GenRad Experimenter Jan 1937

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

-EXPERIMENTER n GENERAL RADIO

ElECTRICA L MEASUREMENTS
~~
TECHNIQUE AND ITS INDUSTRIAL APPLICATIONS
VOL. XI No.8 JANUARY, 193 7

THE MEASUREMENT OF MUTUAL INDUCTANC E


. MUTUA L I NDUCT AN CE is olle T he general theory of coupled cir-
of the most important properties of cuits ilwol ves at one ex treme the
electrical circuits. By means of it the closely coupled iron-core transformers
world's power is transmuted from gen- just mentioned and a t the other ex-
erator \'oltage to transmission line treme t he loosely coupled tuued cir-
voltage and back to the ,'oilage suit· eujts of radio-fre'lucncy amplifiers, of
a ble for our motors and electric lights. wh ich t he neutrod yne recei\'er of Haz-
Yet the measure ment of mu tllal in . elt ine is an historical example. Such
ductance is of little impottance as com· loosely coupled circuits are now being
pared to measuremen ts of self-induc- used as band-pass filt ers al the inler-
tance, capaci tance, a nd resistance. mediate frequ ency of he terodyne re-
The major reason for this sligbting of ceivers.
mutual inductance lies in the in terest · Mutual inductance is uni(lue in that
ing fac t that iron .core transformers arc il ca n exist only in the presence of self-
nearly perfect. The mutual ind uctance induc tance. An important measure of
helween the primary and second ary mutual inductance is its ralio to the
windings has 80 nearly its maximum geome tri c lIlea n of the two sel f· ind nc-
va lue that no ordinary measure ment of tances it ('"onnects, called the coeffi cient
it could distinguish the minute ,li f- of coupling, k.
ference. Such differences are of course M
k - ~- (J)
of great importance a nd are measured - V Ll L2
by some cha racterist ic of the trans-
former itself, the voltage regula tion of which tUay ha ve any value from zero to
a power transformer or its leakage re- ullity. In loosely-coupled lUlled ci rcuit s
actance, and tbe frequency cha racter- t he resona nce curve has a sin gle peak
istic of an audio transformer. for all values of coupling coeffi cient

ALSO IN THIS ISS UE : New Wave Filters page 4


A Handy Voltage Divide r page 6
A Wide-Range R-F Choke page 7

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
1C88 tban a so-called criti('ai coupling.' Tn-E 107 Variable Indur-tors are cali-
A larger coupling coefficient resuhs in brated in such a wa y that the mutual
a flul1cned top \\ith the s teepness of induc tance bct\lcen their rotor and
the sirles mainttlil1ed. The measurement sllllor windings can be ell!!.ily o!Jtaine,1
of this small mulual induclauce is of Crom their dials ('a librated in self·induc·
cons iderable inlporlance in the adjwH. tance. For a 1% full-scale accuracy
ment of i·f fillers. of this dial ca libration. the error in
.1\ 1u tual iJlduc tance Illa , Le measured the mutual induciance will range from
by using 8S a s tandard of comparison a 2.5% to 10% over the portion of the
mutllal inductance, a self-inducta nce, !wa le normally used.
or a capaci ta nce. When the standard When the primary and secondary
mutual inductance is continuously ,'a- windings of 8. mntual inductance are
ri ahle and of BLlCh a range that it can be connected in series. the self-ind uc tance
made equal 10 t1w unknO\\ll 111111ual iu- L of the pair is
duc tance. the Pelici IUlltual·indllctance L '"" L t L2 ± 2M+ (3)
balance is the simplest me thod. As The mulual induc latJce J\1muy he cal-
E culated from the two self·indu ctances

