SUPERHETERODYNE BOOK...Chevrolet Motor Company Crowley Radio Corp. Delco Air lichee Corp. Emerson Electric Mfg. Co. Federated Purchaser, Inc. Foda Radio & Elec. Corp. Ford -Majestic

?die

SUPERHETERODYNE

BOOKAll About Superheterodynes

How They Work, How to Build

and How to Service Them.

by Clyde Fitch

PUBLISHED BYGERNSBACK PUBLICATION

98 PARK PLACE -NEW YORK-10/.---PC;;PCP;XCX:C7a

Sle1.7,'....

IncreaseYOUR Servicing

BUSINESS 25%

Service Manual 'Complete Directory

- all 0

Autom8bile Radio Receivers-

Full Installation and -

Trouble Shooting Guide

Over 200 Pages

Over 500 Illustrations

9x12 Inches

Flexible, Loose -Leaf

Leatherette Cover

$2.50 List

IF you are overlooking servicing autoradios, then you're missing a great dealof business. The auto -radio business

had its greatest boom this past summer andthousands of sets were sold. By this timemany of these same sets require servicingand with hundreds of them right in yourown community, you can build up a goodauto -radio servicing business. In a shorttime you can easily add 25(4 or more toyour regular servicing business.

Every man connected in any way withthe booming auto -radio business will wanta copy of this book immediately. It is de-voted exclusively to auto-: adio servi,"dope" in complete, understand:al-le f,,rrnThe OFFICIAL AUTO -RADIO SERVICEMANUAL contains schematic diagrams,chassis layouts, mounting instructions, andtrouble shooting hints on all 1933 and manyolder model auto -radio receivers. ThisManual contains a "gold -mine" of inform-ation.

List of Sets Covered in the ManualAtwater Kent Mtg. Co.Autocrat Radio CompanyCarter Genemotor Corp.Chevrolet Motor CompanyCrowley Radio Corp.Delco Air lichee Corp.Emerson Electric Mfg. Co.Federated Purchaser, Inc.Foda Radio & Elec. Corp.Ford -MajesticFranklin Radio Corp.Galvin Mfg. Corp.Ge ,eral Electric Co.General Motors Corr,Grigsby-Grunow Co

Chas. Hoodwin CompanyMontgomery Ward & Co.National Co., Inc.Phil,n Radio & Tel. Corp.Pierce -Afro, Inc.Premier Electric Co.RCA -Victor Co., Inc.Sentinel Radio Corp.Spark -W1thington corp.Stewart Radio & Tel. Corp.United Amer. Bosch Corp.United Motors ServiceL. S. Radio & Tel. Corp.Wells -Gar ner CompanyZe Ith Radio Corp.

[Send remittance of $2.50 In form of Check or Money Orderfor your copy of the 1933 OFFICIAL. AUTO -RADIOSERVICE MANUAL. Register letter If it contains cashor currency. THE MANUAL IS SENT TO YOU POST-AGE PREPAID.

GERNSBACK PUBLICATIONS, Inc.

96-98SH PARK PLACE NEW YORK, N. Y.

TheSuperheterodyne

BookAll About Superheterodynes

How ney Work, How to Build and

How Them

by Clyde Fitchcompletely revised

by Robert E. Kruse

GERNSBACK PUBLICATIONS, Inc.

Publishers

98 PARK PLACE NEW YORK, N. Y.

Contents

Chapter

Chapter

1.

2.

PageBasic Principles of the Superheterodyne 4

Signal Frequency Amplifiers 13

Chapter 3. Oscillators and Frequency Changers 17

Chapter 4. Single Dial Tuning Systems 23

Chapter 5. The Intermediate Amplifier 30

Chapter 6. The Second Detector, Audio Amplifierand Power Supply 34

Chapter 7. Practical Superheterodyne Construction 37

Chapter 8. Commercial Superheterodyne Circuits 47

Chapter 9. I. F. Transformer Design 56

Chapter 10. Servicing Superheterodynes 59

Printed in U. S. A.1st Edition

Copyright 1932 by G. P. Inc.2nd Edition

Copyright 1934 by G. P. Inc.

PrefaceTHE following pages were prepared to present in

as simple and clear terms as possible the theory,design and construction of superheterodyne re-ceivers. The purpose is to give the reader a handyreference book and guide that will help him in hiswork, whether he is interested in servicing super -heterodynes or plans to design and build them. Inany event, a thorough knowledge of the subjectwill be found the shortest and most sure route to suc-cessful receiver performance.

The superheterodyne has always been consideredthe supreme type of radio receiver. It is more com-plex and versatile in its actions than other populartypes of receivers, and for this reason, has alwaysbeen found highly fascinating by those who havestudied it. However, a highly technical know-ledge of all of the various components of the super-heterodyne is not absolutely necessary for the prac-tical man; therefore, only sufficient data of thisnature to meet practical requirements are given.

The main bulk of the book treats with modernreceivers of conventional design. As many varietiesas possible of these modern receivers are included soas to give the reader a breadth of vision and not holdhim down to fixed rules. The older types of super -heterodynes, which appeared in great variety sev-eral years ago, are not treated at any length as thecircuits are all practically obsolete due to the adventof modern vacuum tubes.

-The Author.

CHAPTER 1

Basic Principles of the Superheterodyne

The universal adoption of the super-heterodyne method of reception is dueto causes so simple that they can beunderstood by anyone, and may besummed up in the one word-"Cheap-m SS".

While other types of receivers canbe built to equal or exceed the possi-bilities of the superheterodyne, theyinvariably cost more for the same per-formance, and ordinarily use moretubes as well. The reason for thisdifference is also relatively simple andcan be explained in terms readily un-derstood by the non -technical reader.Certainly there is nothing whatever toexcuse the air of mystery which hasbeen woven about the superheterodyne,presumably for reasons of commercialadvantage.

Briefly then, the advantage of thesuperheterodyne lies in the great easewith which high amplification and greatselectivity can be built into a long-wave receiver of fixed wavelength-incapable of tuning adjustment. Sucha receiver is in itself useless at anyother wavelength, and it was a supplyof just such useless long -wave receiversuhich is said to have suggested theidea of preceding them with a "con-verter"-that is to say a device cap-able of accepting a short-wave (orcrdinary broadcast) signal and chang-ing it into a long -wave signal whichcould then be fed into the long -wave re-ceiver, therein to be amplified enorm-ously without difficulty. The advan-tage-the trick, to be explained-thebasic principle, all of them lie in thisconversion -device which changes incom-ing signals to a longer wavelength.Having explained it, the rest of thestory is simple. To make this explan-

4

ation now go to the simplified theoryof the "heterodyne effect."

Basic FormulaThe heterodyne phenomenon not only

occurs in radio and electrical circuitsbut in all other branches of physics aswell. It is based upon the simple factthat when two sources of energy, A andB, vibrating at different frequencies,are combined, other frequencies aregenerated, equal to A + B and A - B.These two additional frequencies arecalled "beats," and while other fre-quencies or harmonics are also gener-ated, this simple formula is sufficientto explain many actions taking place inradio transmission and reception.

As stated above, the action occurswith any form of vibratory energy. Inacoustics, the piano tuner often makesuse of the beat note generated by twostrings slightly out of tune with eachother for tuning the instrument. Inlight, Newton's interference rings arecaused by the same action; namely,light waves of different frequenciescombining and setting up a series ofvisible "beats" or fringes.

The heterodyning phenomenon as ap-plied to radio circuits was first recog-nized and used by Reginald A. Fessen-den and patented by him. (U.S. Pat.1,050,728 Jan. 14, 1913). This inven-tion applied mainly to the reception ofcontinuous wave radio telegraph sig-nals by means of a local oscillator,which heterodyned the received signalsand produced an audible heat note whichcould be heard in a headset.

An extension of this principle result-ed in the superheterodyne receiver.E. H. Armstrong (U.S. Pat. 1,347,885June 8, 1920) made the beat note sohigh in pitch that it was inaudible, yetretained all the characteristics of theoriginal signal, and was amplified by a

+-nol .-- _

THE

high frequency amplifier and then de-tected and amplified by an audio fre-quency amplifier in the usual manner.This method of reception had many out-standing advantages. It was used dur-ing the World War to satisfy the de-mand for a supersensitive receiver thatwould work with extremely small aer-ials.

A study of our basic formula willreveal that the heterodyne acton ac-tually occurs in broadcast transmitters.While it is perfectly correct to con-sider a carrier wave being modulatedby the audible frequencies, as is theusual custom, a clearer understandingof the entire phenomenon can be hadby sticking closely to our basic hetero-dyne formula. By so doing the originof side bands becomes obvious and the*ICI -ion -a the detattor in the receiver(also called "demodulator") is at onceMine_ 3cplained ergis really no differ-iTC-_Wween modu ati and tero-ymng: his act is not thoroughlyecognized by many radio engineers, al -174f

though in the Ultradyne superhetero-dyne receiver developed by R. E. La-cault, and in some more recent re-ceivers using the autodyne principle,the heterodyne action is called modula-tion.

In radio transmission, since radiationof electric energy from the transmit-ting antenna must take place at veryhigh frequencies to be efficient, a high -frequency generator is used at thetransmitter. For broadcasting, fre-quencies frocri 550 to 1,500 kc._per sec -

areiL4ised each ffiftion saving itsown assigned operating frequency.This frequency is called the carrier -fre-quency or "carrier wave". We will callthis frequency f. Now, suppose wecombine with this carrier frequency, f,the sound -frequencies (or music andspeech -frequencies produced in thestudio) and see what happens; keepingin mind that the music_ and speech -fre-quencies range fr-om about_ 50 to 0'00cycles pe..r, secoird: I'Ve will call thesethe audio -frequency band, or just "AB."

1 From heterodyne formula, welearn tha four 1akttc frequencies arepresent; y, the carrier -frequencyf, the audio -frequency band AB, andthe bands (f -4- AB) and (f - AB).

The audio -frequency band AB willnot be radiated from the antenna, be-

- LSUPERHETERODYNE BOOK f ato 14cL

yqrcause its frequencies (50 to 5000 cycles)are too low for efficient radiation. Thecarrier -frequency, f, will be radiated,as will also the frequencies (f AB)and (f - AB). These a r o pro-su e e sire sass-117U--sferri;m the a ove that a groupof frequencies (namely, f, f 4- AB, andf - AB) are radiated from the broad-cast transmitter, having a maximumdifference of f plus and minus 5000, ora total separation of 10,000 cycles, or10 kc. For example, using a carrier of1,000,000 cycles, a band of frequenciesfrom 995,000 to 1,005,000 cycles will beradiated.

At the receiving end the action is re-versed in the detector circuit. For thisreason the detector is sometimes called

5

li

1

1 g

..110111111116. 4,41111

11111111111r41111111

IToOR,

.1 p. V*Fig. 1. Showing rrow a oeat note is pro-duced by adding two currents of different

frequenc es.

a demodulator. I340n this particular casethe side band frequencies and the car- )<'rier are n_ibLed nitheZletrectoranny Eireiodyne action producthe beat frequencies are an exalicate of the original audio freqUeticiesthat we're present at the-transThiTiMr.Qne side band and the carrier are sUf-&tint t9 pr. ce 4tlati that is,when the carrier f is subtracted fromthe upper sideband (f AB) in accord-ance with our basic formula alone

The lower side band f - ),mixe- with the carrier, will give thesame result. In actual figures the re-ceived carrier wave of 1,000.000 cyclesis mixed with either the 995,000 sideband or the 1,005,000 band and pro-du : .010 c cle beat in the detector

( Kmixer circu Peraor4_7011M; i sFigure r slibws graphically how two

different frequencies, A and B, play be

1:41,12C641/:.` liRicetf 4Z1 /is is, /My ,wtsTb5 dfjovu_C2C7-Er 6t-44ge- 6 6(45R-41" Penz.270V-REcTI-04 fir/ A,

5e-

-itere,op/(7-6- THE SUPERHETERODYNE BOOK

t

4

added so as to produce the result shownat C. When the two frequencies A andB are in phase, the result at C is anincrease in amplitude; whereas whenthe two frequencies are out of phasethe result at C is zero (when the am-plitudes of the waves are equal) be-cause at this instant the -two frequen-

effect is accomplished. In circuit A, Fig.2, an electrical current having a fre-quency represented by the curve A inFig. 1 is generated; circuit B, Fig. 2,has a frequency generated in it repre-sented by the curve B in Fig. 1; sincecircuit C is inductively coupled to cir-cuits A and B, currents of both fre-quencies are generated in circuit C and

in the usual manner so that the upper /half cycles are amplified more or ,less than the lower half cycles and abalance no longer exists and a frequen-cy represented by curve M -N appears/in the plate circuit of the detector tube.,

It is interesting to notethat-curve Cin Fig. 1 is identical in shape with a

cies oppoie alkl,ncRtralize each other. curve obtained by modulating a higherFigure 2 shows electriallir how this frequency with a lower one (M-N)-\ifurther illustrating that heterodyning-

and modulatingare two flair -lei for thesame action

n aco tics we can hear the beatnote set up by two vibrating stringsslightly out of tune with each other be-cause of the non-linear action or de-

-tection characteristics of the humanear.

In the heterodyne action at the radiotransmitter, the vacuum tubes presentin the circuit cause the heterodyningand mixing of the frequencies becauseof their rectifying or detecting_Eharac-teristics.

FINES 401.1 -7Irpt Fh aes.I1 26c,

Mt" Explanation of Simple HeterodyneLE -At ,e,1 Radio Circuit

C' -4K G9

GNER amit F Litepe.ttlir.'

p14 -

CSC .60(GENERATOR fire -9

I/ sb A itu.inbfa:N.-aro o

FIG.2 NvLyo

re)

-c-

or19

.

a lectrical circuits for mixing frequencies.

the resultant is as shown at curve C inFig. 1.

irce,Tu Strange as it plgyseem, no_beat_eurz.

f5Arelgit- ix g_rs1i uit-C, although in

o S Ott the...14)2.its,1 a _heat cur -01 ,v,74vE rgnt_is_fxrangejlaly...said to_existin, this

4.1 p5 ciggia, If we had a lamp in circuit Cificia would light due to the presence

n'Ill' of the alternating current, and tune the....

I eL,ikf circuit by means of a variable conden-Itrai-4 ser, we would find that the lamp wouldIpee.. light only when the circuit is tuned to

sr; the_frequentroir eiflief-circuit A or cir.rtrri- euit B, but it will not light when tuneti jeSto a frequency equal to the sum or dif- I.

-63 p.ference of frequencies A and B. Th -:-:7' A - son TEiniiis-is-that-tha-ursper-hilf- Aiii6L Ycles and the lower half -cycles of thegkOla; heterodyne or beat current present in

itig curve C, Fig. 1 exactly_ neutralize each44. other and the beat frequency represent-

ed by the curve M -N, drawn throughetrii?; the peaks of the waves, cannot be .iso-

ktted th'Siorl-tkmplemsthod, It canonly crated -lb non-linear ampli-

ti

Before studying the superheterodyne,suppose we first become familiar witha simple heterodyne receiver. Figure 3is in effect similiar to Fig. 2 with theexception that it applies to the hetero-dyne reception of radio waves and con-tains a crystal detector and head -setfor detecting the audible beat note.The series of curves illustrating thetheory are shown in Fig. 4. Referringto this illustration, let us assume thatcurve A represents a dash as is madeby closing the transmitting key for agiven period of time and then lifting itas shown on the curve. Let us as-sume that we are located at some dis-tance from the transmitting station atwhich this is done, and further that weare equipped with a radio receiving set,the circuit for which is shown in Fig.3. The upper section of Fig. 3 shows aconventional crystal receiving set. Inthe lower portion of the figure, we haveincluded a generating device for pro-ducing an undamped radio frequencyvoltage, which may also be appliedto the crystal receiving circuit asstation r detection\by using a detector shown. We are not at this particular

1401 1711434AT A Mtain. -117 etediirR-- REC.71-(1ER I.Littr A Cc30*E.1,vxA- Itec-ret*TEX (Tu.bE" e a Natt:.) Oiet V EStfrettfcf. a r).- A , C.5 '

074e0AA/C:

'11iE atvf No NE-TE (ckANit fRE9) vidEV te. \ Al 4 .rIkt

Alf,t14* -MP Kr -47,-6 pe4c45 lir..ces7;74..""mn4.61pr 5E-yrz Mt) rtlicyer mpgr?"46cc CResir4fArdo)

4

g-tortr. 444A ber-feaseer7

kcioit.:7icejar joltFEFINETEITIWEBOOK

time concerned with the details of op-eration of this local generating circuit.It might however very probably be a

Tu. Nee dfr FRO?

-ficafgartothe operator at the tra smi ting stationis transmitting dots and dashes in ac-

e cordance with the Continental Code,then we will only hear a series of clicks 17

at the beginnings and ends of the dotsd dashes. Often due to the presence 4

f other noises on the head -set, thesecan not be readily identified.

From the above discussion, we canreadily see that an ordinary crystal re-ceiving set is not suitable for receivingradio telegraph signals transmitted bya continuous wave station. This would

Local frequen* also be true if a vacuum tube receiverGenerator: of the type ordinarily used for radio

rtaPAIMr.': 1' broadcast reception were used. WefeaANcE must, therefore, include something els*

Th Ff4F in the set, if we are interested in MO-ej-ifVf-6',A145111,144rY ceiving signals of the type shown tit

Fig. 3. A simple het, Pocagt wanY7r curve A, Fig. 4.ri'arrr

vacuum tube oscillator so constructedand so operated that the beat frequencyproduced by it could be varied at will.

