300mm.
The measurement could be performed after the network analyser was calibrated for the
different lengths of coaxial cables used to connect the probe and antenna to the network
analyser. The network analyser uses the two measurements to determine the antenna‟s
near-field for each frequency.
The network analyser firstly needed to be calibrated with the different feed lines (coaxial
cables) to get precise readings. One could then take sets of measurement with the probe at
three different heights at a constant distance away from the antenna; refer to Figure 34. The
probe was placed 300 mm across the position at which the antenna was attached to the
case. The probe was lifted and lowered by 300 mm at the same distance from the antenna,
to compare the near-field measurements from different locations.
The goal is to measure the S₂₁ parameters to get the magnitude and phase of the antenna‟s
near-field and compare it to the equivalent simulation. If the results concur then the practical
antenna will have the same performance, gain and efficiency, as those of the simulations.
54
4.4.2.1 Near-field Magnitude:
Each measurement is compared with its counterpart FEKO and CST simulation. The near-
field magnitude and phase was extracted and displayed on the graphs depicted below. The
S-parameters imported into FEKO could not be used in CST, forcing the model to use ideal
chip inductors.
(a) Near-field probe height (X) equal to 0mm:

When the measurement taken with X equal to zero mm (where antenna connects to
case) it was observed that the near-field magnitude of the measurement was similar to
that of the simulations. The maximum difference in the magnitude between the FEKO
simulation and the measurement (refer to Figure 54) is at 428 MHz, where the
measurement is 7 dB higher than the simulation.
When comparing the CST simulation to the measurement, the maximum difference
increases to 10 dB higher than the measurement at 346 MHz. The measurement result
does not decrease far below the FEKO results, mainly varying between the FEKO and
CST results.
The dips and spikes in the CST model‟s near-field is the presence of a reflection from the
ground plane. The frequencies affected by the reflection are dependant on the antenna‟s
height from the ground plane.
Nearfield Magnitude Comparison, X = 0mm

-30
-35
-40
Nearfield Mag. (dB)
-45
-50
-55
FEKO
CST
Measured
-60
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 54: Near-field Magnitude of measurement and simulation with X equal to 0mm
above
(b) Near-field probe height (X) equal to 300mm above feed point:
When the measurement was taken at a height of 300 mm, an improvement was
observed when comparing the measurement to its simulations (refer to Figure 55). The
measurement equals is OR is 2 dB higher than the simulation between 35 MHz and 242
MHz. The difference between the FEKO simulation and the measurement reaches a
minimum equal to 5 dB at 415 MHz. The measurement is higher than the CST and
FEKO results between 48 MHz and 128 MHz, with a difference of 5 dB with the FEKO
55
results and 8 dB with CST. The biggest near-field difference between the FEKO and the
measurement result is equal to 7 dB at 291 MHz.
The largest difference between the CST simulation and the measurements is at 169
MHz. CST calculates a near-field 20 dB lower than what was measured. The CST
simulation‟s near-field also increased to 14 dB higher than the measurement, at 35 MHz.
The rapid increase and decrease in the near-field at 35 MHz and 169 MHz is caused by
the reflection from the ground plane.
The measurement fluctuates between the FEKO and CST simulation, in the
measurement with the probe at a height equal to 0 mm.
Nearfield Magnitude Comparison, X = 300mm

-30
-35
-40
Nearfield Mag. (dB)
-45
-50
-55
FEKO
CST
Measured
-60
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
above
(c) Near-field probe height (X) equal to 300mm below feed point:
The comparison between the measurement and simulations where the near-field probe
was lowered by 300 mm is displayed in Figure 56. The largest near-field difference
between the CST simulation and the measurement is at 35 and 169 MHz. The CST
result at 35 MHz is 20 dB higher than the measurement. The measurement is 17 dB
higher than the CST simulation at 169 MHz.
The Measurement is 7 dB higher than the FEKO simulation at 292 MHz, remaining
mostly above the FEKO simulation across the bandwidth.
56
Nearfield Magnitude Comparison, X = -300mm
-30
FEKO
CST
Measured
-35
-40
Nearfield Mag. (dB)

-45
-50
-55
-60
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
below
Near-field Phase:
A comparison was performed between the near-field‟s phase for the simulation and
measurement case. The three near-field measurement cases were compared to their
counterpart simulation case to validate that the antenna has the calculated far-field
magnitude and phase.
(a) Near-field phase comparison at X equal to 0mm