J l
, La and L. ohtained "ilh the two coils
Hiding and o pposi ng (using the
- si~nll berore 2M),
and +
M s J~ (L~- L.) (4)
M, I Ms
For a coefficient of conpli ng nearly
r unil y where L. is very smH II compared
to L... the error in t he determination of
M is that of L. itself. For smnller coup·
ling I'OCfficients this error increases, as
).'!Gett!! I. Felid "'ut" a l_;nd"ct.~ halance
ah.. a)l~ happe ns II hen the differcnce of
sho\\11 in Figure I, the primaries of both 1\\0 nearly e(I'IalnlllObeNi en tcrs in an)
lIIut\lal indm: tanees, unkno ..... n and
sta ndard. are connected in scries to u
I)()\\er source. Their 5CCOll(larics are
also connected in series to a suitable
oete!!tor, head teJe phoLle5, or other a·c
0l>crated met.er, iu IIlIc h (Iirectioll thai
Iheir illlluced \oltages o ppose. The
s tandard is then \ariell until a balancc
is o bt ai ncd, whcll
.M \ = ,\1.~ (2)
The error in this measurcnwut is e!'-
senlilllly that of the standard , pro-
vided thai the capaciti ve reactances
betw«n the coils of the mutual io-
ductance!! are large L'O IUJlured \1 ith
Iheir mutual reactaoces and that the FIGua!!. 2. TYl' !!. 107 VariaMe Inductor.
impedance of the deteclor is small. Mutual inductance at any !etting is onc. half
the difference of {he .calc reading .nd tbc
'''' - ../li~. or ...., ... vr.L.\~I ... he,e 0. ... ~ DJ ~ .. al oe of k lf·illl.loetaoce 1,11 l:ef O motual in-
u.. d .......... LocI..... v-ipooeo '" "00"1' 'ac,.,.. II) 01
duc tanee &I entered o n the lIan,erlue
.,.ha..-, ...d _ •• ..,..

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
formula . For exa mple. when K = 0.1 , - - -- - - - E-------,
and L, = LEt the error ....,ill be increased
fivefold . The measurement of self·in·
ductance may Le made 011 a Tn-!!!
650·A Im peda nce Bridge with 811 error
of 2 j.th or 2% and 011 a TYPE 667·A Tn-
ductance Bridge wil li a n error of O. J ,uh
or 0.2%. The errors \~hi ch occur ill in -
ductance measurements .... ere discussed
in considerable detai l ill the General
Radio Experimellter (or March, 193-LI
The TYP E 293 ·1\ Universal Ilridge may
also be used with resulwlIl crrors which
lie between those of tbe o ther Lridgcij
mentioned. In brid ges haviug a decade
ratio arm, CrOIll the se ning of '\lhicb the
self.indu ctance is calc ulated, such as FIGUR K 5. Th", Carllpl", 11 l\I"\lIal·l"u"clMIce
t be TYPE 667·A and Tn- I:: 293-A Bridge
Bridges. the illcrease iu I.he error 8S the
two separate bridge ba la nces approach connected inlo the bridge arm 1lI1I y bc
one a uo ther is minimizcd if oTle ra tio mcasured by removing the o ther \\ illil .
arm i8 kept fixcd a nd the cha nge in ing Crom the detector circuit . Denoting
t hese bridge readings by primes.

-
balauce taken up by minimum cha nge8
in the other. The error is then that of U'
- A- 8
thc change in resistance of the decade A'
M ~ L,v (6)
ratio arm. A+ IJ
){utual induc tance ma y be (.'Ompared
B' - 8 ir A ,. A' (7)
with a self-induc tance on the Campbell
mutual·inductance bridge sbo\\'n in
A +
B L,v
Figure 5. Such measuremCll ls are easil) maJc on
the T\"l'E 667·A LnJucta nce Bridge by -
AL" - IJL N
M - (5) connecting olle \\ indillg of the mutual ,..
A +11 inductance to (he unkno\\ II termina ls
Z
c:
The self-inductance Dr of the \\ indin g a nd the olher wi nding in serie8 with ,..
lim detector. The errOfll arc slightly ::a
-<


<
o

><

- z
o

F IGURE 3. TYPE 650·1\ Imped i llce Bridge FICURE 4. Tn·1!. 667.'\ Inductance Bridge 00

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
,-------E-------, feTCllce betwee n the two bridge set·
tings iii reduced by at least It factor of
three and there is a greater chance of
keeping boUI the bridge balances wit hin
It single decade setting l(lwcr tha n the
highest onc used.
A capacita nce may be llsed a8 a
s tandard for comparison with an un-
=