For the purpose of our explanation,let us assume that the frequency pro-duced at the distant transmitting sta-tion is 500,000 cycles. Suppose nowthat before we start the local frequencygenerator shown in Fig. 3 we listen inthe head -set of our crystal receiverwhile the operator at the distant trans-mitting station closes the key, trans-mits a dash and lifts the key to con-clude the dash as shown in curve A,Fig. 4. Before the incoming signal ar-rives, no current will flow through thehead -set. When the signal does arrive,a voltage shown in curve A is appliedto the crystal rectifier system and asteady current, (curve B, Fig. 4) willflow through the head -set.

When the key is closed at the trans-mitting station, a slight click will beheard in the head -set. This is due tothe rise in the value of current throughthe head -set due to the application ofthe voltage as shown at the beginningof curve B, Fig. 4. After the currentthrough the head -set reaches this newvalue, it remains steady until the keyis lifted at the end of the dash. Thissteady current will maintain tensionupon the receiver diaphragm, but as thecurrent, and therefore the tension, doesnot change, no sound will be heard untilthe key is lifted, when another clicksimilar to the one heard at the begin -

dosed hereromp nay opened here

(61

X/

bailey, of onownnly oonfinwouo we

Current Mho hood set in crystal rece,rrof set(Ora/ c listor nor operoriny

oo "'Apse::: jean-dZio or from fool

pP ESSCV it*

Yo/ apt .TPrease d tz-,4x,t,byp 71 ritrctrz <041.0; AVDff p Keg e

ACurrent ?two bead set by oennwnetwon

of incoming and loco/ frovionews.

Fig. 4. Graphs showing the currents pre-sept M Fig.014Wi, rd/'tAet

Let us now throw into operation our,local generator and adjust the frequen-cy produced by it to have a value fair-ly close to that produced by the distanttransmitting station, say 501,000 cy-cles. Curve C of Fig. 4 shows the volt-age which will be impressed upon thedetector from this local oscillator alone.Since we can control the design of our

a\141 ning of the dash will be produced. If generator, the voltage impressed uponMOON', wAgN 2 raps. AKt ii(Li), A Datcrog At tato be usgrx.5E4 pa, iv altDfA, to- tOfriAr Rt ciutigif tie Mt% CITIAMAKIftgal

(flit Ift pia:*M

SuipeAker cmcwr) Iru. at bit A$ DErt-GRA, 410iii4goo ee'

THE SUPERHETERODYNE BOOK

the detector from it can readily be madeconsiderably greater than that impress-ed by the incoming signal from the dis-tant station.

Let us now assume that the opera-tor at the distant telegraph stationagain transmits a dash by means ofthe key. We will now have present inthe crystal detector circuit not only thevoltage impressed by the local oscillatorof 501,000 cycles but also the incomingvoltage of 500,000 cycles. These twovoltages when added together will givea voltage such as is shown by curve D,Fig. 4. You will note that this is nowa eine modulated wave. The variationsin amplitude due to the mixing of thetwo frequencies are of a frequencyequal to the difference between them,namely, 1000 cycles.

The application of a voltage to thecrystal detector as shown by curveD, Fig. 4 produces a current throughthe head -set such as is shown by curveE, Fig. 4. Note that until the key is

amplit de of the into ng signal volt-wi the amplitude

of the loca oscillator voltage can bemad manImes this value. You willreadily notice plitude of the1000 cycle tone produced in the head -setby heterodyne reception as shown bycurve E is considerably greater thanthe amplitude of the "clicks" occurringat the beginning and end of the graphas shown by curve B. This is due tothe effect of the local oscillator volt-age. Therefore, by the use of hetero-

ne rece. o " %-atitafirt. u e atcatiowitoour si The am-plitude a-thetitirigial in thehead -set from the incoming voltagealone is usually assumed to be propor-tional to the square of this voltage.However, the amplitude of the signalproduced in the head -set by the com-bination of the incoming voltage andthe local oscillator voltage is propor-tional to the product of the two volt-ages. Since the local voltage may be

The basic superheterodyne layout. Note that it is divided into six componentsReading from left to right the various boxes contain the following. "Signal fre-quency amplifier" is a T.R.F. amplifier adjusted to the wavelength of the transmittedsignal. "Frequency changer" is a rectifying amplifier, or "mixer" or "first detector,"and in some cases may be combined 1n the same tube with the "oscillator." Therest of the system, to the right of the label "Fig. 5" is simply a long -wave T.R.F.coif Flo*, IA vatraiverwmpeinigivii-kway, incapable of bripg;v47

rkg..- closed at the transmitting station, the many times greater than the product-r54"1" current through the head -set is steady of the two it may be many times great -got and the receiver diaphragm is undert6(1- ;'constant tension. This steady current

/ is produced by the local oscillator.When the voltage impressed upon thedetector begins to vary in amplitudedue to the effect of the incoming signal,then_ we have the current through the

rhead-set varying as is shown in curverE.- These variations will take place at

a frequency of- 1000 cycles, and thiswill cause the head -set diaphragm tovibrate at this frequency. Since 1000cycles is well within the range of au-dibility, a 1000 cycle tone will be pro-duced in the head -set.

There are a number of important de-ductions which can be made from thecurves shown in Fig. 4. In general, the

Sr 5t5 4r.".r r It

er than the square of the incoming volt-age. Heterodyne reception possessesthis great advantage over other typesof reception-the process of hetero-dyning in itself introduces great am-plification. It is limited, of course, bythe maximum amount that the detec-tor can handle without overloading.

The Superheterodyne

It only remains to carry this processone step farther to produce the super-heterodyne action. To do this, adjustthe local frequency generator so as tomake the beat or difference frequencyinaudible-say 175,000 cycles instead of1000 cycles, and replace the head -set

THE SUPERHETERODYNE BOOK 9

with a vacuum tube amplifier designedto amplify the 175 kc. current. Thisamplifier is called the intermediate am-plifier. The output from the intermedi-ate amplifier is then fed into anotherdetector (the second detector) afterwhich the signal may be heard in thehead -set, greatly amplified by the in-termediate amplifier. When receivingtelegraph signals from a continuouswave transmitter, only the key clickswill be audible in the head -set when us-ing the superheterodyne method of re-ception, because in this case the beatfrequency is above the audio frequencyrange. When speech or music is beingbroadcast, however, this will be heard.

Instead of using just a head -set con-nected to the second detector the mod-ern superheterodyne employs an audiofrequency amplifier and a loudspeaker.

Summing up, we have the basic sup-erheterodyne layout as illustrated inFig. 5. In a modern receiver it com-prises first, a signal frequency ampli-fier. This is simply a tuned radio fre-quency amplifier designed to amplifyat the signal frequency. It is the sameas the R.F. amplifier used in a tunedR.F. receiver. Its main purpose is toimprove selectivity and eliminate "im-age frequency interference", which willbe discussed later. In addition, the sig-nal frequency amplifier gives us a cer-tain amount of gain and therebyreduces the amount of amplificationnecessary in the intermediate amplifier.In some superheterodynes a signal fre-quency amplifier is not used.

The output from the signal frequencyamplifier is passed into the frequencychanger, or first detector, as shown bythe arrow, where the signal current ismixed with the current generated bythe local oscillator (also shown in Fig.5) and a resultant current of a differ-ence or beat frequency is produced be-cause of the non-linear characteristicsof the first detector, as was previouslyexplained.

The oscillator is simply a vacuumtube oscillator of constant amplitude

!output but whose frequency may belvaried over a wide range. The outputof the oscillator is cou led to some por-tion-orthefiis etector cricifir.

Next comes the interitialifelrequen-cy amplifier. This is a vacuum tubeamplifier designed to amplify the cur-

rent of the intermediate or beat fre-quency delivered to it by the first de-tector. This amplifier is a fixedT.R.F. amplifier, as the intermediate'frequency is always kept constant, byvarying the frequency of the oscillator,rEgear-dl-eigof the -frequen--c-r-iir theincoming signal.

The second detector and audio fre-quency amplifier follow the intermedi-ate frequency amplifier, and are de-signed in accordance with the generalpractice used in other types of radioreceivers and therefore need not bedescribed in more detail here.

Now that we have an outline of theentire superheterodyne, we can startwith the broadcast station and show inactual figures the entire heterodyne ac-tion, from start to finish, all based uponthe fundamental formula given in thefirst part of this chapter. Suppose westart with a broadcast station having acarrier frequency f of 1,000 kc. Fort1,, fsimplicity, we will assume that a sin- r'imp-,gle sine wave _audio note of the high -1,est frequency for which the station is '''Iiro,designed is being transmitted. This to K6-11trwill be a note having a frequency of k1'5 kc. We will call this fl.

There will be present at the trans-mitter the following frequencies:

(1) f __1,000 kc.(2) fl 5 kc .-4160(3) f 4- fl 1,005 kc. f pe(4) f - fl 995 kc.

of which frequencies (1), (3) and (4)will be radiated and reach the receiver.

At the receiver another frequencywill be generated by the local oscillatorhaving a frequency of 1,175 kc. Wewill call this fo. In the first detectoror mixer circuit fo will combine withthe above frequencies (1), (3) and (4)so there will be present at the receiverthe following:

(5) fo 4- f 2,175 kc.(6) fo f _____ 175 kc.(7) fo (f + fl) 2,180 kc.(8) fo - (f fl) 170 kc.(9) fo (f fl) 2,170 kc.

(10) fo - (f - fl) 180Because of the selective or filtering

action of the radio frequency trans-formers of the intermediate amplifier,which are tuned (broadly) to 175 kc.,only frequencies (6), (8) and (10) ofthe above group will be passed throughto the second detector. In other words,

r

10

t..

...-. rizrE-g Era & `"-Zr'"

b t., ieo Al A? r 64,-,Ver-re-relrm

HE SUPERHETERODYNE BOOK p f 0 pg. c I r< ca,

frequencies of 175, 170 and 180 kc.reach the second detector, where theyare ti2j=4,_ and by further heteE9ti..neaction produce the following:

(11) 175 170 7, - 346 kc.(12) 175 - 170 5 kc.(13) 180 + 175 355 kc.(14) 180 - 175 5 kc..(15) 180 4- 170 11)4 350 kc.(16) 180 - 170 10

AERIAL--+

5 KC.R.F. AMPLI-FIER ANDMODULATOR

(MIXER)

LOOS1000

595 KC.

KC

1000K

RADIO TRANSMITTERFIG.6

R.F.OSCILLATOR

The frequencies present in a 1000 kc.broadcast station.

Of these frequencies, only (12), (14)and (16) will pass through the audiofrequency amplifier to the loudspeaker.In other words, a powerful 5 kc. signalwill reach the speaker correspondingwith the original 5 kc. note (2) thatwas present at the broadcast station.The 10 kc. frequency (16) appears as asecond harmonic. It will not appear ifone side band was suppressed, as (16)above would not be present. This en-tire action is graphically illustrated inFigs. 6 and 7.

These figures do not include all ofthe various frequencies and harmonicsthat will be generated due to the mix-ing process, but they include the im-

portant ones. Of the frequencies thathave been conveniently dropped in thisexplanation, because of selective filter-ing, we will have more to talk aboutlater, as they are sometimes the sourceof trouble in actual practice and man-

- ifest themselves by heterodyning withother frequencies in the system andproduce annoying audible squeals in theloudspeaker. They will be discussedin more detail in the chapter on "Ser-vicing".

The AutodyneThe autodyne circuit is one in which

a saving of one tube is effected by com-bining the first detector and oscillator.It is necessary, however, to draw asharp distinction between two mannersof doing this thing with totally differ-ent results. The older, and altogetherbad, method is represented in Fig. 8.Here the triode is attempting the im-possible task of tuning to the samefrequency as the incoming signal (forbest gain) and at the same time de -tuning from it by 175 kc.-if that hap-pens to be the frequency at which thefixed long -wave intermediate amplifierworks. This is an absurdity, and itis evident that one needs to associatetwo tuned circuits with the tube's gridso that one may be adjusted to the in-coming signal, and the other be off-set by 175 kc. (or whatever our inter-mediate frequency may be). The dif-ficulties of doing this in a triode of thedirectly -heated sort shown in Fig. 8

are considerable, although these diffi-culties were mainly evaded in the in-genious bridge arrangement of Fig.11, due to the original author of thisbook And here shown partly for his-toijatil interest, partly because it isstill a most excellent arrangement forreceivers using battery -heated filaments

150 KC.2170 KC.

170 KC.31.80 KC

175 KC.2175 KC. 345 KC.

1005 KC.

t,-.1,4,

5 KC -1--.5 KC355

SIGNALFREQ.

AmP'F'R CHANGER OSCILLATOR AMP F

150 KC.

2.12.DET.

KG.AUDIOAMP

LOUDspKR1000 KC. 170 KC. c,5K

10KC.995 K4. 175 L< 350 KC.

10 KC.SUPERHETERODYNE RECEIVER

FIG.7

The various frequencies present in a superheterodyne receiving the station illustratedin Fig. 6.

to be with us, too.)

Food, Decorating, Home Building and improvement,

Gardening, Travel, Health, Money Management, Family Cars

..

.they're

the features that made Better Homes and Gardens the world's most

popular full size magazine during its first fifty years, and they're

the features that keep it so popular today!

Won't you accept this special Invitation Offer today?

Simply detach

your Invitation Savings Token and slip it

into the "Yes" pocket on

your Savings Certificate.

Detach the certificate, seal in the postpaid

reply envelope provided, and mail today!

That's all -- there's no need to send any money now.

We won't bill

you until after you've received your

first issue.

And be sure to note

the special guarantee printed on your Savings Certificate.

If that

first issue in any way disappoints you, it won't cost you a penny!

Return your Invitation Savings Token in the "Yes" pocket of

your Savings Certificate today!

Cordially,

Barbara Joy

for Better Homes and Gardens

-11-V(41° -77-7 cfro

, 05'6 256 25-e2frbirr) Co P. wiTte ..5`©o 64( ToArr/. /ftact;?..,),- ptf --Tictory e -rz' 7 ?"xt.)47.c 74756_

THE SUPERHETERODYNE BOOK ' .5 11fic4V 440

of the high -economy type working fromdry cells. It is covered by U. S. Pat-ents 1,667,513 and 1,762,221. Circuit Ais tuned to the incoming signal, anddoes not affect circuit B (tuned to theoscillator frequency) because the "re-turn tap" of A IA_placed at the centerof B, and magnetic &Uplink betweenthe two coils is avoided by shielding orcareful placement of the coils.

Complex Autodynes

In Fig. 9 is shown another ingeniousmethod of combining the mixer andoscillator functions in a simple tube ofnormal construction. This is due toE. V. Landon and the circuit is thatof the Majestic type 15B receiver. Thedifficulty with this circuit is principal-ly in the fact that there is some inter-locking of the various trimming con-densers, i.e. adjustment of one slightly

FIG. 9

-71 24

I 1415 i l I

R2' 11Cl

6.0.1 COD.C1 --Ganged tuning conder ers with trim-

mers.C2-Trimmer of first I. F. primary,acting as coupling to R.F. coils.C3-Series trimmer for oscillator.C4-Trimmer of first I. F. secondary.C5-Trimmer of second I. F. secondary,disconnected with switch in "local"

position.cs I. F. cathode bypass.-Audio coupling, .01 mf. mica.

CS-Cathode bypass, 900 mmf.RI and R2-Volume control, 10,000 ohms,and 350 ohms.R3-Decoupling resistor feeding first de-tector, 2,000 ohms.R4-30,000 ohm dropping resistor.R5-"Bleeder" 25,000 ohms.

also

herso 7;Pia

The autodyne circuit.

upsets the others, to the occasionalbafflement of the Service Man. Thiscircuit will now be described in somedetail as it is still of the utmost use-fulness for indirectly heated tubes ofthe tetrode (screen -grid) class (andworkable for triodes of that type also),likewise the set illustrates a numberof points which are of importance in

51-S 24

C6 Rio

8M5 1 a NIF

47

L16;LIS ./

COILE

511

11,6-Second detector cathode bias resistor,, 40,000 ohms.

R7-Second detector plate feed and coup-ling resistor, 0.3-megohms.

RS-Audio grid coupling resistor, 0.3-meg-ohms.

R9 and R12-Audio bias voltage divider, .2and 1 meg.R10-R.F. grid -filter resistor, 0.1-megohm.1111-Tone control.R12-First Detector cathode bias resistor,10,000 ohms.Ll, L2, L3, L4, L5-Tuned input system.L6, L7, LS-Oscillator coil system.L9, L10, L11, L12-I. F. coils.L13, L14-Output transformer feeding mov-ing coil L15.Ll-Speaker field used as filter choke.

The selectivity curve for this receiver is shown in Fig. 10. The sensitivity of thistype of circuit will obviously vary with frequency, and in a typical set ran from 20microvolts at 1,500 kc. to 60 microvolts at 550 Ice.

12 THE SUPERHETERODYNE BOOK

600 KC

11100 KC. --....

'Kw

t000

A'I.F L.F

100

10

1

Selectivity curves. The curves are shownin the usual manner, the lowest point be-ing at resonance and the two extremes30 kc. off to either side. The height ofthe curve ind'cates the ratio in whichinput must be increased to maintain thesame output which was obtained at reson-ance. This is, of course, an indicat'onof the degree of discrimination against

unwanted signals.any receiver and may well be touchedon here.