The antenna‟s phase measurement results at a height equal to zero mm are illustrated in
Figure 57. The measured phase has a maximum difference equal to 100° and 120°
between 26 MHz and 46 MHz. The measurement has a larger phase between 100 MHz
and 166 MHz and in the higher frequencies between 386 MHz and 426 MHz. The
measurement follows a similar gradient to that of the simulation. It can be accepted that
the antenna‟s phase will be equal to that of the simulation.
Nearfield Phase Comparison, X = 0mm

0
-50
-100
-150
-200
-250
-300
FEKO
CST
Measured
-350
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 57: Near-field Phase of measurement and simulation with X equal to 0mm
57
(b) Near-field phase comparison at X equal to 300mm above attachment point

From 25 MHz to 500 MHz, the antenna‟s measured near-field phase remains mostly
above the FEKO and CST simulation. The measurement decreases below the CST
results, becoming equal to the FEKO results between 288 MHz and 330 MHz, as
Figure 58 illustrates.
Nearfield Phase Comparison, X = 300mm

0
-50
-100
-150
-200
-250
-300
-350
FEKO
CST
Measured
-400
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 58: Near-field Phase of measurement and simulation with X equal to 300mm above
(c) Near-field phase comparison at X equal to 300mm below attachment point

The last measurement that needs to be compared to its simulated counterpart is where
the near-field probe is positioned 300mm below the attachment point of the antenna on
the case (refer to Figure 59).
The results of the measurement and simulations between 185 MHz and 410 MHz are
similar. The biggest difference is at 41 MHz where the measurement is 28° higher than
CST and 43° higher than FEKO. This is caused by the reflection of the ground plane.
The comparison between the simulations and the measurement in the three cases
confirms that the measured antenna and both antenna simulations have similar
frequency characteristics. The antenna will thus meet the design requirements.
58
Nearfield Phase Comparison, X = -300m

0
-50
-100
-150
-200
-250
-300
-350
-400
-450 FEKO
CST
Measured
-500
50 100 150 200 250 300 350 400 450 500
Frequency (MHz)
Figure 59: Near-field Phase of measurement and simulation with X equal to 300mm below
4.4.3 Power handling
√ [5]
With the use of [5] we can calculate whether the inductors will be able to handle the required
current that will be supplied by the powersource. 30 Watt will be supplied to the antenna,
represented by P in in equation [5]. The antenna‟s impedance will be 300 Ω according to
simulations, giving a I(rms)equal to 0.45 A.
When the datasheet of the 1008HT inductors is consulted we see that the 1008HT-R56T
inductor which we are currently using in our design can handle an rms current of 0.24 A. The
inductors have a minimum resistance equal to 5 Ω at 20 MHz, giving a total equal to 30 Ω for
the six series inductors.
Where the parallel resistor is the lowest (200 Ω) the current will have a magnitude of 0.39 A.
The current through the inductors will be at a maximum where the parallel resistor is equal to
680 Ω, with a magnitude of 0.43 A. The current‟s magnitude will decrease from the feed
towards the end of the antenna.
The current inductor would be able to handle the required power needed for a short period of
time, but a parallel configuration of inductors can be used to supply the necessary
inductance and would be able to reduce the current through the inductor, thus increasing
their operational duration.
59
Chapter 5
Conclusion
In this project‟s initial phase several antenna topologies were investigated. For each
topology, the component values and positions were optimised numerically to find the design
best suited for a low frequency, wideband jamming antenna.
The simulations of the different topologies in FEKO revealed some design criteria. The
addition of inductor and resistor loads (RL loads) on an antenna increases the antenna‟s
ability to radiate energy at different frequencies. When a large number of RL loads are
spaced from one another, to avoid inductor coupling, the antenna‟s gain will have an
improvement at low frequency and deteriorate at high frequency.
The top-loaded antenna exhibits a high efficiency because the horizontal side arms allow the
antenna to have a current distribution larger than zero at the top of the antenna. The results
from the antenna with a helix structure, folded structure and top-loaded antenna confirmed
that the antennas performance cannot be improved by extending its length while complying
with the height specification. The proximity of the added wire to the existing wire also affects
the antennas performance at different frequencies because of coupling (inducing of currents
and capacitance) between the wires. The cases of the antenna with a helix and folded
structure illustrate the coupling between wires located in close proximity of each other.
The use of a RL load was studied using small inductive loads in the form of beads which
increased to inductive loads on the boundary of the required operating frequencies (RL chip
loads). The RL load‟s position, magnitude and quantity was investigated on a monopole
antenna and compared to the design in which the antenna‟s length was increased.
The best design topology was found to be that of a normal monopole antenna, with four RL
sections added along the length. Each RL section consisted of six inexpensive, commercially
available inductors in series and one parallel resistor over the inductors.
Using FEKO to obtain the RL values, the realisation of the antenna was handled in two
steps. First a suitable inductor was sourced and the design was checked by importing the
measured inductor parameters into FEKO to ensure that the specifications would still be
met, using a practical ferrite inductor. The second step was to build and measure the
reflection coefficient to test compliance with the VSWR specification. Both the simulated and
measured antenna data meet the VSWR specification over the complete bandwidth.
No gain measurement facility for the required frequency range was available to test the
experimental antenna‟s gain against the specifications, due to the low operating frequency. It
was thus decided to do near-field probe tests and compare them to the FEKO simulated
near-field results. The antennas far-field characteristics would agree with those of the
simulation if the measured near-field agrees. These comparisons showed satisfactory
agreement in magnitude and phase for the measurements at different heights at a constant
distance from the antenna.
60
Therefore it could be accepted that the antenna will have a gain and SWR equal to that of
the simulation. The antenna will have a 10 dB increase in its gain while having a slightly
lower SWR. The components used are capable of enduring the required power handling and
can be implemented in a parallel configuration for longer working periods.
Future work will include the use of components better suited for the required power handling.
Further research on the influence that a human body and an urban environment would have
on the practical antenna must still be performed.
61
Bibliography
[1] Zhang F. F., Sun B. H., Li X. H, Wang W. and Xue J. Y., Design and
investigation of broadband monopole antenna loaded with non-foster circuit.
Progress In Electromagnetic Research C, Vol. 17: 254 – 255, 2010
[2] Hansen, Robert C, "Optimum Loading of Short Whip Antennas", IEEE