-
=
~
known mutual imluctu llce on the Carey
Foster mutual-inductance hridgeshown
in Fi gure 6.
(8)
The resis Lallce P includes thai of tbe
\\ imling cOlUlcc led in that orlll, which
is somewbat difficult to obtain. This
bridge circuit may be set up on the
FIGl1RK 6. The Cfl.ccy FO!tcr \lutu lll-lmluc- T\TI' E 293-A Universal Bridge, using
t ll l1 00 Uri<lge
8S a capacitance s Lnlldard a suitable
less Ihnll for the pre\,jous case because Tn>E 509 Mica Condenser.
for a given mutual inductance the dif- - H. F. Fn;w

NEW WAVE FIL TERS


• THE R E is a stead y de mand for A filt er for general al.)ll ica tioll , how-
wave filters offeriug a moderate a mount e ver, should be designed 80 that in the
of diserimiuation he tween the dcgired vallcys between the peaks the filter
and undcgired frequ eucici and having never gives less thall a certai n required
a modcra te sharpness of cUloff. For Htlenuation. Prom tbis sta ndpoint tbe
sOUle years the General Hadjo Com- height of the lowes t vn lley measures the
pany has been supplying 'fYl'E 330 " discrimination," a nd if there is more
Fillers for general-purpose laboratory than one va lley the best Jesign will
use. For 11 numlrer of uscs, ho\\ e\-cr, a make t.hem all of the same height. Hav-
sharper cutoff is required, aud afler a iug defined the " discrimi nation" as the
considerable periot.l of supplying in - minimullI va lley height, the "sharp-
di vidually·de igllC"d filters for purtieu- ness" caD be defined a8 the (requency
lar applica tions, the TYPE 830 Filter ratio in which the discrimination is
is offered 8 S a stock item. obtained. Both these (luantitie8 are
1n Jiscussing fi lter performance the indicateJ in the typical filter charac-
quantities of interest are usually the teri8tic of Figure I .
sharpness of clItolf or, more briefly, the Por certa in applications the Ulost
"sbllrpness," a nd the " discrimination" important coosideration is sharpness
or Jilferencc between the insertion loss of cutoff. For other applications a wide
iu decibels for the wanted a nd un- frequency loten -al is avai lable for the
wanled fr equencies. For certain appli- trans ition (rom the pan band to the
ca tiOllS, peaks in the altenuatiOll c urve alleuuation region, but considerable
are important for the suppression of discrimioation is required bet""eell
particular frequencies or harmonica. frequeucies in the two regrous. The coo-
4

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
dilioos on the design which a re ncc-

- essary to obtai.n a sha rp cu toff a re


very di lTereot (rom those which will
gil·e 11 la rge a moun t of discrim ina tion.
It turns out, in (act , that tbe greater
the sharpness which is obt ained by I
proport ioning the clements in a pM -
ticula r configuration, t he less will he
,,j
the discri minat ion.
I n t he Tn E 830 Wave Fill ers the
endeavor has bee n Illude to reach a
com promise bet weeu sharpncSI! a nd d is-
crimination which will be satisfactory
for a large number of applica tioll s. T he ,. f•
•au""..,.
insert ion \OS8 characteristic when the F ICU MK I. In&erlion losa c1Hm, cl c ri~ , ic of a
fil ter is working bet ween cons ta nt filter indica tin g ,,·h. 1 it I" .,au l l,)" •.liilCrimi na.
resistance termina tions is showlI in lio n . nd sharpnetil. The ra tio r;
(or ill rlldp.
rocal, .Iel»ending o n whet hcr ,(Ie filter ill 10.. '
Figure 2. Jt will be seen t hat the
l,uM or high,pIIH) i, II ",ell~ ure of ~ h arl'" eiWI
" discrimi nation" is sli ghtl y over 40
decibels a ud the "sharpness" is such T"PE 830 Wa \' e Filters are available
that t his va lue is reached al 1.5 limes in both high. pas8 and \ov..pass types.
lhe cu torr (or the IO""pass t ype8 or .67 Slock models have eut·orr frequencies
of the cutorr (or the hi gh.pass t.ypes. of cither 500 or 1000 cycles a nd image
High.pass and low .pass t ypes call im pedances of either 500 or 5000 obms.
be used in ta ndem to provide a ba nd · Comple te specifica tions are given on
IlB. SS fi lter ",here tbe interval to be page 11 4 of Catalog J.
covered is one oct a ve or more. Tbe
T,,,,, 830 Fille .... e "'Obul.«u.... ond ... M uDd .... , .. ,.", . uf
a ttenua tioll cha racteristic of 811Ch a 1~~ A_.ica~ T.le',bo ........ Telo-t •• ph C..... p.n1 .., •• 11 f, ..

combination is shown in Figure 3. ",m....Moa , • ..,.....h In_,'p"",,. _~'. """.~.