As we have said, it is easy tomake a detector oscillate - but itis quite another trick to tune thedetector to a signal and at the sametime keep it oscillating at a frequency175 kc. (175,000 cycles) removed fromthat frequency. This is done in theMajestic set by use of two independenttuned circuits-one of which invites thedesired signal into the tube, while theother attends to the business of oscil-lating. The coils Ll, L2, and L3 arethe antenna coupler. LS is the tunedcoil of this arrangement and is tappedinto the primary of a second tunedtransformer, L4 -L5. The whole thingis what was once so passionately ad-vertised as a "band-pass input" by sev-eral advertising departments whosenames we don't remember. In plainradio -language this is dual -preselection,made necessary by the distressing easewith which a type 24 detector over-loads. In order not to aggravate thattendency, the volume control leaves thedetector tube alone and controls byshunting the antenna coil, and bychanging the bias of the intermediatefrequency tube which is of the Ballan-tine-Hull variable -mu sort (Majestictype G -51-S) and therefore unworriedby such operations. Now we have thesignal in the input grid of the detectorand need only to mix with it the 175kc. off -tune oscillation. This oscilla-

tion is provided by the coils L6, L7 andL8 which constitute a T.R.F. systemoperating 175 kc. off -tune. The coil L6is the. feedback coil and has been placedin the cathode -lead to avoid confusingthe tuned input system. Coil L8 cou-ples the plate to the oscillator tunedcircuit L7, which has the usualseries and shunt adjustment conden-sers. The series one, C3, is in turnshunted by a combination consisting ofthe coil L8, the primary of the firstI. F. transformer L9, and the I. F.trimming condenser, C2, connectingthem; thence through the by-pass con-denser to chassis-which completes thecircuit back to the other side of the ser-ies oscillator trimmer. It will be seenthat I. F. primary trimmer, C2, servesalso as a coupling between the oscil-lating feed -coil, L8, and the detectorplate. Because both the oscillator andthe I. F. system make use of the oscil-lator trimmer C2, the two adjustmentsare not independent, though the choiceof sizes is such as to make the inter-locking small. The plate supply of thedetector comes through the I. F. pri-mary, which may be regarded as anR.F. choke, if one wishes. The restof the set is orthodox and one need butexplain that the switch Sw. providesa "local" range by disconnecting C5which detunes the I. F. stage andhence intentionally spoils the efficiency.

Multiple -function TubesOf late, multiple -function tubes have

appeared which are not normal tubesused in two manners like the above,but instead are simply one glass bulbcontaining the parts of two separatetubes-an oscillator and a "mixer -de-tector", ordinarily in the form of a5 -grid tube which is therefore calleda "pentagrid converter." To introducethese here would be confusing, andthey are withheld for a later chapter.

CHAPTER 2

Signal Frequency Amplifiers

Since almost any desired degree ofamplification may be produced in thefixed -frequency intermediate ampli-fier, it may appear senseless to use(also) a tuned radio frequency ampli-fier ahead of the "mixer" (converteror first detector). However this am-plifier does appear in all of the bet-ter modern superheterodynes in theposition just mentioned, which is thatindicated farthest to the left in Fig.5 by the box labeled "signal frequen-cy amplifier" (which is merely an-other name for a T.R.F. amplifier).

This amplifier serves three verydefinite purposes:

1-It greatly improves the quiet-ness of the receiver, since it amplifiesthe desired signal more than thenoise -background, before either ofthem reaches the "mixer" (labeled"frequency changer" in Fig. 5.).While recent mixer tubes decreasethis problem somewhat as explainedin chapter 3), any improvement insignal-to-noise ratio is worth while,especially at short waves where thisis the normal limiting factor of recep-tion.

2-Similarly, the T.R.F. amplifier(also called pre -amplifier), by favor-ing the desired signal, decreases thepossibility that a strong off -tune sig-nal will arrive at the mixer withenough strength to overload that tube,thereby "cross -modulating" the desiredsignal. Should this take place, thetwo will be detected together andthereafter no degree of selectivity inthe I.F. amplifier can separate them.Therefore, selectivity ahead of mixingis important also for this reason.Note that the desired selectivity canbe accomplished by pre -selectionalone, amplification being non -essen-tial. Hence, the T.R.F. tube may beomitted and a scheme of two cascaded

13

tuned circuits used such as that shownin Figs. 9 and 10. However suchschemes, without exception, causesome loss in the desired signal, sincetheir tuned circuits unavoidably havesome resistance. In consequence, thesignal to noise ratio mentioned initem 1, above, is damaged or reducedsomewhat. For this reason, the moreexpensive sets will invariably use acomplete T.R.F. amplifier, not merelytuned circuits in cascade.

3-Finally, the use of 2 or moretuned circuits ahead of the mixer, in-stead of one only, can easily preventstill another type of interferencecaused by off -tune signals operatingthrough the effect known as "image -frequency interference." This effectcan be explained as follows: referringback to chapter 1, you will recall thatin a receiver with an I.F. amplifiersystem tuned to 175 kc., we could re-ceive a 1000 kc. broadcasting stationwhen the oscillator was tuned to1,175 kc., since(osc.) 1,175 kc.-(signal) 1000 kc.

-175 kc.and accordingly, the incoming 1000kc. signal is converted to 175 kc.and passes easily through the I.F.amplifier. BUT unfortunately itis also true that

Antenna input circuit for uniform gain.


Fig. 13. The well-known curves of selec-tivity: A, that of a single circuit; B, thatof three. The wider the bottom of thecurve, the more opportunity for "cross -

modulation."

(signal) 1,350-(osc.) 1,175=175 kc.Hence, the same oscillator settingwill likewise convert a 1,350 kc. sig-nal to 175 kc., and this, also, willpass through the I.F. system. Sharp-ening the I.F. system's selectivity doesno good in preventing this effect.Changing the intermediate frequencymerely causes it to appear at anotherpoint; always we are faced by thefact that there will be two channelswhich are "tuned in"; these two sta-tions being in every case separated bytwice the intermediate frequency.

Protection against this effect mustbe provided by selectivity ahead ofthe mixer-it is ineffective after. Itmay manifestly be provided either bya full T.R.F. amplifier, or else bycascaded tuned circuits without R.F.tubes as in Figs. 9 and 19 alreadymentioned.

Design of the Pre -amplifier(550-1500 kc. type)

In the simpler receivers intendedonly for the normal broadcast band of550-1500 kc., or a trifle more, oneordinarily attempts to secure fairlyeven amplification over the tuningrange of the pre -amplifier, for obviousreasons. Enormous masses of litera-ture have been written on flat -gainT.R.F. amplifiers (refer to the files ofProceedings of the I.R.E. for good ex-amples) and we can mention here,only briefly, a few methods.

In Fig. 12 the natural tendency ofthe amplifier gain is to fall off atlonger wavelengths, because the tun-ing capacity C, is then larger and thevoltage across it correspondinglylower (an explanation for this effectwill be found in any standard radiotext book). It is a tendency univer-sal in T.R.F. amplifiers and painfullyprominent in older T.R.F. sets. It maybe compensated for in several ways.

A-We may make Ll (Fig. 12) ofsuch inductance that with the antennaconnected, it will resonate at about430 Ice. (700 meters). As we tunetoward the long -wave (500 kc.) endof the broadcast band, we approachthe tune (natural frequency) of theresonant antenna and the gain goesup. This is the high inductance an-tenna coil scheme commonly used.

B-With the above scheme, oralone, we may use __the device alsoshown in Fig. 12, of feeding the "B -plus" to the R.F. amplifier tubethrough a choke coil, L4, which isresonant at about 430 kc. (700 met-ers), and coupling the tube to the

A signal frequency amplifier circuit. Th s requires a four gang condenser and isvery selective.


tuned grid -circuit of the next tube,by means of a very small capacity,C1, usually not more than 10 micro-microfarads (.00001 mf.). The ac-tion here resembles that just de-scribed; as one tunes toward 500 kc.,the tube shown acquires a load ofhigher impedance (as L4 approachesresonance), hence its output voltageincreases. The deficiency of thisscheme (which is due to the late CarlTrube) is that the tuned -circuit, L3and C, is loaded with' more fixedcapacity than would be the case fora two -winding transformer method ofcoupling. In consequence, the sametuning range 550-1500 kc. ordinarilyrequires a 500 mmf. tuning conden-ser instead of a 350 mmf. unit, whichdecreases the 1500 kc. amplificationsome 30 per cent in each stage. Thisis not a serious loss, especially in asuperheterodyne, where the I.F. sys-tem has "gain to burn." It is mere-ly mentioned to anticipate carpingcritics.

Short-wave Pre -amplifiersIn a superheterodyne receiver

working at short waves the pre -ampli-fier or signal -frequency amplifier ismuch more essential than in a setoperating on the 550-1500 kc. range.Its omission in many receivers is duepurely to a desire of the manufac-turer to build to a price, or perhapsto the fact that few radio listenersrealize the vast improvement whichthe pre -amplifier can effect at shortwaves. It suffices here to say thatall three types of interference citedabove (noise, cross -modulation andimage interference) are much worseat short waves; hence the pre -ampli-fier - not merely a pre -selector -should be present.

LA- 115 T. 142. 36 D.S.C.1. 126 T.142. 30 ENAMEL

Antenna coil data for circuit Fig. 15.

Since the selectivity and amplifica-tion of a T.R.F. amplifier falls off asthe wavelength becomes shorter (fre-quency becomes higher), a correctlydesigned multi -range receiver shoulduse more stages of T.R.F. amplifica-tion in the short-wave ranges than inthe broadcast range of 550-1500 kc.(example, General Electric model K-80). The practice of dropping outthe pre -amplifier when leaving thebroadcast band is merely a way tocheapen the receiver; with reducedresults as a consequence.

Cross -modulation in the T.R.F.Amplifier

It is necessary that selectivity bebuilt up stage -by -stage in the T.R.F.

A three gang condenser circuit.

11'10.0

5K 7.5

5.0

0g 2.5

"5-0

-SENSITIVITYCURVE A -FCURVE B"-FIG.45

-

800 800 1000 1200 1400FREQUENCY IN KILOCYCLES

FIG.16

Sensitivity curves of circuits Figs. 14 *nd 16

FIG. 18

T . 22 T. N2.30 ENAMEL WOUNDOVER L3

L3..110 T. NZ 30 ENAMELLi- (14 T. N2 30 ENAMEL

Oscillator coil data for circuit Fig. 15.


voosELY COOOLEO

WESTINGHOUSE

FIG.I9

oo 7

(WAVE TRAP CIRCUIT TO WIPEOUT IMAGE FREQUENCY)

DET

STROmBERG- CARLSON

nnEG.

KOLSTER

Various band selector circuits used in commercial superheterodynes.

amplifier at a rapid rate, otherwiseunwanted strong interfering signalswill not be sufficiently held down,and will become amplified to the pointof overloading either an R.F. ampli-fier tube, or the first detector (mixer).The use of the Ballantine-Snow "var-iable -mu" or "super -control" types oftubes (such as the 51, 35, 58, 78etc.) largely prevents this, but insome cases the volume control mustalso act to reduce input from the an-tenna, besides controlling the gainin the tubes.

Practical NotesPractical examples of 550-1500 kc.

pre -amplifiers are shown in Figs. 14and 15, the corresponding gain curvesappearing in Fig. 16. Observe thatthe addition of the extra tuned cir-cuit in Fig. 14 has had the usual con-sequence of an un-compensated tunedcircuit-a sloping gain curve, as a

penalty for the greater selectivity.Coil data for circuit 15 appear inFigs. 17 and 18, the coils being suit-able for a model 311-H tuning gangmanufactured by the Radio CondenserCo., to whom Fig. 16 is due. Thealuminum shield cans are indicatedand must vary but little from the di-mensions stated, otherwise the coilinductances will be altered and tedi-ous cutting and trying result. L2may be a small "universal" coil suchas used for R.F. chokes in broadcastreceivers, while C2 may Lave a capac-ity of not more than 500 mmf. in anycase, but is dependant to some extenton L2 and may be very much smaller.

The circuits of Fig. 19 have allbeen used by the makers listed andare quite sound, but less necessarywith variable -mu tubes, which ordin-arily permit either dropping out oneof the tuned circuits, or the inter-position of another R.F. stage.

CHAPTER 3

Oscillators and Frequency Changers

As we have explained, the purposeof the oscillator is to supply an ad-justable radio frequency voltagewhich at all times is different fromthe frequency of the received signalby just the amount which is repre-sented by the intermediate frequency-or rather the tuning of the I.F.amplifier. For some years, I.F. am-plifiers have been built to work at175 kc. and oscillators have conse-quently run 175 kc. above the fre-quency of the received signal-sinceit happens to be more convenientmechanically to do this than to runthe oscillator 175 kc. below the re-ceived signal frequency.

Thus we may set up examples ofthe operation of such a receiver:R.F. and mixerfrequency.

1500 kc.1250 kc.1000 kc.750 kc.550 kc.

Oscillator f -quency

1671425 kc.1175 kc.925 kc.725 kc.

Design of Oscillator and T.R.F.Circuits

From this we can at once find outsomething about the tuned circuits ofthe set. The frequency to which acoil -condenser combination tunes isdependant on the capacity, C, of thecondenser (plus the stray capacitiesin the set) and the inductance, L, ofthe coil. However, the resonance fre-quency does not vary directly withthe product of L and C; it varies withthe square root of LC. Thus a con-denser with a capacity of 250 mmf.connected across a coil with an in-ductance of 100 microhenries willtune to a frequency of 1000 kc.,which is to say, a wavelength of 300meters. If we quadruple the capaci-ty to .001 mf. (1000 mmf.) the tun -

17

ing will not shift to 250 kc. (1200meters)-but only to 500 kc. (600meters). Thus, increasing the capac-ity to 4 times the former value hasincreased the wavelength to only twicethe former value-and 2 is the squareroot of 4, illustrating the above state-ment.

In the receiver tabulated, we wishto tune from 500 to 1,500 kc. 1,500is 3 times 500, and since the fre-quency change is in proportion tothe square root of the tuning -capac-ity change, we evidently need a tun-ing condenser with a 9 -to -1 capacityrange. This 9 -to -1 range must bethe result when all the stray capaci-ties in the set are connected acrossthe system, and usually such a rangeis available with a tuning condenserwhose maximum capacity is about 350mmf., although some circuits requiremore, as mentioned in the discussionof Figs. 12, 14 and 15.

The case for the oscillator is dif-ferent. Referring to the table again,1675 kc. is only 21/8 times 675 kc.Therefore, the condenser (plus straycapacities) needs a range of capacityof only 21/z x 2% = 61/4 to 1. Or-dinarily we allow the minimum capac-ity of this condenser to be about thesame as that of the condensers usedto tune the T.R.F. amplifier-the min-imum being the capacity in the cir-cuit at 1500 kc. Thus, we evidentlyneed a condenser section for the os-cillator which has a smaller maxi-mum, which is produced by means ofa fixed (or adjustable) series con-denser; or (for bandspread sets) is"tapped down" a few turns on its coilto lessen the tuning range. Thequality of tracking for these systemsis in the order of their listing.

Also, since the minimum capacitiesare the same, it must be evident that

18 THE SUPERHETERODYNE

the oscillator coil, tuning to 1675 kc.,can have only about 80% of the in-ductance of the T.R.F. coils whichtune to 1,500 kc. with another con-denser of the same minimum capaci-ty.

This has been set down in some detailbecause the reader will wish to workwith other ranges, and with other val-ues of intermediate frequency-but al-ways with these same principles. How-ever, the reader is cautioned thattheory is not too dependable here,since it cannot anticipate stray cir-cuit capacities correctly and must besupplemented by painstaking cut -and -try.

Short-wave OscillatorsSpecial problems arise in connec-

tion with oscillators for short-waveranges, especially where it is desiredto use the same tuning condensers forthe 550-1,500 kc. range also. The un-happy consequence of this commoncombination is that one must use atuning capacity of about 350 mmf.max. (see foregoing paragraphs) forthe 550-1,500 kc. range, which is fartoo much for good amplification atshort waves. Changes in the tuningcapacity usually produce mechanicalinvolvements and increased circuitlosses, frequently leading to unreli-able functioning at short waves. One

BOOK

DET

osc

oer.

Fig. 21, left, illustrates magnetic coup-ling of detector and oscillator. Fig. 22,right, oscillator pick-up coil linked with a

coil in cathode lead of detector.

therefore frequently makes the bestof a bad matter and uses the 350mmf. (or thereabout) condenser inall ranges, and puts much thought andcare into the design and test of anoscillator which is "sure fire" atshort waves, without producing ex-cessive output at longer waves. Ifthe various ranges are obtained byswitching instead of plug-in coils, theutmost care is taken to minimizecapacities in the switch, not princi-pally because of the added micro-mi-crofarads, but because these addedmicro -mikes produce all manner ofundesired stray couplings and someadded loss. Where possible, the

Figure 23 (left) operates with grounded cathode and takes its bias from the 40,000ohm gridleak RI, oscillation amplitude and harmonic production being limited by the

6,000 ohm grid series resistor R2.Figure 24 (right) operates with higher initial bias due to the 6,500 ohm cathode re-sistor R4, hence its limiting resistor R3 is but 600 ohms. The values are for a 27,56 or similar tube. In the broadcast range of 500-1,500 kc., L2 will have about 1/5the number of turns of Ll, the latter being center -tapped for the grid connection.If the tuning condenser C2 has the same capacity range as the sections used to tunethe mixer input and other signal -frequency circuits, the remaining capacities will beapproximately; C3 adjustable 5 to 50 mmf. (ordinary trimmer found on condensergang), used here to trim minimum, i.e., to 1675 kc. when other circuits are set at1,500 kc.; C5 fixed, depends somewhat on C2, but about 750 mmf. (.00075 mf.); C4,

to correct C5, range about 15 to 75 mmf.; Cl, 500 to 750 mmf.


CL

I MEG.

24 OR 57

POTENTIOMETER(ANY CONVENIENT

SIZE) "..\

0-IMA., D.C.MILL IAM-METER

BATTERY(NOT A C )

-ZZ 4.5V.,

CL2 9v.

FIG. 25

+65-90vz

11111 417W;

A. V.T. voltmeter for adjusting the out-, put of the oscillator to optimum value.

switch blade is kept grounded andonly the contacts are raised "off-ground"-the reference of course be-ing to R.F. voltages off chassis; D.C.voltages being unimportant.