Transactions on Vehicular Technology, Vol VT-24, No. 2 May 1975.
[3] Stewart, W.D. “Notes on Modelling Short Inductively Loaded Antennas”,

Lumped Load Models v. Distributed coils, 2004
[4] Harrison, C. W. “Monopole with inductive loading”, IEEE Transactions on

Antennas and Propagation.”Vol 11: 394 – 400, July 1963
[5] Yestrebsky, T. “Application notes 23, MICRF001 antenna design tutorial”, The
infinite bandwidth company, July 1999
[6] Trusk, Chris “Magnetic Materials”, Classic works in RF Engineering, Artech

House Inc, 3 – 6, 2006
[7] Deasy, Richard E and Cedar Rapids, Frequency range enhanced monopole
antenna, Nov 1991
[8] Technical Reference Manual for OMNI-A0124-01
[9] Chip Inductor – 1008HT Series (2520) Datasheet, Coilcraft Inc, 2011
[10] Devoldere, John, ON4UN's Low Band DXing, American Radio Relay
League,1987
[11] Brown, Bruce F., “Optimum Design of Short Coil-loaded High-Frequency

Mobile Antennas”, The ARRL Antenna Compendium, Vol. 1, ARRL,
Newington, CT., 1985
62
Appendix A: Photographs of measurements and Schematics
Figure A.1: Antenna layout
63
Figure A.2: 1008HT-R56T inductive and resistive loaded monopole antenna
Figure A.3: Rohde & Schwarz FSH6 Network Analyser
64
Figure A.4: ADT4-6WT transformer mounted on case
65
Appendix B: Results of SWR measurements

5.5
4.5
3.5
SWR
2.5
1.5 Antenna on ground plane with human body

Antenna on ground plane
Antenna raised without a dragwire
Antenna raised with a dragwire
0 100 200 300 400 500

Frequency [MHz]
Figure B.1: Standing wave ratio from measured and simulated data with phase shift and 300 Ω
reference impedance, antenna on a ground plane with (blue) and without (green) human body
in close proximity, raised from ground plane with (cyan) and without (red) drag wire and
simulation of corner mounted antenna (magenta)

3.5
2.5
SWR
1.5
Antenna on ground plane with human body
Antenna on ground plane
Antenna raised without a dragwire
Antenna raised with a dragwire
0 100 200 300 400 500