1• ••..-..... ..... _10..-, ,..ot\ I. I""~ ..... oI'P11eol
- W. N. 'fUTIL E """""....
• · ,.
,
f- " Z
• •
,.c:
• • '"-<
.J\ ,
• · ,
! ,
• ·
• •
<
o
· ·
••
-- - - •
· - -
........ OOCY ...... «u

F,GUHI: 2: ((,,/,). Charnclt:ri~ l iCil or !OOO-cyd.,. l"w-l' ~eI! ut,,1 hi g h_l.a~@ mlUll
.._ .. c.oo - - ><

z
o
FICURE 3 (rig '"). The 5O(l.f'yclc, high ' I'I'iIII and IOOO-t:ycle. 1ow-I 'u U In_ CUll I,,: com l,inl"! 10
r.)rm a Im,,, I-I,,,&ti fi lh'r (:over, ,,S one oel u,'"

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
A HANDY VOLTAGE DIVIDER

=
...
=
'"

.1 N B U I L 0 I N G speeio lizctl labora- SPECifiCATIONS


tory C(l uiplIlCII I and ill circuit develop- Power Dissipation : TYPE 154.A, 6
lIlellt "ork, the engineer often needs walls; TYPE 15·~·B, 12 walls.
un iuexpclIsivc. com pac t. and easily Accuracy of Resistance : ± 10%.
adjustable volLoge divider (or supply- Mounting: Standard 3·bole; see ac·
ing plate Dud grid voltages to vacuum companying sketcb.
lubes. Ln several recently-developed Dimensions : TYPE lS'~·A, lYs incbes
General nadia inslrumcllts, the uni ts (maximum diameter) x 1 ~ inches
shown in tbe accompanying photo- (height). TYPE 154·B, 37'8 incbes
graph have been used (or this purpose (maximum diameter) x 231 inches
and have proved 80 useful Ihat they (height) .
are 00 .... olTered for general sa le. Net WelEht : TYPE 154.A, 6 ouncC8;
Tbe resistance winding and tbe TYPE 154-B, 8 ounces.
moulded bakelite forms are identical
in construction with those used in
TnE 471·A and TWE 314-A Hhe<wllat-
Potentiometers. Three adjustable rOla·
ry contacts are provided, each cover-
ing one-third of the total resistance.
Adjustments are made by loosell ing a
thumbscce\\ and movinl,{ the arm to
the desired (,osi tion, after which the
adjustment is locked by lightening the
thumbscrew.
Ty pe 'rollli Ucsisumce Code Word Price
I M-\ 5,000 11 DlVIDAIOIY $6.50
UH-A 10,000 11 OIVIDBDAT 6.50
151-A 20,000 11 on' locAl'g '.50

154-8 50,000 11 OIVIDIi:YRK 8.50


154-8 100,000 11 OIVlDFACT 8.50
lS-l-8 200,000 11 OU'lDGIRL 8.50

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
A WIDE · RANG E R·F CHOKE
.THE UTI LITY of most general.
purpose radio· frequen cy chokes is
limited to operation o\'cr a compara.
tively narrow range of frequencies.
Ordinary methods of winding produce
Ii high value of distrilmlcd capaci.