The very serious problem of fre-quency instability enters here. In the550-1,500 kc. range this is not toopainful, since a change of an entireper cent ( !) in frequency only putsone into the next broadcast channel.But, if we happen to be receiving at25 meters and the frequency wandersonly 1/4 of 1 per cent we go wander-ing off across a number of stationsin succession, whereas a 1 per centshift would cause us to shift theequivalent of 12 ordinary broadcastchannels-which is intolerable.

Such frequency wanderings are dueto several things-A-Aging effects.

Changes in the coils with time andweather.

Changes in the tubes with age.Gradual changes in trimmer con-

densers.B-Heating effects.

Tube warmup.Warmup of tuning elements after

set is turned on.C-Irregular effects.

Changes in tubes (replacements).Antenna changes.Coil -shields shifting with respect to

IA coils.Line -voltage changes.Variable contacts between metal

parts.Against all these the short-wave

oscillator must provide if it is a good -enough job-and many are not. A

large part of the unsatisfactory per-formance and alleged "fading" ofshort-wave signals is due to suchthings.

Experience has shown that effectsof type A can be minimized by theuse of high-grade construction, whileeffects of type B and C are anticipat-ed by the use of proper oscillatorycircuits and by disposition of partslearned through care and experience.Here careful laboratory work is es-sential.

Oscillator CircuitsAlthough design conditions intro-

duce various disguises, the oscillatornormally uses some form of the Hart-ley circuit; or its close relative, thetuned grid coil with an untuned platetickler. In Fig. 21 and in Fig. 22 areshown schemes for magnetically coup-ling such oscillators, to a screen -griddetector, via the grid (input) circuit.In Fig. 21, this is done by magneticcoupling to the tuned coil, while inFig. 22 it is done by magnetic coup-ling to a "pickup" coil, connected inseries with the cathode-which is,after all, a part of the input circuitand hence equivalent except as tomechanical convenience. In boththese diagrams the lower end of thedetector grid input coil is, of course,understood to be grounded to themetallic chassis of the set.

An increasing tendency towardfixed screen -grid voltages has large-ly done away with the formerly com-mon scheme of feeding the oscillatorR.F. voltage into the detector screen -grid. One must not jump to conclu-sions on observing that an oscillatordraws its plate supply from somepoint apparently connected to R.F.and detector screen -grid. This iscommon practice because the oscilla-tor commonly runs with a plate -supplyvoltage of about 90 and this is thescreen -grid voltage of most R.F. anddetector tubes. However, the D.C.takeoff to the oscillator is usuallymade near a bypass condenser whichminimizes any possibility of R.F.coupling to the screen -grids. Fre-quently the filtering at this point isincreased by taking off the oscillatorsupply through a 100 to 5,000 ohmresistor, as a decoupler.


__

-

"B+" 8+" "B-4--90V. -B-- Sov. -A-- - 90V. -C-- .

-_,

ORDINARYOSCILLATOR.

ELECTRON -COUPLEDCIRCUIT

T PLATE GROUNDEDINSTEAD OFCATHODEFl Cr. 26

In the electron -coupled oscillator, the oscillatory circuit itself has not changed, thescreen -grid merely taking the place of the plate used in A and B. This leaves theplate free to act as explained in the text.

Still another coupling scheme is toconnect the oscillator control -grid tothe detector control -grid through aresistor, usually with a stopping con-denser in series, as the biases arenormally different. A minor varia-tion of this is to connect some ele-ment of the oscillator to the mixerinput grid through a small capacity.This is especially useful in multi -rangereceivers since the coupling capacityhas a reactance which decreases withwavelength, hence partly compensat-ing for the naturally weaker oscilla-tions at shorter waves-which darkstatement will be explained within afew paragraphs. An example of thispractice appears in the National "FB-7" receiver.

Regardless of these variations theoscillator circuits are all essentially asshown in Figs. 23 and 24 with theconstants stated.

The data given under Figs. 23 and24 hold, regardless of the method ofcoupling the oscillator to the mixer.At short waves it will be found thatthe tickler coil L2 will need compar-atively more turns-sometimes almostas many as Li, and that the grid tapwill need to be carried further up Ll.

Proper Oscillator ConditionsWhile the "variable mu" Snow-Bal-

lantine tubes such as the 35, 78, etc.,are not good second detectors, theymake excellent mixers (first detec-tors) provided the R.F. voltage fedto the control -grid from the oscillatoris about 1 V. less than the D.C. bias

on that grid. Since there are so manytypes-and new ones appear on Tues-days and Fridays-it would be fool-ish to give data here. Use the condi-tions set down in the tube -data -bookof the manufacturer, who will cheer-fully send details on any tube uponrequest. Then adjust the oscillatorto put the proper R.F. voltage on thegrid. This adjustment is simply amatter of changing the coupling be-tween the oscillator and the mixer,while reading the R.F. voltage at themixer -grid with a simple vacuum -tubevoltmeter of the "peak type." Thecircuit of a device of this type isshown in Fig. 25, the only part ofwhich needs explaining is Cl, whichis a "2 -plate" midget -variable con-denser, set at maximum.

The procedure will be illustratedby example. Suppose we intend touse a 35 tube as a mixer. From anRCA data sheet we find that thistube as a mixer should operate with250 plate V., 90 screen -grid V., and abias of minus 7 V. We have builtit up in a receiver and now wish toadjust the oscillator so that it sup-plies to the 35 control -grid 6 peakR.F. volts (1 less than the 35 bias).Using the setup of Fig. 25, clip CL1to the 35 control -grid, CL2 to thechassis, remove the oscillator tubeand adjust slider S to obtain a platecurrent of 1/10-ma., then read thevoltmeter V. Now start the oscilla-tor and move S until the reading ofthe milliammeter is again 1/10-ma.


Read V. once more and subtract thetwo readings. The difference shouldbe the peak R.F. voltage applied tothe 35 grid, unless stray pick up hasspoiled your readings. To check thisturn Cl to minimum. If the readingsare repeated you should now get asmaller result-which is incorrect ex-cept as a check. If Cl has little orno effect, shield the whole rig by en-closing it in a metal can with onlyCL1 coming out through a thin micawindow 1 in. across, the rest of thegrid circuit of the meter -tube beinginside. It may be necessary to de-crease R to .1-megohm to obtainsensible readings. The meter thendraws too much power from the os-cillator, but is still vastly better thanguesswork.

All this must be done with care (itis really no novice's job) but an ob-serving user w.11 improve his receiverwith it just the same.

Electron -Coupled Oscillators

The difficulty of getting good fre-quency stability from normal oscilla-tors working at short waves, andcoupled in the ways mentioned, hasled to the introduction of the Dow"electron -coupled" oscillator and itsdescendant, the "pentagrid conver-ter." One form of the electron -coupled oscillator appears in Fig. 26.

Since the plate of the circuit shownin Fig. 26B is bypassed to chassis, itis unable to develop any R.F. volt-age differing from that of the chassis.It follows that the plate of Fig. 26Cis screened from capacity coupling tothe oscillator, even though the oscil-lator is in the same tube. From thisit follows, in turn, that variations inthe plate voltage or load cannot havemuch effect upon the oscillator fre-quency-which the effect we desire.

At the same time, R.F. power doesreach the plate of Fig. 26C, by thefollowing process. Since oscillationis taking place in the grid -cathode -screen combination, the control -gridmust evidently be swinging up anddown at radio frequency. It conse-quently is varying the electron streamwhich flows from the cathode throughthe grid to the plate-hence thestream arrives at the plate bringing

R.F. power with it. The plate ac-cordingly is associated with the restof the tube -action only through theelectron stream - hence "electron -coupled."

The Pentagrid Converter

The inconvenience of having thetube's cathode "off ground" as in Fig.26, immediately leads to the thoughtthat we ought to have a tube with anextra grid to be used for oscillationso that we would not need to put thetickler between screen -grid and ca-thode. This would be a tube witha cathode, two oscillating grids, ascreen -grid and a plate, in otherwords a spec.al sort of pentode. Suchtubes have been made.

The next step is to put one moregrid in, using it to feed the incomingsignal from the distant station intothe tube-whereupon we have com-bined the "mixer" with the electron -coupled oscillator. This is the "pen-tagrid converter," which has meritsaltogether aside from that of merelycombining two tubes. A well-knownsort is the 2A7 also 6A7).

Not only does it give us the goodstability of the electron -coupled os-cillator; it also provides an oscillator -mixer coupling that is automaticallyright. The circuit of such a tube isshown in Fig. 27, while Fig. 28 showscoils which will work over the 500-1,500 kc. broadcast band with a 175kc. intermediate amplifier.

7,1

300OHMS

50.000OHMS 0

(TUNE)

-250MMF.

c)

CONTROL IS NOT .,--.T

0.511-14

o (CONTROL BIAS ClUP TO 45V. IF

WANTED, OMIT C ) 13+J B+ B_:F G. 27 100V. 250V

I.F.T.1

Cti3.I.

ORLARGER

C4

C 1540X

L2/ C2

The method of connecting the pentagridconverter as a frequency changer.


Short-wave Oscillator Coils

The table, Fig. 29, at the end ofthis chapter gives coil dimensionssuited to various short-wave bands,but using a triode oscillator (type56 or the like, with a pickup coil con-nected into the cathode circuit of anormal screen -grid detector (type 57or the like). To adapt these oscillatorcoils to the 2A7 or similar tube, sim-ply omit the pickup coil, since anelectronic pick-up is built into thetube. The plate coil is also omittedand in its place is wound a coil cor-responding to L2 of Fig. 27. This iswound directly over the grid coilwith only one layer of oiled silk be-tween, and spaced to have about 3ithe length of the grid coil-the gridcoil being, of course, Ll (Fig. 27)which is connected to the first grid.The number of turns for the fourcoil -ranges is ,correct-but rememberto space out the turns as described.

This set of short-wave coils is in-tended to work with tuning conden-sers of 150 mmf. maximum capacity,the oscillator section being like theothers but having a .001 mf. "pad-

GR DCOIL

(Li -MG 27)92 Toms, N4.32

ENAM. WIRE -CLOSE WOUND

FIG 28

ONELAYER.

OF OILEDSILK

PLATECO L

\". (L2 -FIG. 27)36 TURNS, N4.31 ENAM.

WIRE CLOSE WOUND DIRECTLYOVER -LOW -ENO OF LL

An oscillator coil for the 1500-550 kc.band with a 175 kc. I.F. amplifier.

f..01-

OGT CAL

'Att.410....4Vo%

. DAC

An, SOSA non.

AAA 74770I. ,7

Turawleteil AAA>

40 SOA NAASRAC AAA

DATA -

AAA OS NAM

DOWVA joal'AV aoSTM7A40

4=14 A' 'Irowl

....?frar AAAA

FIG.29P. DIA

Coil data for oscillator and detector coilsfor a S.W. superhet.

ding" condenser in series with it,corresponding to C4 and C5 of Figs.24 and 25. This can be an adjust-able condenser, of that capacity, orelse a fixed one, shunted by an adjust-able one, as in Figs. 24 and 25.

The "detector coil" winding is de-scribed for a receiver without an R.F.amplifier, but one can be-and shouldbe - added, using a variable -muscreen -grid tube of the sort repre-sented by types 35, 58 and the like.The "det. coil" of the table thengoes between this R.F. tube and an-tenna, while between the R..F. tubeand the detector (corresponding toL4 of Fig. 28) there will be a similarcoil-with this difference only-the"ant. coil" is omitted. In its placethere is wrapped one turn of oiledsilk over the grid coil, on which iswound a coil exactly like the gridcoil (spaced the same as the gridcoil) but only 1,fi as long and accord-ingly with only % as many turns.This is connected into the plate cir-cuit of the R.F. tube. (Incidentally,the coil at Fig. 17 can easily be madeinto a good R.F. transformer [500-1500 kc.] by such means.) The cir-cuit of this receiver is shown in alater chapter as the 'Simplified Mit-cherl Receiver."

CHAPTER 4

Single Dial Tuning Systems

AS we have mentioned in earlierchapters, the problem of tuning a

superheterodyne receiver with onecontrol -knob, differs from the sameproblem in a T.R.F. receiver, for thereason that the oscillator requires atuning range which differs from thatof the other tunable circuits in thesuperheterodyne receiver.

In the example given early inChapter 3, we pointed out that whilethe T.R.F. stages must tune from 550to 1,500 kc. to cover that range ofreceived signals, the oscillator must,at the same time, tune from 675 to1,675 kc., provided that the interme-diate amplifier operates at 175 kc.If the I.F. amplifier works at a dif-ferent frequency-and it often does-then the oscillator must have anotherrange, always such that its frequencyis at all times above the T.R.F. fre-quency (frequency of received signal)by an amount just equal to the I.F.(This statement neglects those infre-quent sets in which the oscillatorworks "below" or "across" the re-ceived signals.)

Off hand, one might judge that itwould be necessary only to reducethe inductance of the oscillator coilby removing turns, using a variablecondenser exactly like that used forthe other stages. This is impossible,since the resulting curve will not stayat a uniform distance from the T.R.F.curve, but will, instead, maintain afixed percentage relation-which isuseless.

An early method of accomplishingthe proper "tracking" was proposedby McLaughlin in 1924. He used anoscillator tuning condenser like theT.R.F. condensers but used a variableoscillator inductance in the form of avariometer, driven by the condensershaft. Figure 30 gives the details,

28

because even today, they are of in-terest in clarifying the problem.

The mechanical inconvenience ofthe McLaughlin device caused it to belargely abandoned in favor of an-other scheme, in which the coils areall fixed, the oscillator inductance be-ing somewhat smaller than the others.The oscillator tuning condenser ismade to have a slightly lower mini-mum capacity and a considerablylower maximum capacity by connect-ing a fixed condenser in series withit. Assuming the constants to havebeen chosen correctly by a somewhattedious process of calculation, pluscut -and -try, we have the situation ofFig. 31, where A is the T.R.F. tuningcurve, B the one we hope to haveattained in the oscillator. The Bcurve is seen to lie 175 kc. above theA curve. It does not look so, butthis is optical illusion, exposed by afew measurements on the graph. Ac-tually, the results are seldom so per-fect, one attains more nearly such acurve as C, which is not evenlyspaced from A. One may then eitherlocate it as shown, with both ends ly-ing on the correct curve B, or else(and more usually) one chooses theLeast average error, by making bothends lie above B and the middle be-low B. This has given acceptable re-sults in hundreds of thousand of re-ceivers.

Since it is nearly impossible tohave either the variable or the fixedoscillator condensers just right as tocapacity one must in practice use thecomplication of Fig. 32, where C isthe variable tuning condenser with itsusual "trimmer" or "screwdrivermakeup condenser," marked C3,while the series fixed condenser C1is likewise provided with a "trimmer,"C2. Actual values for a 175 kc. case


-COIL DATA- R.F.C.L -59 T. NO. IS D.C.0 ,3./8" DIA.L2 S T. f46' FROM LIL3,- S T. " .0/4 TUBE INSIDE L4L4-54,4, T... ,3418" DIA.L5-30 T. s' " " " 1/4" FROM L4C1,C2...0005-MF. EACH

FIG.30 8+

An early single dial circuit using specialcoils and identical condensers.

were given in Fig. 23 for a 175 kc.I.F. system, others will appear later.

Incidentally, Fig. 31 illustrates an-other, and more refined, method whichis gradually becoming standard, someyears after it should have beenadopted. This is to start at the be-ginning-by tun-ing condenser of the proper size andshape for its job so that no extrasare required and only the usual small"trimmer" is attached to it. Theconstants for that arrangement ap-pear on the diagram of Fig. 31, andthe necessary plate -shape is discussedlater.

For the present, we wish to say aword more about the scheme of Fig.32, which can be used by the amateur,both in that form and in a moresimplified form.

If 175 kc. is the intermediate fre-quency, and 550-1,500 kc. is the tun-ing range, then the fixed series con-denser must have about 2 times thecapacity of the tuning condenser atmaximum setting, although this rulemust be used merely as a first ap-proximation, subject to later adjust-ment. Fig. 23 and 24 are illuminat-ing.

The oscillator -detector system willthen look like Fig. 34, although theinductive coupling between the twomay be replaced by any of the otherfeed methods suggested in earlierchapters. Oscillators, by the way,

should be shielded as a concession tothe comfort of the neighbors who mayotherwise be subjected to squeals.Only the coil is especially in need ofsuch shielding, and the radiation isvery short -ranged, unless one dealswith a primitive superheterodynewhich lacks T.R.F. stages to preventradiation from the antenna.

Special Condenser SectionsThe rather idealized tuning curves

of Fig. 33 have been saved up to thispoint, as they are not of the greatestpractical importance. It has latelybeen recognized that such rigidly"straight line" tuning is of question-able value, even if it were readilyobtainable. It is, moreover, not read-ily attained, partly because of theclumsy condenser plate -shape whichis indicated, and partly because thebest of the designs is to some extentdefeated by the uncontrolled varia-tion in the set.

Special plates do, however, serve avery real purpose in the oscillatorsection of a gang tuning condenser.

Design of Oscillator CondenserPlate Shape

The method of calculating the plateshape of an oscillator condenser so thatit will accurately track with the con-densers used in any R.F. tuning sys-tem and maintain a constant frequencydifference of 175 kc. or any other valuehigher than the R.F. condensers atany setting in the range has beenworked out and is used in many com-mercial receivers.

The first step in the procedure is todetermine the tuning characteristics ofthe signal frequency amplifier as shownin Fig. 31 and with the intermediatefrequency known (175 kc. for example),the theoretical oscillator curve can bedrawn, as is shown by curve B. Toavoid errors from lump minimum cap-acities which may exist in the circuit,the condenser is designed so that asmall capacity will have to be addedby means of a trimmer condenser sothat the sum of the minimum distrib-uted capacities can be controlled andset at a fixed starting point.