Frequency [MHz]
Figure B.2: Standing wave ratio of measurements, with a 300Ω reference impedance and phase
compensation, antenna on a ground plane with (blue) and without (green) human body in
close proximity, raised from ground plane with (cyan) and without (red) drag wire and
simulation of corner mounted antenna (magenta)
66
600
400
Impedance (Ohm)
200
Antenna on ground plane with human body [Real]
-200 Antenna on ground plane with human body [Imag]

Antenna on ground plane [Real]
Antenna on ground plane [Imag]
Antenna raised without a dragwire [Real]
Antenna raised without a dragwire [Imag]
Antenna raised with a dragwire [Real]
Antenna raised with a dragwire [Imag]
Simulation of corner mounted antenna [Real]
-400 Simulation of corner mounted antenna [Imag]
0 100 200 300 400 500

Frequency [MHz]
Figure B.3: Impedance of measurements, antenna on a ground plane with (blue) and without
(red) human body, raised from ground plane with (magenta) and without (green) drag wire and
simulation of corner mounted antenna (black)
67

A Wideband Monopole Antenna Design

Uploaded by

Copyright:

Available Formats

A Wideband Monopole Antenna Design

Uploaded by

Document Information

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

A Wideband Monopole Antenna Design

Uploaded by

Copyright:

Available Formats

What is the focus of the thesis?

What is the focus of the thesis?

What antenna topology was found to be the best overall design?

What antenna topology was found to be the best overall design?

A Wideband Monopole Antenna Design

Thesis presented in partial fulfilment of the requirements for the degree

Supervisor: Prof. KD Palmer

Date: March 2013

Copyright © 2013 Stellenbosch University

All rights reserved

4.1 PRACTICAL MEASUREMENTS 42

FIGURE 1: GMJ9000V PORTABLE JAMMER WITH REFERENCE ANTENNA [8].................................................................... 8

A typical wideband man-pack currently used to deactivate RCIEDs

Since the jammer is mounted on a soldier‟s back, it shouldn‟t

When this 0.9kg antenna is folded to improve mobility it has a

Figure 2: Measured OMNI-A0124-01 Antenna Gain from literature [8]

Figure 3: Measured OMNI-A0124-01 Antenna VSWR from literature [8]

1.2 Thesis outline

 Bandwidth from 20 MHz to 500 MHz

 The voltage standing wave ratio should be less than 3:1.

 It must be foldable to a length equal to or less than 0.85 m

 The feed power handling should allow for 30 W

 It must weigh less than 1 kg

2.1 Monopole antenna loaded with non-Foster circuit

They used a technique presented by Sussman-Fort and Rudish regarding non-Foster

The bandwidth could be increased if the inductor‟s inductance could be changed to

Zhang et al proposed a non-Foster circuit constructed from inductors and capacitors

Figure 4: non-Foster circuit schematic

Figure 5: Reactive value required for an antenna length of

Harrison [4] analysed the loaded monopole using superposition of asymmetrically

In conclusion, the previous authors‟ theories were combined to design an improved

2.3 Frequency enhanced monopole design

2.4 Effect of Ground-Plane on antenna performance

2.5 Realised Gain:

Realized_gain = Directivity X Radiation_efficiency [Def1]

Realised_gain =|s21|²X directivity

Clearly, realised gain is always <=true gain.

2.6 Conclusion from Literature

Zhang et al advised that to achieve an improvement in the antenna‟s impedance and

Simulations and Analysis of Antenna Designs

A good design should thus

Simulation 22 is the final antenna design using ideal inductors of 2 µH.

Table 1: Percentage simulation met the VSWR and Gain specification

Figure 10: Design topologies investigated

Overview of simulations performed:

 Top-loaded antenna with four resonant RL bead sections at varying locations

 Top-loaded antenna with four RL chip sections at varying locations failed to

 Including the 1008HT-R56TInductor S-parameter model into the FEKO

3.1 Construction of Inductors

Figure 11: RL Chip section

Figure 13: RL Bead and Chip Diagram

3.2 Topologies containing Bead inductors

3.2.1 Monopole antenna with RL beads at fixed and various locations

Figure 14: Monopole antenna

Straight Mono with 4 Bead loads at Fix Step

 Antenna with four RL beads: 0mm, 800mm and 1600mm

Figure 16: Top-loaded Antenna with

Toploaded Mono with 4 Bead loads at Var Step

Toploaded Mono with 5 Bead loads at Var Step

3.3 Topologies containing chip inductors