tance, while separating the winding


into 8cweral individual sectiolls in series
results in such a large number of reso-
nant frequencies that at least ooe is
usually loca ted in the band where it is
desired to work. The TyrE 119-/\
Radio-Frequency Choke has been de-
signed to e limin ate these faults.
The winding is composed of a large
number of spiral. wound pies, each only
OD C wire wide. By this me thod, the
capacita nce is much more uniformly in duc tance o f 250 IIIh . Thcsc cons ta nt s
distributed than with tbicker pies. IIrc practically indepe ndent of {re o
resulting in only one resona nt point. (Iuener and no minor resonances a re
The efTeclil'c capacitance is ex tremely found up to at least 50 megacycles.
..... low, 80 that the choke is use ful at fairly In addi tion 10 its {unc tion as a choke.
hi gh ra dio frequencies. Lhis unit bas been found ex tremely
The specifications for 'l'vN; 119-A useful as the inductance e leme nt In
Radio. Frequency C hoke ure lis ted hi gh.impedanre low .pass filters.
below. It will be noted that tile capaci-
tance o f 2 j.lp.f is ex treme ly 10 .... for a n
SPECIFICATIONS
Indu ctance : 250 millihenrys.
Capacitance : 2 p.p.{ .
D· C Resistance : 450 ohms.
.z-..
Maximum Current : 60 milliamperes.
Dimensions : Base, l ~ x 1% inches;
...
c:

height, IYs inchcs, o\"cr ·a ll . '"


-<
Mounllng : See sketch.
Net Weight : 3 ounces.


Code Word P rice <
11 9_A IMACE
o
' 1.50

z
o

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988
MISCELLANY
.AMONG RECENT VISITORS [0 order 1.0 insure yonr name remain·
to the General Radio Laboratories was ing on the list. the return card should
Dr. Georg Keinath , Director of the be filled in eomplel.ely and ma lled
[nst rumcnt. Lahoratories of Siemens promptly.
Ilnd I-Ialske. Dr. Keinath is the author
of "Die T echnik E lectrische Mess- • AN INGENIOUS solution to a dif.
= geriilf'," a well-kn own lexi o n electrica l ficult problem is illustrated ill the ac·
= measuring ins lrumenlii. companying photograph.
""
= .THE " R. M . A. ENGINEER , "
The wllee! shown in the photograph
is an e ~perimental part for a new in ·
published by lhe Engineeri ng Division tegrator now under development in the
of the Radio Manufacturers' Associa - Electrical E ngineering Research Lab·
tion made its first appearance with the oratory at :M assachusetla Institute of
Fa ll Meeting of the lR E at Rochester. Technology. Tbe problem is 1.0 measure
This publicAtion, which wi ll carry the friction losses. Since the shaft runs
s tandimlization reports and similar in enclosed jewel bearings. it is not ac-
(lata, as .... ell as technical articles, will cessi ble for speed measurem ent. Fur-
be issued tbrce times a year. Voio,m e 1. tbermore, the power level involved is so
No.1 is a n achievement of which the small as 1.0 preclude the possibility of
board of editors lIlay be proud. measurement by ordinllry means.
• PERIODICALlY it becomes The solution devised by Dr. S. U .
necessary to revise the Experimenter Caldwell and 1Hr. J. J. J aeger of M assa·
and catalog mailing lis t. Shortly after chuselt s Institute of Technology is
you receive this issue of the E;t:peri. simplc. A General Radio Strobotac was
/IIcrller, a post card "subscription " re- se t at a convenient value of speed, and
newal form will reach you. The re\'ised lbe disc was dri ven by a stream of
G.G mail ing list wi ll be made up from compressed air until its speed was
the returns received from t hese canis. above that of the flashing Strobo tac
lamp. The disc was then allowed to
coaSI to a standstill. Whenever its
speed or a multiple of it coillcided with
tlle rundamental or a subharmonie of
the St robo tac fla~h speed, the disc ap·
peared to stand still momcntarily. In-
tervals between these points were
!"illled with a SlOp watch, and a spee(l-
vs. -time curve for the comple te dccel·
eration period was plotted. From this
curve and the moment of inertia, the
fri ctional loss was easily calculated.

30 STATE STREET
GENERAl RADIO COMPANY
CAMBRIOGE A, MASSACHUSETTS
-
BRANCH ENGINEERING OFFICE-90 WEST STREET. NEW YORK CITY

IET LABS, Inc in the GenRad tradition www.ietlabs.com


534 Main Street, Westbury, NY 11590 TEL: (516) 334-5959 • (800) 899-8438 • FAX: (516) 334-5988

You might also like