Next an inductance value for the os-cillator circuit is selected which will

THE SUPERHETERODYNE

allow the oscillator circuit spectrum tobe covered with a condenser of practicalsize. Inductances of 150 to 200 micro -henries are found in commercial re-ceivers. In designing an inductance wemust not lose sight of the fact thatabout 20% must be added to compen-sate for the loss of inductance whenthe coil is placed inside a shield.

Knowing the inductance we can cal-culate the necessary capacity values forthe oscillator condenser at various dialsettings. When these capacity valueshave been. ascertained and the capacityvalues of the R.F. tuning condensersknown at the same angle of rotation,the radii required to give the desiredcapacity curve to the oscillator conden-ser plate can be calculated by means ofthe following formula:

Cor

Where,

ro_n

Cs

nCs=capacity of R.F. condenser

at any given dial setting.Co=capacity of oscillator con-

denser at correspondingsetting.

r =radius ofplate.

ro =radius ofdenser plate.

n =pairs of activeR.F. condenser

1600

1500

1400

1300

1300

1100

1000

900

800

700

600

500

6

R.F. condenser

oscillator con -

surfacesunit.

CURVE "A.R.F. TUNER CURVE FOR

175 KC. SUPERHETERODYNE.ANT. AND COUPLING SECOND -

241.23 MN.. TUNINGCAPACITY- 366 MMF. MAX.

CURVE: B"OSCILLATOR TUNING CURVE

OSC. COIL SEC.K143. 63 MN.050. TUN. COND.= 355 MMF.FIXED PADDING CAPACITY

- 14 MMF.

try

0 20 40 60 80 100DIAL DIVISIONS

FIG.31

Typical tuning and oscillator curves.

in

BOOK 25

A padding circuit for single dial control.

This formula may be used for solvingany required curve and is not limitedto this particular problem.

Curve A of Fig. 31 is a typical curveof a tuning condenser manufactured byThe Radio Condenser Company.

Fig. 35 shows a capacity curve, A,of the oscillator condenser made bythis same company. This curve wasplotted with the oscillator trimmer con-denser at the minimum position. Thisis not the capacity curve this unit hasin actual use, because when used theminimum is raised to 50 mmf at 12divisions by the lump stray capacity ofthe receiver and by means of the trim-mer. This graph should be comparedwith that of curve B, showing thecapacity curve of the R.F. tuning con-denser as measured apart from the cir-cuit.

The oscillator condenser is designedto be used with an oscillator coil sec-ondary having an inductance of 143.6micro -henries. When the minimumcapacity of the circuit is set to a valueof 50 mmf. by means of the trimmer,the proper frequency distribution ofthe oscillator spectrum is that shownin Fig. 31, curve B.

In designing a receiver using thesecondensers, a check should be made tomake certain that the R.F. amplifierclosely follows the graph of Fig. 31,curve A and that the oscillator circuitfollows graph B of Fig. 31. Both ofthese measurements should be madewith the intermediate amplifier accur-ately adjusted to peak at 175 kc. Afinal check may be made by connectinga small 10 mmf. carefully calibratedcondenser in parallel with the oscillatorcondenser, and supplying a modulatedsignal of known frequency by meansof a signal generator in the usual man-ner.


Turn the dial to the point where1,500 kc. is a resonant frequency of theR.F. amplifier (equivalent to 12 divi-sions on Fig. 31A) and reduce the trim-mer of the oscillator condenser untilthe capacity added by the small vari-able test condenser has been allowedfor, and gain is maximum. The platesof the test condenser should be set atabout 90 degrees when making this ad-justment so that capacity can be eitheradded or subtracted.

Readings should then be taken atvarious points throughout the 1,500 to550 kc. range and the gain brought to amaximum at each reading by means ofthe small test condenser. The varia-tion from the initial setting should benoted and correction made to the os-cillator condenser by bending the slot-ted plates. The variations, in properdesigns, will never be greater than canreadily be corrected for by bending theplates. To avoid body capacity the ad-justment of the test condenser shouldbe made with a long insulated handle.

The Precise gang condenser, Fig. 36,for use in a superheterodyne having anintermediate frequency of 175 kilocyclesis designed to work under approximate-ly similar conditions.

Multiple -function tubes

The use of a multiple -function tubeto combine the oscillator, the mixerand electron coupling in one envelope,does not ordinarily change the picture

Fig. 33. The ideal (not actual) tuningcurve of a superheterodyne - absolutestraight -line -frequency variation with theoscillator 175 kilocycles above the detector

tuner.

very materially, except insofar as thetube itself may require somewhat in-creased feedback coupling to give thesame oscillator performance as waspreviously obtained from triode oscil-lators. The changes in the oscillatorwindings which then become neces-sary are suggested in Fig. 27 and theassociated text.

Short -Wave ReceiversSo far, we have blandly talked

about 175 kc. I.F. systems, and 550to 1,500 kc. tuning, as if there werenothing on earth but the standardbroadcast band. In the process, wehave mainly learned that for the am-ateur it is well to try no tricks, butto use an oscillator condenser likethe other condensers, put twice itsown capacity in series with it, re-duce the oscillator inductance 22 percent from that of the T.R.F. coils(compare L3 with Ll in Fig. 18),and then finish the alignment byusing a trimmer across the oscillatorvariable condenser to get the high -frequency end straight (best responseon stations near 1,500 kc.) and mak-ing the low -frequency alignment withthe trimmer across the fixed series or"padding" condenser. The proprietyof this is seen by again referring toFigs. 32 and 34, from which it is ap-parent that the minimum capacity ofthe tuning condenser must be ad-justed by the trimmer across it whilethe maximum must be adjusted by the

INDUCT-IVELY

COUP-LED

151. DET.L C '27 I FT

EKC.175

, 550 -1500 Kc.-- 725 - 1675 KC. 05C.

Fig. 34. The tuning circuit. Constants forsingle -dial tuning are given in the text.

THE SUPERHETERODYNE

trimmer connected to the series con-denser. (It is unfortunate that thelabels in Figs. 33 and 34 are notidentical. )

Going now to a short-wave re-ceiver, we seem for the instant to beat sea, for we not only have a dif-ferent tuning range, but also a dif-ferent intermediate frequency. In acombination broadcast- and short-wave receiver, the intermediate fre-quency may be found (in 1934) tobe near 450 kc., as high as it can bemade without getting overdose to thebroadcast band, and thus spoilingbroadcast -band reception through in-terference. In pure short-wave re-ceivers, we find the I.F. to be abovethe broadcast band, usually at about1,575 kc. This seemingly senselessdifference may pass away shortly, andhas arisen mainly because the "pure-ly short-wave supers" seem, as a rule,to be deficient in the number of tunedcircuits (or to lack T.R.F. and pre-selection altogether), hence are ab-normally subject to bad image inter-ference, (See chapter 2), which canbe improved materially by using ahigh intermediate frequency, thusthrowing the image farther away. Thecorrect cure, however, lies in amplepre -selection, as illustrated by thediagrams of Chapter 8.

As the 1,575 kc. intermediate sys-tem gives relatively poor amplifica-tion, another fairly common dodge isto use an intermediate frequency inthe broadcast band, merely adjustingit to lie halfway between two broad-cast channels, so chosen as to avoidstrong stations. This requires an ad-justable, or tunable I.F. amplifier,hence one may use a broadcast re-ceiver for that purpose. The entiresystem then becomes a broadcast re-ceiver preceded by a short-wave oscil-lator -detector combination. Usuallythe latter device is separate and issold under the name of a "converter."True short-wave receivers have ren-dered them obsolete and none areshown here. For those still inter-ested, the following comments areoffered:

In choosing a converter to attach toan existing broadcast receiver, onecan roughly gauge the probable re -

BOOK 27

300

5z- zoo

4oO

I I I

CURVE."A"05C. TUNING COND.WITH SPECIAL SHAPEDROTOR PLATES

CURVE -13.R.F. TUNING COND

0 10 20 30 40 50 60 70 80 90 100DIAL D VIS ONS

FIG.35

Typical capacity curve of oscillator con-denser.

sults from the following rules.Choosing a Converter

A two -tube converter will as arule show rather serious noise level,unless the detector is either a triodeor a screen -grid tetrode used as a tri-ode (screen tied to plate). The 35and similar tubes seem better thanthe 24 and the like, even when usedas triodes.

Converters in which the first tubehas a resistance input seem ratheruniformly to be quite noisy, just asdo receivers with that antique typeof input. Frequently results can bemuch improved by discarding such analleged R.F. amplifier stage and con-necting the antenna through a smallcondenser (100 mmf. or less) to theplate prong of this tube (which isbest left in place though not used).An actual R.F. stage, instead of a"noise amplifier" stage, is of course,an asset. Such a stage has a tunedcircuit feeding it, or at least a goodR.F. choke-never a resistance.

Need it be said that a converterwith any dial but a very smooth andeasy one is much worse than no con-verter at all?

Short-wave SuperheterodynesGoing to the actual short-wave or

all -wave superheterodynes, it becomesmore and more the case that one can-not specify much and must largelycut and try for dimensions. Each re-ceiver description is accordinglybound to differ in details of coils andcondensers from others meant for thesame wavelength range - especiallywhere there is an attempt to get


maximum amplification by keepingthe capacities of the T.R.F. and oscil-lator down. The effect then is to al-low the "stray" circuit capacities tobecome relatively more important, andas they differ with constructional de-tails, the statement above is sub-stantiated.

Thus for example, we have in theGeneral Electric K-80 series of all -wave receivers, a case where straycapacities have been swamped out bydeliberately using ample tuning andtrimming capacities, the decreasedamplification being made up by addedR.F. stages-which incidentally helpsthe noise and image situation. Herewe have, for example, a tuning rangeof 3,900 to 10,000 kc., which, withthe 445 kc. I.F. used, calls for anoscillator range of 4,345 to 10,445kc. The tuning condenser is a nor-mal 335 mmf., one of broadcast -range type, which accounts for thetuning range of 3,900 to 10,000.Our previous rule of using twice asbig a series condenser would be allwrong as the oscillator range is nowalmost as big as the T.R.F. In factthe series condenser has to be .000225mf., about seven times the tuningcapacity. For the range of 8,000 to18,000 kc. it becomes .00234 mf.!!

Tuning range,kc.

150-450(200 to 750 meters)

550-1,500 (broadcast)1,500-4,500

(66 to 200 meters)

This is the general tendency in theshort-wave region, but trial beats cal-culation.

Typical constants for short-waveranges using an I.F. near 450 kc. anda tuning -range of about 3 to 1; thatis, for any one range the highest fre-quency (condenser set at minimum)is about 2 times the lowest. (Exam-ple 550-1500 kc.) Both pad andtuner have trimmers, of course.

Another Method of AttackIn other designs it has been pre-

ferred to minimize tuning capacityas previously explained. Thus, in theNat.onal "AGS" receiver a tuningrange of about 4,100 to 7,200 kc. isobtained with a tuning capacity near100 mmf. through somewhat specialconstruction which reduces "stray"capacity, so that the 100 mmf. con-denser produces about 3/5 of the tun-ing range which was obtained in thereceivers mentioned above with Ca-pacities near 350 mmf. For a range -switching set, this would be objection-able, because of switch complexity-for a plug-in set, it is desirable asthe per -stage amplification in theR.F. system is materially increased,which means fewer stages - hencefewer plug-in coils per set. One may

Tuning capacitymax.

350 mmf.350 mmf.

350 mmf.

Series pad.capacity

100 to 150 mmf.500 to 800 mmf.

.0010 to .0012 mf.

Fig. 36. The Precise three- and four -gang condensers. The arrows indicate fhe oscillatorunits with special shaped plates.


answer that more sets of coils areneeded - but few want the entirerange anyway. In this case, the ratherlimited range per coil is not greatenough to modify greatly the trim-ming and padding rules.

"Bandspread"

Some years ago a fad was begunfor receivers which enormouslyspread out some small tuning range,usually an amateur transmitting band.This has led to rather grotesque tun-ers, usually of plug-in variety, whichare useless except in their narrowbands. Such "bandspreading" can beaccomplished in two simple ways.

The first is to build a normal short-wave superheterodyne, and to connecta small "vernier" condenser of about5 per cent of the tuning capacityacross the oscillator tuning conden-

ser, this small condenser to be pro-vided with a microdrive dial of somesort. This does not cripple the per-formance of the set for other work.

The second, and commoner methodis to use either very tiny tuning con-densers or (what is equivalent) toconnect a "padding" condenser acrossthe coil, and to connect the tuningcondenser across but a few turns, ata tap provided for that purpose. Withplug-in coils, it is possible to providean extra prong so that either a reg-ular coil or one of these so-called"bandspread" coils may be pluggedin. This is done in the National"AGS" and "FB7" receivers for boththe R.F. and oscillator coils. Thelayout of such coils is a matter of al-most pure cut -and -try. Printed de-scriptions, unless followed with pain-ful exactness, do not suffice.

CHAPTER 5

The Intermediate AmplifierFROM this point on, the reader will

find it to advantage if he will striveto keep in mind the fact that wehave now reached that part of thesuperheterodyne which is nothing inthe world but a long -wave T.R.F.receiver-complete. The so-called "in-termediate amplifier" differs in noway from a normal T.R.F. amplifier;the "second detector" correspondsexactly to the detector of any T.R.F.receiver, and the audio system isnormal in all ways.

Let us therefore remember that oncewe get beyond the "mixer" we are talk-ing about nothing but a long -waveT.R.F. receiver. In fact, the simplestpossible sort of T.R.F. receiver, for itiu without even a tuning adjustment.

Choice of Intermediate FrequencyThis long -wave amplifier is set

once for all on some frequency chosen to reduce image interference (seeChapter 2), to avoid strong localinterference (including the standard550-1,500 kc. broadcast band and the500 kc. ships), and yet low enoughto permit good amplification withoutgetting into an audio frequency andletting through noise.

In the past, this has simmereddown to 175 kc. which is outside theaudio range, yet low enough so thatthe second and third harmonics of350 and 525 kc. miss the standardbroadcast band. The selectivity andimage ratio are acceptable for the550-1,500 kc. range, though aboutthe same thing can be said for suchfrequencies as 135 kc. and 145 kc.

For short -waves, as already pointedout, the image ratio becomes bad,and one uses an I.F. around 445 to465 kc., missing the broadcast band-or else gets further image im-provement by giving up some ampli-

30

fication and jumping clear over thestandard broadcast band to 1,575 kc.or some such value.Number of tubes and tuned circuitsSince one of the main objects of

using a superheterodyne circuit is toobtain high amplification and selec-tivity in the intermediate amplifier-where there are no variable conden-sers, one finds that as a rule thisamplifier employs high -gain tubes andtransformers, which not only havetheir secondaries tuned (as in thecase of a normal T.R.F. amplifier) butalso have some of the primaries tuned.The high -gain tubes are commonlyscreen -grid tetrodes or screen -gridR.F. pentodes, meaning tubes suchas the 35, or else tubes such as the78. The R.F. pentode gives abouttwice the amplification while draw-ing twice the plate current and some-times a stage can be saved by usingit.

In Fig. 37 is shown an I.F. ampli-fier employing variable -mu screen -gridtetrodes in two I.F. stages, followedby a triode detector of the high-levelbias type-that is, a detector operat-ing at a high signal level and prob-ably followed by only one audiostage. Observe that the system showssix tuned circuits. Even if regenera-tion is carefully worked out of thesystem, this will produce excessivesharpness and destroy tone fidelity,unless the selectivity curve is care-fully adjusted. This is done by sospacing the primary and secondaryof each transformer as to provideapproximately "critical selectivity"-i.e., such coupling is used as will justbegin to broaden the curve. Over -close coupling will result in a two -humped tuning curve, hence onesometimes uses a de -coupling washer


A two stage I.F. amplifier circuit. Note that three tuned I.F. transformers are employed.

as in Fig. 39 and 42. If the curve isstill too sharp at critical coupling, itmay be slightly broadened by "stag-gering" the tuning deliberately. Todo this by ear is exceedingly diffi-cult, and in fact an oscilloscope(preferably of the cathode ray sort)is necessary for a proper job. Bylistening carefully a very observingperson with ample time can do a fairjob of such an adjustment by ear.Normally fewer tuned circuits are em-ployed in the I.F. and this problem isless severe.

The "Stenode Radiostat"While on this subject of I.F. selec-

tivity, one should mention the Robin-son "Stenode Radiostat," which is asuperheterodyne whose I.F. amplifierhas been enormously sharpened byusing in it a "stenode bridge" whichis simply an I.F. transformer em-ploying one ordinary condenser -tunedcircuit, plus a second tuned circuitwhich consists of a quartz -crystalresonator ground to the intermediatefrequency. Without going to crystaltheory, one can say that such a crys-tal plate is equivalent to a tremend-ously sharp tuned circuit, so that the"sideband cutting" becomes very ex-treme, the higher audio frequencies

almost entirely disappearing and noth-ing but rumbles, booms and gruntscoming out, if a normal audio ampli-fier is used. Accordingly the audiosystem must restore the audio fidelityby first flattening the low notes downto the level of the attenuated highnotes, then taking the resulting weak(but fairly undistorted) audio outputand amplifying it up to normal level.Therefore a typical circuit (Fig. 38)shows not only the equalizing trans-former (marked "filter" in Fig. 38)but also shows an unusual number ofaudio stages, 3 in this case. The netgain of the system, when capacitycoupling is erased by proper settingof the crystal -bridge condensers, is tosuppress greatly the ordinary back-ground noise of reception.

The "Single Signal" ReceiverIt must be apparent that if the

Robinson equalizing audio systemwere replaced in a stenode receiverby a normal audio system we shouldfind only low pitches in the output.If, in addition, the crystal bridge wasunbalanced by deliberate setting ofNC2 we should find a lopsoided trans-mission curve to result, virtually wip-ing out one sideband of the signal,an effect first used by Robinson, and

F g. 38. The "Stenode" rece'ver above shown differs from ordinary superheterodynes inthe circled filter which accepts frequency variations not exceeding a hundred cycles.


Fig. 39. The construction of the "Tran-sitor" I.F. transformer, available in three

designs.

explained by Batcher on the basis ofWheatstone -bridge theory - the so-called crystal bridge being nothingbut a wheatstone bridge in whichcondensers CIA and C1B are bal-anced against the capacity of NC2and the crystal capacity . . . the sys-tem going off balance at crystal res-onance-where the crystal is not acondenser but a tuned circuit, equiv-alent to both capacity and inductance.

Such a highly selective transmis-sion curve can be used for continu-ous -wave telegraphic reception, to ad-vantage, by exaggerating the IS. sel-ectivity to the point where a bandonly about 50 cycles wide is passed.When a second oscillator (beat -noteoscillator) is fixed at a point about1000 cycles from the I.F., we findthat ordinarily only one signal at atime can be heard, and that this sig-nal appears only once as the tuning -knob is turned, the lopsided I.F.curve preventing the image from be-ing effective to any great degree.This does not exactly describe the"Single Signal" receiver of Lamb, butindicates the general nature of itsoperation.

Crystal Filters for BroadcastReception

Despite optimistic claims, your re-visor has not yet heard a crystal -filtersuperheterodyne, which, on voice pro-duced anything but exceedingly badand low -intelligibility speech, exceptwhen equipped with a Robinson audioequalizer. The usual "single signal"receivers when used on voice areeither used in a desperate attempt toget some fragments of speech out ofan intolerable mess of interference,or else are used with the crystal fil-tering deliberately broadened in somemanner, either by switching the crys-tal out (leaving a plain I.F. trans-former) ; by unbalancing the crystalbridge to decrease the sharpness; orby switching the crystal to producea narrow band of suppression (in-stead of narrow band of transmission)and placing it on the unwanted in-terfering station. In Fig. 43 isshown a commercial crystal filterworking at 500 kc. with these adjust.ments provided.

Fig. 40. The Hammarlund I.F. transformer.The shield is removed so as to show the

inner construction.


I.F. coils used in the circuit of Fig. 87.

Ordinary I.F. TransformersIn Fig. 39 is shown a commercial

175 kc. transformer with a copperdecoupling washer to produce approx-imately critical coupling. The tuningof the windings is done by the com-pression condensers on the base andthe frequency may be shifted to 165or 185 kc. The output of this trans-former drops about 8 decibels oneither side of resonance. Severalstages of this would ruin the audiofidelity, hence the other transformersare less sharp and drop only about.3-d.b. at 5 kc. off resonance.

In short-wave receivers where gainis at a premium, it is not uncommonto find the I.F. coils made of litzen-draht.

Another type of I.F. transformeris shown in Fig. 40.

Still another variation uses smallair -insulated variable condensers in-stead of the compression type. The-oretically, the losses are lower and ifthey can be made to retain adjust-ment as well or better, such conden-sers should show some advantage overthe compression type.

The dimensions of one 175 kc.transformer appear in Fig. 41, thecoils being of the "lattice" or "uni-versal" type of winding so built thatif open circuited the grid coils reson-ate at about 850 kc. while the platecoils resonate at 880 kc.. In use,they are tuned by 150 or 200 mmf.condensers. The sharp transformeris shown at the right.

For home-made transformers-whichare emphatically not worth while at thepresent prices of good manufacturedones-one may begin with the construc-tion of Fig. 42, using 800 turns ofNo. 36 S.C.C. "scramble wound" intoeach spool, making one transformer

COPPERPLATE

IGO TO ZOO KC.NC 36 3.C.C.; C 140 IA MR )

33

FIG. 42Data for a home built I.F. transformer.

as shown, the others without the cop-per washer and with slightly widerspacing between coils. This is for175 kc. For 450, 500 or even 1,575kc. the number of turns is reduced,and the trimmer, or tuning condenseracross the windings also reduced;otherwise serious loss in amplificationresults. Here again, stick ratherclosely to a printed description.

PHASESHIFTINL

CON D

OFF-SERIES'"PARALLEL"

SWITCH

SELECTIVITYCONTROL

CONE)

I F TRIMMING

FIG 43

A National crystal filter. The switchat the left permits the use of a seriesor parallel connections as desired, or thecrystal may be switched -out, dependingon the selectivity requirements. Thecrystal itself is in the round container,the I.F. peaking condenser is reachedthrough the lower opening and the phase -shifting condenser through the upperone. The milled knob acts upon the sel-

ectivity -control condenser.

CHAPTER 6The Second Detector, Audio Amplifier and

Power SupplyTHE second detector, audio ampli-

fier and power supply systems ofsuperheterodynes differ little from thecorresponding parts of T.R.F. receivers.Any of the various detector circuits andamplifiers may be employed, the par-ticular type of circuit largely depend-ing upon the tubes used.

While there are some slight changesnecessary in superheterodyne operationas far as the second detector and theaudio amplifier are concerned, the pow-er supply system is identical to that inother receivers. It simply has to fur-nish sufficient filtered power at thevarious voltages required by the tubes.For this reason only the importantpoints concerned with the second de-tector and audio amplifier will be dis-cussed. A. study of the various com-mercial diagrams in Chapter VIII willreveal the similarity between these cir-cuits and those of T.R.F. receivers.

The Second Detector

The purpose of the second detectoris to detect the incoming signal presentin the intermediate frequency carrier.The theory is disclosed in Chapter I.In the majority of receivers a grid-biased or power detector is used. Thismay be a type 27 tube, transformercoupled to a single stage of push-pull

Second detector and audio amplifier circuit.In a later receiver, we might find the 27tube replaced by a 56 and the 45 tubesby 2A3s. Similar changes without differ-ence in operatibn may be included in

Fig. 45 and 46.

34

audio amplification using two type 45or 47 tubes; or if a single output tubeis employed, a screen -grid type 24 de-tector resistance coupled to the outputtube is sufficient.

In superheterodyne receivers theplate circuit of the second detector re-quires greater attention as regards fil-tering of R.F. currents, because thesecurrents are of a much lower frequen-cy (175 kc.) than is found in T.R.F. re-ceivers. This precaution is absolutelynecessary because of the large numberof heterodyne beats and harmonics inthe plate circuit of this tube. In manyreceivers heterodyne whistles in theloudspeaker originate at this point.(See Chapter I.)

In the detector plate circuit of mostreceivers an R.F. choke L and eitherone or two fixed condensers C are usedfor this filter. The diagram is illus-

values of theparts vary considerably in different re-ceivers, but condensers of .001 mf. aresuitable in most cases. The choke shouldbe larger than that used in T.R.F. re-ceivers.

The circuit of Fig. 44 shows a type27 tube operated with a plate voltageof 200 and a grid bias resistance of25,000 ohms, bypassed by a .1-mf. con-denser. This circuit shows a push-pullaudio transformer feeding two type45 tubes.

Combined 2nd Detector and AutomaticVolume Control

In some receivers the second detectoris connected as a diode, and is used toserve the double purpose of a detectorand automatic volume control. Theplate and the cathode of a type 27 tubeare connected together for operationin this manner. Rectification takesplace between the grid and cathode.Because of the high damping of the cir-cuit the I.F. transformer feeding thedetector is not tuned.

This type of circuit is shown in Fig.

THE SUPERHETERODYNE

45. The signal current, after being rec-tified, passes through the resistors Rand R1, causing a potential drop acrossthem which in turn applies a negativebias to the I.F. amplifier tubes. Thesetubes have a fixed bias, R3, wherebythey are adjusted for maximum sensi-tivity. As soon as a signal comesthrough, an additional bias is suppliedby the current passing through R andR1 which tends to reduce the gain inthe I.F. amplifier. The amount of thereduction in gain is proportional to thestrength of the rectified current in thedetector. This automatically maintainsa somewhat even level of volume fromthe loudspeaker.

The manifest difficulty of filteringa 175 kc. carrier out, while passing100 cycles (1 kc.) unimpaired to theaudio system, combined with the needof automatic volume control, and alsothe desire to limit the number oftubes in the receiver, has led to therecent introduction of the "Wunder-lich" tube (push-pull detector andtriode amplifier in one envelope) andto tubes which combine a diode de-tector, or a push-pull pair of diodedetectors with an audio triode or pen-tode. The thought in each case isthat the push-pull detector (whethergrid -leak type as in the Wunderlichtube or diode as in the 55 et. al.) isself -balanced and produces essentiallyno R.F. output-hence removes thefiltering difficulty to a large extent,even when resistance coupling is used.

At the same time, the D.C. detectoroutput (rectified R.F.) emerges fromthe cathode of the tube and can bepassed through a resistor across whichappears a voltage proportional to thiscurrent (therefore proportional tothe incoming signal's carrier) hencesuitable for feeding back to the ca-thodes or grids of R.F. and I.F. tubesfor the purpose of A.V.C. (automaticvolume control).

Finally, the audio output emergesfrom the plate of the triode or pen-tode section, available for furtheramplification. The designer and con-structor have their problems enorm-ously simplified thereby. A typicalcircuit is shown in Fig. 46, and com-mercial circuits in later chapters willshow further details.

BOOK 35

Combined second detector and automaticvolume control.

When such a tube is to be "graftedinto" an existing receiver it must beremebered that the detector will showless audio output than was obtainedfrom a tetrode in the same position,but will be superior to a triode de-tector. Therefore, consider the sensi-tivity available before changing.

A. V. C. KinksIn general the addition of A.V.C.

of any type is advisable only in setsof excellent sensitivity, say with asensitivity of better than 20 micro-volts, i.e. less than 20 microvolts ofR.F. signal with 30 per cent modula-tion needed to produce 100 milliwattsoutput. The tubes to be controlledshould be of the variable -mu type ifpossible, as the ordinary tetrodes(such as the 24 type) block too easi-ly. If the receiver does not use va-riable -mu tubes, in at least two sock-ets (R.F. or I.F.) it is better to for-get the idea altogether. It is usuallynot advisable to control the mixertube or the oscillator, although it isdone in factory -made sets with specialdesign.

Having picked the I.F. and R.F.tubes (usually two) to be controlled,one cuts the lower end of the tunedcircuit loose from chassis as shown inFig. 46B and introduces the R.F. fil-ter shown in Fig. 46C. This changemay seriously upset the tuning range,unless the greatest possible care istaken to keep the wiring of the samelength as before. Remember to countthe distance through the stopping con-denser, Cl, which should be as com-pact as possible, with a capacity ofnot less than .01 mf., preferably ofmica. Resistor R, may be of about


R.FT

1.5 MEG

T'\ABOUT

58 RPENTOOE

.5-1)-

MEG.

FIG.46

0.5- 1MEG

I.F

58 1 FPENTODE

LASTI. F

VOLUMECONTROL

257DIODE,

PENTODE

000ONmS

BIAS RESISTOR_

MS." COgREAUDIO OUT

AUDIOGRID

AUDIO

TONECONTROL

.000OHMS

.5- wit

DE -COUPLER AND(SCOUR) 250V DROP RESISTOR

In this diagram, the detector is not push-pull, hence the output must be filtered free ofR.F., especially as resistance coupling to the audio amplifier is used. Close study willshow that the 2137 diode -pentode works as follows: A-the I.F. signals rectified by thediode plates (acting as one) and the cathode, and the resulting D.C. and audio flowthrough the 50,000 ohm resistor and the volume control (about 14-megohm), as does theremaining R.F. B-the audio voltage across these resistors is fed, along with the R.F.voltage, through the volume control slider to the "audio grid"-but the D.C. is stopped bythe 0.1- to 0.5-nif. condenser. C-the audio grid, with the cathode and the plate, makeup an audio amplifier whose output goes through the 50,000 ohm resistor, thence the audiois fed to the audio tube's grid, but the choke blocks the R.F. D-meanwhile, the D.C.voltage across the volume control and 50,000 ohm resistor, is in part picked off and fedback to the R.F. and I.F. grids to be controlled. If the signal increases, so does the rec-tified output of the diode, hence this voltage raises and lowers the amplification in pro-portion, holding level constant. E-finally, the D.C. plate current of the tube flows throughthe 800 ohm bias resistor of the 2B7 and provides bias as usual. Since the amplifiedaudio also flows through this path, a high capacity bypass is provided, because of the

high mu of the tube. Figs. 46B and 46C are explained in the text.

0.2 megohms but is not critical. C2is merely a good non -inductive by-pass of about 0.2-mf., and the lengthof wires to it are not especially im-portant.

Time ConstantsThe "time constant" of an A.V.C.

determines the rate at which the sys-tem "takes hold" when the signalstrength changes. If, for instance, Itin Fig. 46C were made several meg-ohms and C3 were 2 "mikes," weshould find the system always taggingbehind, allowing a strong station toblast for a minute or more, and fail-

ing for an equally long time to godown after a weak station.

On the other hand if the system istoo snappy, we may easily get intothe maddening action formerly called"motorboating" - because it soundslike that devil's invention, the un-muffled outboard motor.

General rules are hard to lay down,but if either of these effects are met,start changing resistors and conden-sers one at a time, carefully observ-ing results on both tuning and fad-ing each time a change is made, so asnot to spoil the A.V.C. while curingthe motorboating.

CHAPTER 7

Practical Superheterodyne ConstructionBY this time the reader will feel

weary o f abstract theoreticalconsiderations and desire to see themassembled in a workable receiver. Forthe sake of simplifying the generalunderstanding of the subject, we shall,therefore, first show a perfectly work-able receiver whose design is some-what out of date-but only in thesense that it does not employ lateeconomy tricks. Its performance ishigh-grade in every way and comparesadmirably with the latest. It is in nosense either cheap or obsolete-but asound stra'ghtforward superhetero-dyne, covering the normal broadcastrange of 550-1,500 kc. By way ofconfirming this, compare the circuitwith the very high-grade commercialreceiver of Fig. 64. This diagram isshown in Fig. 47. It includes a singlestage signal frequency amplifier usinga type 35 tube, a type 24 first detector,a type 27 oscillator, two stage I.F. am-plifier using type 35 tubes, and a type27 second detector, transformer cou-pled to two type 45 tubes connectedin push-pull. A type 80 rectifier isused.

All the values are indicated on thediagram. If the reader cares to em-ploy different tubes he can easily doso if the proper voltages for the tubesare applied. This would probably neces-sitate changes in resistor values.

The single control tuning arrange-ment illustrated requires specially cutcondenser plates for the oscillator asoutlined in Chapter IV. If the designerprefers, a condenser padding arrange-ment also described in Chapter IV, maybe incorporated instead.

In midget set design, where spaceis a vital factor, it is possible to dropsome of the tubes from this circuitand still retain a high degree of se-lectivity and sensitivity. While thecircuit shows nine tubes (counting the

37

rectifier), it is possible to drop thesignal frequency stage and cut the totalnumber of tubes to eight. The use ofa single 45 output tube, or preferablya type 47 pentode, cuts the numberdown to seven. Even one of the I.F.stages may be dropped and the set willstill perform with six tubes. By com-bining the first detector and oscillatoras shown in Fig. 29, the total numberis reduced to five.

The rest of this chapter will be de-voted to the theoretical considerationsin the design of a receiver followed byactual constructional data of homebuilt sets. In selecting a set designthe reader is also referred to the chap-ter on commercial superheterodyne cir-cuits, from which source he will obtaina variety of ideas actually used in com-mercial receivers.

Theoretical ConsiderationsThree of the most important factors

to be taken into consideration in thedesign of a superheterodyne receiverare:

1. The sensitivity required to ob-tain the required power output fromlow signal inputs;

2. The degree of selectivity neces-sary per stage to give a satisfactoryover-all selectivity in the receiver; and

3. Mechanical and cost considera-tions such as chassis size, coil -shieldsize, number of tubes, etc.

The logical way to design the set is.first, to determine the required degreeof sensitivity. If we know the totaloverall gain required for a given out-put, we can ascertain the required gainper stage. We shall have a fair ideaof the grid swings on successive stagesat full power output, which will enableus to design our circuits for minimumtube distortion and maximum selectiv-ity and stability. The solution of the1st factor listed will be a guidepost in

OP

00

0005

-R

.F.A

MP

MF

..3

5

La,'

L2r

Li

LA,:i

i.7

1;. /

xx

MF

.MF

.

--

--.-

.

500

OH

MS

"'

100,

000

OH

MS

t+

5cO

,40,

A0s

0

TA

C:C

TN

VE

EA

,

l ST

[JE

T'2

4

Xx

8000

OH

MS

AC

.X

i 2

20,0

00O

HM

S

Y TO

.45

FIL

. 80

I, F

. AM

'35

fr''3

5

900

9000

OH

MS

OH

MS

FIG

.47

xx

-150

0O

HM

S

+45

V+

605

'--6

MF

.E

A.

P.F

.0

2HP

DE

T13

'27

T

M F

.

20V

OS

0

50 0

00

AU

DIO

AM

P.

45

+20

0 v.

+25

0

X 24.O

,RS

AN

D'2

7IL.

TU

BE

S

X

A c

ompl

ete

supe

rhet

erod

yne

circ

uit i

ncor

pora

ting

the

desi

gnfe

atur

es d

escr

ibed

in th

e va

riou

s pr

ecee

ding

cha

pter

s.

THE SUPERHETERODYNE

the determination of factors 2 and 3.Instead of using the conventional

level of 50 milliwatts output, we shalluse the rated power output of thetube or tubes as indicated in the vari-ous tables supplied by tube manufac-turers.

If the power tube selected is of the45 type, the power output will be 1,600milliwatts at the maximum rated volt-age. This means that if we want apower output of 1.6 watts (1,600 milli -watts) to be fed into the speaker, theinput signal voltage on the grid of the45 must not be greater than 50 voltspeak (the value of the grid bias). Anyincrease of voltage on the grid will bethe cause of undesirable distortion and,of course, must be avoided. It is bestto use R.M.S. values in calculating thevarious signal voltages, and as theR.M.S. voltage of 50 V. is .707 X 50= 35.35 volts, we find that the R.M.S.value which can be applied to the gridof the 45 is 35.35 volts.

Most radio sets today feed the audiooutput of the detector into the grid ofthe power tube by means of resistancecoupling; in this case, the detector willhave to deliver 35.3 volts to the gridof the output tube.

Figure 48 shows the circuit of apower detector, resistance -capacitycoupled to the output tube, and we findthat in the case of a screen -grid de-tector and a 45, E3 will be 35.35 R.M.S.volts. No gain can be expected fromthe resistance -capacity unit, so thevoltage at E2 must also be 35.35 volts.Figs. 49, 50 and 51 show the possibleaudio output of three standard tubesused as second detectors. These curvesshow the A.F. output volts (R.M.S.)

BOOK 39

Fig. 48. Elementary circuit of power de-tector and output tube.

of the 24, 27, and 32 tubes plottedagainst the R.F. input volts (R.M.S.)and are very useful in view of the factthat they give the required operatingpotentials for these tubes used as de-tectors and the required R.M.S. valuesof the incoming signals to "kick" thepower tube. Figs. 49, 50 and 51 alsoshow the points where grid currentwill start due to overloading of thegrid by the incoming signals.

Referring to Fig. 49, curve B, wefind that a signal of 3.24 (R.M.S.) voltsis necessary on the grid of the 24 de-tector to fulfill the requirements of the45 for maximum power output. Thesignal on the grid should not exceed4 volts R.M.S. or the grid will drawcurrent, thus causing distortion. Inthe case of the 27, Fig. 50, we find thatit would require an R.F. input of 12volts to deliver an A.F. output of 13volts. This tube will not satisfy thecondition of maximum power outputunless a high -primary -inductance A.F.transformer, with a turns ratio of atleast 3.5 to 1, is used. A bad featureof such a tube is the fact that grid cur-rent starts to flow at about 12.5 to 13(R.M.S.) R.F. volts. Under all con -

u)iaiot.1o

'; 50

X START

I

OF GRIDCURRENT

B

"' 18itci 16

IT 14

0> 12

,IR 10

x . STARTCURRENT

X

OF

lirGRID

C

oil-2 28

?. 24

k 20

EbEcoESG

32 TUBES174-67.SV.

v.G V.

IIIP.rB'24 TUBES '27 TUBESa 40Sj S 60 '160CURVE -AA CURVE -B IVCURVE

Eb ECGA 45V. -5V. 112

';'. 30 t, 6 ). 12.roArffM

Eb 180V. Eb 180V.Ess 45V. ESG.z 75V.EcG.-4.5V. EcsF V.

rA /4Y mt=AKA.." 2

w 40op CURVE B 90V. -9V.

CURVE C 180V...

0 .PLATE

Mrir4ORE ',a PLATE PMFur 10

LL 0

LOAD .2- MEG. a 2-,.. 0

LOADs.2-MEG. g 4LLD1E211 rAdll.0 ra

g 0 1 2 3 4 0 0 2 4 6 8 10 12 0 0 2 3 4'A RADIO FREQUENCY INPUT- 3 RADIO FREQUENCY. INPUT- 0 RADIO FREQUENCY

VOLTS R.M.S. .1 VOLTS R.M .5 . a INPUT VOLTS R.M.S.

Fig. 49, left. Curves showing the possible gain that may be expected from a 24 tube.Fig. 50, center. The same curves as in Fig. 49, only for a 27.fig. 51, right. Same as for Figs. 49 and 50 but for a 32.


ditions, it is advisable to work the tubeat some value below that which causesthe flow of grid current.

It it is desired to use a type 47 pen-tode as the output tube with a screen -grid second detector, we find that anR.F. signal input of less than 2 volts

more stages of conventional T.R.F.ahead of the modulator tube (first de-tector), it is not absolutely necessarythat the entire burden of amplificationbe borne by the I.k . amplifier. Ifthere are two stages of T.R.F. aheadof the first detector, then there will be

Fl F. I40

El =.00002-V.

RF2 DET.1 I.F. 1 0.0.2 DET.244 11+ w.. 11+ AUDIO OUTPUT

02..0008V.E3 0.032-Y. 4 0 032-V. ES 0.352.V. 06.4.0 V.

Fig. 42. Elementary circuit showing the various voltages developed at each stage.

will be sufficient to deliver a poweroutput of 2.5 watts.

If push-pull circuits are used in theoutput stages, the A.F. signal voltageswill have to be doubled and, as the out-put of the detector cannot be increasedwithout severe distortion, it is neces-sary to add an additional A.F. stageor high ratio transformer so as not tooverload the detector.

Calculation of Gain

Having determined the minimumR.F. voltages which must be suppliedto the grid of the detector to deliverthe maximum power output, we are ina position to determine the total gainwhich must be obtained from the I.F.amplifier.

Modern radio receivers of the super-heterodyne type have an input sensi-tivity of less than 5 microvolts permeter and, with the standard height ofthe antenna set at 4 meters, we findthat the absolute sensitivity will beabout 20 microvolts (a microvolt be-ing one -millionth of a volt). Thus, ifwe desire a receiver that will deliverabout 4 volts of R.F. signal to the de-tector from an input signal of 20 micro-volts, the total voltage gain of the am-plifier will be

R.M.S. volts on grid of detector

R.M.S. volts input from antenna4 volts

or 200,000 gain.000020 volt

As a certain amount of amplificationcan be, and is, obtained by one or

a voltage gain of about 1500 (assuminga gain of about 40 V. per stage)which must be considered in designingthe I.F. amplifier. The reader will re-cognize the necessity of using pre -amplification before the modulator asthis phase has been covered in anotherchapter.

Now let us see just what the finalfigures will be with the added gain ob-tained in the pre -amplifier.

If the input of the receiver is .00002 -volt and the pre -amplifier has a gainof 1,500, then the input to the first I.F.transformer will be .00002 X 1500 or.03 V. The 4 V. required by thedetector, divided by the .03 V. inputto the I.F. amplifier, will then be thevoltage gain required by the I.F. am-plifier, which is 133.3 V.

As it will be impossible to obtain again of 133.3 in a single intermediatestage, it will be necessary to use twostages working at a gain of about 65,or three working at 44 per stage.

In the example cited above, the am-plification due to heterodyning is ig-nored, as various conditions developwhich cause the gain of this portionof the circuit to vary over wide ranges.The sensitivity and output will be af-fected by the strength of the receivedsignal, by the power output of the localoscillator, and by any change in oper-ating potentials which may take placeas the receiver is functioning. Hetero-dyning, however, will cause an increasein gain and give greater sensitivitythan these calculations will indicate.

The check for the correctness of thecalculations can be made by multiply-


1401/ A C.

Fig. 53. Circuit of a homemade midget superheterodyne designed by Mr. H. G. Cisin.

49

41

ing the gain in the pre -modulatorby the gain in the I.F. amplifier; thus1,500 X 133.3 gives a value of 199,950.

Figure 52 shows a skeleton circuitwith the voltages developed in thevarious circuits. Two stages of I.F.amplification are shown and, as eachstage is not working at the maximumpossible gain, the I.F. amplifier willbe very stable.

If an actual condition exists wherethe gains and voltages are measuredand found to be as indicated in Fig. 52,the volume control on the pre -amplifierend of the receiver will be full and thegain on the I.F. amplifier cut awaydown.

If the signal frequency amplifier islimited to one stage, it will be neces-sary to increase the gain of the I.F.amplifier if the same level of sensitivityis to be maintained.

Building an A.C. Midget Super -Het.

The following instructions are pre-pared from a set designed and built byH. G. Cisin and represents a typicalcircuit for a home -built midget super-heterodyne. This is an eight tube set(including the rectifier), employingscreen -grid tubes, band selector tuningand other interesting features of de-sign. The complete data are given inthe various illustrations, Figs. 53 to 58inclusive, and the list Of parts. It willbe noted that single dial tuning is em-ployed without the usual form of os-cillator condenser padding. If thebuilder does not use the particularcondensers specified he may use other

makes with the usual padding as out-lined in Chapter IV.

First, the aluminum sheet should bedrilled for socket holes, mounting holes,etc. It is then folded so as to form achassis of the required dimensions, Fig.56. The eight sockets are mountedfirst. Wafer type sockets mountedfrom underneath are recommended. Thetwo condenser blocks are mounted, thenthe midget condensers (7) and (35)and the by-pass condenser (38). Theresistors (45) and (51) are mounted inan upright position. The Super-Tona-trol (18) is mounted on the front chas-sis wall. Next the two midget conden-sers (8) and (36A) are mounted in thepositions indicated. The three pairs ofintermediate coils are mounted on thethree duplex condensers and the threeunits are then fastened in position asshown. The various small parts, suchas the R.F. choke (36), the wire gridresistors (12, 17, 24), the "V" resistor(51A), and the five metallized resis-

arFig. 54. Layout of the midget super -het.


35";;',sr0O 7

tors, are soldered in place during theprocess of wiring. In order to facili-tate the wiring, the transformer (39)and the special coil (3 -A -B -C -D -E) aremounted after wiring in most of theother parts.

Before starting the wiring, the chas-sis is turned right side up and thethree -gang variable condenser (4, 5, 6)is mounted at the front center. Thepower supply transformer (50) andalso the seven binding posts are thenmounted. The power switch (53) andthe amperite (52) are wired in first andthen all the filament circuits. All pairsof filament wires should be twisted.The grid circuits are wired next. Notethat the double grid connections aremade at the socket terminals. Flexibleconnectors terminating in clips areused for making the connections to the

SPECIAL COIL SPECIFICATIONS.I *29 ENAMELED WIRE-

0141,1'1+4w 4- 4.024-'1.1E1 1111;111',11.111

TOMotTO TOC""14 TO TO

PLATSCHASSIS CHASSIS STATOR

TO CHASSIS 4"(*CIL,COW DTO

ANTENNA TOTOSTA

TToaS 8toF*6&*8 Gortsicpir °SCH.

COND.

/ IQ- Si -4FIND-ALI. INTERMEDIATE COIL

Fig. 58. Coil data for the midget super -het.

11)1.1'1111 11111111111111111111111M1

Fig. 55, left; view under-neath the chassis.

Fig. 56, right; dimensionsof metal chassis.

Fig. 57, below; front viewof chassis.

CHASSIS DETAILSA.G. MIDGET SUPER -HET

FRONT VIEW OF CHASSIS4,

A. C.pol ID GETSUPER-HET

caps of the screen -grid tubes. The flex-ible connection going to the cap oftube (11) should be fastened directlyto the stator of the variable condenser(5). The other three flexible connec-tions from the trimmers are broughtup to the caps of the screen -grid tubesthrough small holes drilled in thechassis. Plate circuit wiring follows,then the wiring of the cathode circuits,negative returns and by-pass conden-sers. Finally, the antenna circuit,band selector and power supply wiringis completed. Connections are broughtup to the power supply transformer andthe variable three -gang condenserthrough small holes drilled in the alum-inum chassis. All other wiring is con-cealed beneath the chassis deck.

In wiring the transformer (39), onlythe secondary is used. The special coil(3) is wound on a composition form11/2 ins. in diameter using No. 28enameled wire. The finished coilsmount underneath the chassis, direct-ly under the variable gang condenser.After the coil is wired in and all wir-ing has been checked the speaker shouldbe connected to the set, aerial andground connections made, tubes insert-ed and the line plug connected to the110 V. A.C. outlet.

The first step in adjusting the re-ceiver is to regulate the trimmers to


about % of their maximum capacities.Next, turn the tuning dial, with thevolume control on full, until a broad-cast station is heard. The R.F. choke(36) should be turned to the same fre-quency as the intermediates, by ad-justing condenser (36A). Then adjustcondenser (8) so that the oscillatortuning tracks with the band-pass tun-ing condenser. If the oscillator doesnot track the intermediate tuning con-densers should be adjusted to a higheror lower frequency. The use of astandard test oscillator will speed upthe work of adjustment.

List of Parts Used in the A.C.Midget Super -Het.

1-.350 mmf. (each section) CardwellTriple Variable Condenser, type317-C (4, 5, 6)

3 -400 -ohm Electrad Wire Grid Resis-tors (12, 17, 24)

1-Electrad Super-Tonatrol, No. 3 (18)1-Electrad Fixed Resistor, type B-8

(51)1-Electrad Fixed Resistor, type B-250

(45)1-Electrad, type V-20 Center -tap Re-

sistor (51A)3-DeJur-Amsco small duplex Varitors,

type 341 (13-15) (20-22) (25-27)2-DeJur-Amsco single Varitors, type

X -71-A (8, 36A)2-.001-mf. Aerovox mica type 1460

(7,35)1-.1-mf. Aerovox Condenser, (38)1-Aerovox Condenser Block, type B-4

(19, 29, 30, 34, 54) each 1-mf.2-Aerovox Electrolytics G-5-8 (Two

4 mf. units in parallel -46) (2 mf.-49)

1 -2000 -ohm I.R.C. (Durham) Metal-lized Resistor Powerohm, typeMF-4 (10)

1 -6,000 -ohm I.R.C. (Durham) Metal-lized Resistor, type MF-4 (31)



1 -250,000 -ohm I.R.C. (Durham) Met-allized Resistor, type MF-4 (37)

1-Amperite Line Voltage Control,type 8-A-5 (52)

1-Special Coil, Wound according todirections-Antenna Primary (3A);Antenna Secondary (3B); Band Pass

48

z

4 'al 0

00.00)x-flynl^r


Secondary (3C); Oscillator GridCoil (3D); Oscillator Plate Coil (3E)

3-Pairs Find -All Intermediate Coils(14A -14B) (21A -21B) (26A -26B)

1-Trutest R.F. Long -Wave Choke (36)1-Trutest Power Supply Transformer

for furnishing Plate and FilamentVoltages (50)

5-UY-type 5 -prong Sockets (9, 11, 16,23. 32)

3-UX-type Sockets (40, 41, 53)1-Special Amperite Socket, or 1- or-

dinary UX-type Socket (52)1 -27 -type Tube (9)

INPUT COILS ( 2 FOR EACH RANGE)Ll OR 13 (WIND OVER

OTHER WINDING WITH ONE LATER-7-OF OILED SILK

BETWEEN

1/2*-..112 OR L4

OSC. COILS ( I FOR EACH RANGE)

L6-r1

JUMPER 2.56 SCREW

STIFF WIRECONNECTION

TO L5

V4" HOLEFOR BAKELITE

SCREWDRIVER.

35 MMF, SHUNTTRIMMER (.CROSS

L2 on 14

FIXED MICA(MAKE UP WHEN

NEEDED)

VARIABLE, PARTOF C5

(SEE FIG. 61)

FIG. 60

Coil details for the circuit of Fig. 69

4 -24 -type Screen -Grid Tubes (11, 16,23, 32)

2 -45 -type Power Tubes (40, 41)1 -80 -type Full -Wave Rectifier Tube

(55)7-Binding Posts (1, 2, 42, 43, 44, 47,

48)1-Aluminum Sheet, 18 to 20 gauge, 17

x 19 ins., cut and bent so as toform a chassis 10 x 12 x 3% ins.high.

1-High-Grade Input Push -Pull AudioTransformer, Using Secondary Only(39)

1-Power Switch, toggle type (53)1-R o 1 a Midget Electro-dynamic

Speaker Model K-6-equipped withPush -Pull Voice Coil Transformer(40-75 volts)

AN ALL -WAVE SUPER-HEWRODYNE

We will now go to a more recenttype of receiver, which at the sametime gives good performance at short-waves, by virtue of plug-in coils. Itis thoroughly realized that switchinghas mechanical advantages, but thedifficulties of getting good perform-ance with switching are ,not to be ig-nored, and the home builder will, as arule, do far better with plug-in coils,where each range can be attacked asa wholly separate problem, not inter-locknq., it, other bands.

The general nature of the circuitis shown by Fig. 57. By reason of itstuned input coil and its stage of va-

COIL DATA FOR FIG. 59

INPUT COILS NOTES :-ALL LL & 13 COILSSPACED TO BE

1 INCH LONG.ALL L2 8.. L4 COILSSPACED TO BE

11/2 INCH LONG.ALL ,L5 COILS SPAC-ED TO BE 11/4 IN. LONG

ALL L6 COILS SPAC-ED We IN. CONG.

WIREN9.24 D.S.C. FORL2, L4, LS.N2. 36 D.S.C. FORL1.L3,L6

OSCILLATOR COILSRANGE

IN METERS Ll & L3 L2 V L4 L.5 L6 CS ( SEE NOTE2 BELOW)

10-20 3V2 5 5 3 NONE,CONNECTSTRAIGHT THROUGH

20-40 8 12 II 5Y2 .002- FIXED

40-80 16 25 22 11.001 -ME MICA INPAR. WITH ISOMME

VARIABLE.

80-160 32 * 50 45 22 t,

160 -.320 60 * 100, N2.36ONE

LAYER

90, NE.36,ONE

LAYER46

300MMF. MICA INPAR. WITH 130MMF.MF. VARIABLE

275-550 60 LIE

170, N236-scRAmBLE

WOUND'.

153,N2.36-SCRAMBLE

WOUND"80 I

NOTE: 2 :- THIS IS THE SERIES TRIMMER- BETWEEN L5 ANO THE 100. MMF CONDENSER.IT MOUNTS IN THE COIL AS IN FIG. 60.

* ANT. SERIES CAPACITOR NOT OVER 50 MMF, FIG. 61


riable-mu R.F. amplification followedby a tuned detector input, this re-ceiver is relatively quieter and morefree from image interference than arethose primitive sets which use no R.F.stage at all, or else use a "couplingstage" with resistance input. The in-put circuit is ordinarily thought to bedifficult to "track" because of varia-tions due to assorted antennas, butis here avoided by use of the spec-ial antenna winding due to DanaBacon, and introduced in National's"FB-7" receiver. In this winding, theterminals are reversed, putting theantenna end of the primary near theground end of the secondary, wherecapacity differences are ineffectual.The top end, which could have an ef-fect, is grounded and has a constantcapacity effect, regardless of the an-tenna.

The rest of the set is tolerably or-thodox. The mixer is of the penta-grid-converter type already described,that is it consists of a mixer fed byL4, which is electron -coupled to theoscillator working with coils L5 andL6, the combined output being in turnelectron -coupled to the plate whichfeeds the I.F. transformer L7, L8.Since only a single I.F. stage isneeded and used, all windings of bothI.F. transformers are tuned, the fre-quency being chosen as describedhereafter. The I.F. output is fed tothe diode -pentode second detector andA.V.C. tube, whose operation has al-ready been described. The D.C. out-put of this tube provides A.V.C. forthe R.F. and I.F. tubes as previouslydescribed, while the audio output isfed through a resistance coupling tothe first A.F. tube, a triode suited toheadset reception when desired. Withthe headset plug withdrawn, its out-put is fed to the 59 triode, which issuited to loudspeaker driving. Ifconnected as shown, a magneticspeaker may be put into the plate cir-cuit, preferably through a condenserand choke coupling device. However,the fidelity of the system is ampleto deserve a good dynamic speaker,for which must be provided a trans-former suited to both the speaker andthe 59 tube-remembering that the59 is connected as a triode here. If,

for any reason, anyone imagines thatmore audio gain or output is needed,it may be attained at some sacrificeof fidelity by connecting the 59 as apentode, (see maker's tube data), atthe same time changing the outputtransformer to provide suitable im-pedance matching also replacing thecathode bias resistor by one of 500ohms, and using a bypass of at least12 mf., low -voltage dry electrolytic.

TrackingIt is reasonably clear that differen-

ces between individual coils in theT.R.F. circuits (L2 and L4) will re-quire that the trimmer condensers beset differently for each coil. There-fore, they must be attached to thecoils and be removable with them. Infact, they are ordinary "screwdrivertype" condensers of about 20 mmf.maximum capacity placed inside thecoil forms which carry them, as inFig. 60.

To avoid special oscillator tuningcondensers, we use the scheme of Fig.32, slightly modified to avoid an ex-tra plug -contact. The dimensions ofthe coils are shown in Fig. 61.

The I.F. transformers are perfectlynormal ones of a design intended tooperate near 450 kc. Both primaryand secondary of both transformersshould be tunable as shown. It isstrongly recommended that these bepurchased ready made, with shields.

While on this subject of shielding,it is well to mention that the coilsL1 -L2 and L3 -L4 must plug into sock-ets which are separated by goodshielding; the best (though most in-convenient) form being to place a

L2 ORL4 tiTO A.v.0

LINE o.1-mEG

COIL BASE

LI ORL3

TO 58 GRID ANDTUNE CONDENSER

0.5.Mr-= 5016"

0

FIG. 62

TO 2A7 INPUTGRID

TOP5L5x4TREF.


Plug-in coil connections for the oscillator.

metal cylinder of about 3 in. diam-eter and 4 in. tall around each socket.The screen -grid tubes, in fact all thetubes except the A.F. and rectifiertubes, should also have the customarymetal shields. The chassis must, ofcourse, be metallic, and the customaryprecautions must be taken to screenthe detector and audio tubes fromtransformer vibrations, or stray mag-netic fields from the transformer orchokes of the supply system. Leav-ing them loose until everything elseis done and then juggling their rela-tive positions is helpful.

Power SupplyThe filter capacities may be larger

than shown, anything up to 10 mf.each is an improvement. If a dynamicspeaker is to be used, choose a typewith a 6000 to 7500 ohm field andconnect this across the second filtercondenser, at the points Z,Z.

Under no circumstances use a mer-cury -type rectifier tube. The 80 andthe 5Z3 are the best types to date.

Mechanical FeaturesNo one thing has been the cause

of so many short-wave failures as abad tuning drive. The writer, in allseriousness, is of the opinion thatonly the "string drive" and the Na-tional triple -pinch (Model A and N"Velvet Vernier") are altogether upto the job. If any unevenness what-ever can be felt-throw it out! Thevernier ratio need not be more than5/1 if it is smooth-but a roughone can be 200/1 and still be hardto adjust. The tuning condensershould be a 3 -gang type with a per -

section capacity of about 150 mmf.If such a gang cannot be had, donot tolerate a flimsy makeshift-least of all one with a sectional shaft.Instead, use a stout, thoroughlyshielded, broadcast type and removeabout 1/2 of the _plates from eitherthe stator or the rotor (not both) ofeach section, treating all sectionsalike and bending nothing.

Adjustment

Having set up and found that with-out tubes approximately normal volt-ages appear everywhere, put in thetubes, run up the volume control and,with the 200-400 meter coils in place,trim the R.F. circuits (trimmersacross L2 and L4) ; then the 4 I.F.trimmers, working for maximumbackground noise. The tendency willbe for the A.V.C. to allow strongnoise as long as no signal is tunedin, hence do this in the daytime, nearthe lower end of the tuning rangeand tuned off of any broadcastingstation. In the case of the oscilla-tor, start with C4 set about half wayout and trim with C5 only. Nowswing to the 400 meter end of thistuning range and re -trim the oscilla-tor with the aid of C4, also checkingthe adjustments of the trimmersacross L2 and L4. If they have tobe moved, leave them halfway be-tween the second and first settings.

Now tune in a signal, one that will"stay" put" and try first a 50 foot an-tenna, then a 10 foot one. If theA.V.C. is working properly and thesignal is fairly strong, there will notbe a great deal of difference. Ifthere is-check over detail by detailto find the error. Especially checkback to see if the socket connectionsare as shown in Figs. 62 and 63;since a surprising amount of fair re-ception is possible with one coil quiteincorrectly connected.

If something seems not to be work-ing, try localizing it as follows: re-move the R.F. and mixer -oscillatortubes. You now have simply a long -wave receiver. Connect an antennato the plate prong of the 2A7 socketand see if the receiver sounds "alive"in this way, also retrim the I.F.against noise in this way.

CHAPTER 8

Commercial Superheterodyne Circuits

Achapter such as this is behind thetimes as soon as printed. The in-

tention, therefore, is not so much toshow the very latest as it is to showcircuits that are absolutely sound,free from anything freakish, and re-presentative of several general typeslikely to be with us for some time.

Bosch Model 250 and 251The Bosch model 250, (Fig. 64)

besides being an excellent example ofa straightforward superheterodyne,illustrates several practices not yetmentioned.

1-It shows the use of a heatertriode as a diode -triode detector andA.V.C. tube.

2-It shows the tendency for thebetter sets to give up the audio pen-tode and return to the distortionlesspush-pull triode with a first A.F.stage.

3-A tuning meter or resonance -indicator is illustrated. In this case,it is in the cathode return of theR.F. and first I.F. tubes, the maximumchange in meter reading obviously in-dicating resonance. Without this

meter, a highly effective A.V.C. setis frequently mis-tuned, thus missingthe quiet reception that is possiblewith such a set when "on the nose."

4-An excellent example is givenof the de -coupling of the variousstages by taking their plate suppliesoff from different parts of the filter,also by using decoupler resistors,such as the one in the lead to theaudio driver plate.

5-The use of dual speakers is il-lustrated, also the method of excitingtheir fields with the least possible in-terlocking.

6-A type of tone control is shown,which not merely cuts down the highnotes, but also lifts the lows.

7-A dual -tuned input circuit whichis easily aligned is shown.

8-An oscillator of the Colpittstype is used.

A number of the features incoe-porated in this receiver are attain-able in other ways, which have al-ready been described, or which areillustrated by Fig. 69 in this chapter.

The circuit diagram of the Bosch model 250 receiver demonstrating A.V.C. and resonanceindicator.

47


3-16

110 7,AC.

OVEN 240000 RANGE OSC 4 ...ERcR F.78 ,3.4-a 6A7

-48%

FIG 65E L 0

'1.4a!° `7:7.

NOTE

CSTAPrZtlAUS640 ONLY 77% OETHE VALUES SHOWNFOR 60 CYCLES

.24.1EG(EACH)

RFC US-Air)

't39 07.425Z5

FILAMENT SEQUENCE

AUDIO .01 -ME43

WucECON

78 6A7 43

The A.C.-D.C. principle of operation is illustrated by the RCA -Victor model R-22 set.

The Victor R-22 Universal A.C.-D.C.Receiver

Quite a different sort of receiveris the recently developed "universal"sort which may be operated fromeither a D.C. or A.C. line at will,and are without transformers in thepower supply system. These sets arebased on indirectly heated tubesdrawing at 6.3 V., all theheaters in the set being wired inseries. Not the least important tubein the set is a rectifier of the samefilament demand, although at highervoltage. This rectifier has two in-sulated cathodes and two plates, sothat it can be used as a 4 -terminal or"bridge" type of rectifier; approxi-mately doubling the input voltage,less resistance drop. Thus 110 V.A.C. fed into the device will emergeas something like 160 V. after alllosses are accounted for. When work-ing on D.C., the rectifier tube be-comes needless and if left in wastesvaluable voltage-of which there islittle enough anyway. In the Victorreceiver shown in Fig. 65 it is, there-fore, switched out in part. In othersets of this general sort, the 25Z5is used as two half -wave rectifiers,one of which feeds the tubes and theother feeds the speaker field. Whenused on D.C., these things stay wherethey were and merely lower the volt-age slightly. Thus the scheme shownin the diagram may supply 150 or160 V. with A.C. line feed, droppingto perhaps 90 with a D.C. line, whilethe other scheme just mentioned will

show perhaps 98 and 108 respective-ly, the output tube being consideredboth times. Obviously one type per-forms more nearly the same, but onecan object that it is always handi-capped, while the switching type atleast has a chance, on A.C.

Referring to the diagram of Fig.65 will show that there is no noveltyin the scheme once one has becomeaccustomed to the abnormal powersupply. Otherwise, it is an orthodoxsuperheterodyne.

The National FB-7 and AGSShort-wave Receivers

To break the monotony we now gooutright to a pure short-wave re-ceiver, intended for nothing else, andin fact so designed that it cannot beconveniently adapted to 550-1,500 kc.use. This is the National "FB-7" re-ceiver, a diagram of which appears asFig. 66.

The principle points of interesthere are the two electron -coupled os-cillators, one of which performs theusual function of a superheterodyneoscillator - frequency conversion,while the second performs the job ofmaking a carrier audible as a whistleby beating with it in the 2nd de-tector-which after all is merely an-other heterodyne frequency conver-sion of 1000 cycles-i.e. 1 kc.

It should be observed that this os-cillator can be readily cut off; alsothat the receiver has another switch(SW. 2) for the purpose of cutting


The National FI3-7 receiver is exclusively

off the plate supply during transmis-sion, so as to avoid noise.

The connections of the tuning andtrimming condensers in the an-tenna and oscillator coils are seen tobe unusual. This is for the purposeof "bandspreading" as has been dis-cussed in an earlier chapter. How-ever, the coils are provided with baseshaving enough prongs so that all ter-minals come out, hence it is possibleto plug in coils having the tuningcondenser connected across the wholewinding in the usual way. This ex-pedient makes the receiver both acontinuous range and a bandspreaddevice.

In this case, the receiver is notequipped with an R.F. stage or a pre -selector, but, (in a more ambitiousform known as the "AGS") it hasboth of these things. The three plug-in coils thus made necessary are in-serted as a unit. This set also hasA.V.C.

General Electric K-80 SeriesQuite another sort of receiver, in-

tended for a different purpose is theGeneral Electric K-80, a diagram ofwhich appears in Fig. 67. Unlike theAGS and FB7 receiver it is not in-tended for the use of a transmittingoperator, but primarily for broadcastlistening, at standard as well as short-paves-and in the export form forlong waves in addition.

Almost automatically, it is based onan I.F. near 450 kc., but it differsin every important way from most

49

for short waves. The wiring is shown here.

receivers at present available for thesame purpose. There has been a per-fectly unaccountable tendency tobuild receivers which use stages ofR.F. pre -amplification, and fewer tun-ed circuits at short waves than atstandard waves. Since noise, imageinterference, and relative amplifica-tion all demand the exact oppositeof this, it is pleasing to see that inthe K-80, the R.F. stage used in thebroadcast range is not merely re-tained, but is supplemented by an-other extra stage-a tuned one-inthe shortest waveband. This is a longstep ahead, and greatly improves re-ception.

In other words, the set conformswith things already said, and there isno need of repetition. The constantslargely appear on the diagram.

The Stromberg-Carlson No. 33Automobile Receiver

The Stromberg-Carlson automobileradio receiver is diagrammed in Fig.68, not only to show a set of this class,but also to illustrate another use ofthe 6A7 and 6B7 tubes-which differfrom 2A7 and 2B7 only in filamentvoltage and current. We have here the6A7 acting as a pentagrid converter,feeding an I.F. transformer as usual.However, there is no I.F. tetrode asusual. Instead of I.F. is fed at onceinto the pentode part of the 6B7 andby it is amplified at the same frequen-cy, then detected by the two diode sec-tions-the reverse of the examplesshown before. There is another differ-


ence, the two diodes work separately,the output of one being used to feedthe audio amplifier, while the otherdiode has the audio and I.F. filtered outand "wasted," only the varying D.C.being used -of course, for A.V.C. pur-poses. We are not yet finished with thedifferences between this and ordinarycircuits; the A.V.C. control applies notonly to the single R.F. stage but alsoto the 6A7 converter and to a lesserdegree to the 6B7 itself.

The plate supply has intentionallybeen omitted, as it is a rather conven-tional vibrating pole changer plus fil-ter, the main point of interest beinga relay which throws a load across thefilter until the tubes heat up and takeenough power to prevent excessivevoltages.

The Howard Model 45 Receiver

Dropping back again to simpler tubetypes, there is shown in Fig. 69 a re-ceiver which illustrates clearly ascheme of A.V.C. that can, with care,be added to most receivers having goodgain. It must be understood atype 51 tube is the same as a 35, andthat either can in this diagram be re-placed by a 78 without important con-sequences. Similarly the 27 may be re-placed by a 56 and the 47 by a 2A5,although the last change makes an un-important change in the biasing meth-od, which is obvious.

The values of the components of thisreceiver chassis, Fig. 69, are as follows:Resistors R1, R3, R5, 1/5-meg. (1/4 -watt); R2, R6, 500 ohms (1/5 -watt);R4, 6,000 ohms (1/4 -watt); R7, 30,000ohms; R8, volume control, 1/4-meg.; R9,1/4-meg.; R10, 3,000 ohms; R11, 2,000ohms; R12, R13, .15-meg. (1/4 -watt);R14, 2 megs.; R15 -R16 -R17 -R18 -R19,voltage divider, 9,900 ohms; R20,R21, 10 ohms (center -tapped); R22, 200ohms.

Condensers C4, C5, C6, C7, I.F. trim-mers; C8, C9, C10, C15, C16, 0.1-mf.;C11, 250 mmf.; C12, .001-mf.; C17,C18, 0.25-mf.; C19, C23, 0.5-mf.; C21,.05-mf.; C24, 1. mf.; C25, C26, 8 mf.(420 volts); C27, 4 mf. (420 volts).

In the interest of obtaining best re-sults with the A.V.C. receiver, it is im-portant that the type 27 control tube

V9 be a selected one, with a definiteplate current cut-off when tested at180 V. plate and 20 V. bias on the grid.This cut-off should be less than 5 micro-amperes. If there is no means avail-able for checking the tube (in the formof a special tube tester), an immediatecheck for tube performance can be ob-tained in the set itself.

For instance, disconnect the antennaand short-circuit the aerial lead, leav-ing the control tube out of the socket,and note the swing of the tuning meter.Then insert the tube in the socket andif it is a good A.V.C. tube, there shouldbe no change in the position of thepointer on the tuning meter. If thereis a change in the position of the tuningmeter pointer, namely, a swing towardthe right, it is an indication that theA.V.C. tube does not have a definiteplate cut-off instead, it is drawing platecurrent and as a result the bias voltageon the regular R.F. and I.F. tubes hasbeen raised, with the consequent cut-ting down in plate current.

The automatic volume control func-holding the second -detector in-

put voltage at a definite level, a systemwhich is different from that in otherreceivers. A reduction of back -groundnoises, between stations, will be noted.

The A.V.C. tube is so connected bymeans of a 2 meg. resistor, R14,that the grid is at absolute "B-" po-tential. The cathode of the tube is con-nected to a point on the voltage dividerwhich is at 24 V. positive, with re-spect to "B-" or the grid. There thenexists between the cathode and thegrid a potential difference of 24 V.with the grid negative by this amount.The plate of this tube connects toground by means of two .15-meg.resistors, R12 -R13. Since ground isconnected to 124 V., positive (withrespect to "B-"), there exists betweenthe cathode and the plate a potential'difference of 100 V. In order tobypass any R.F. energy which may ap-pear on the plate, a non -inductive con-denser C22 is connected from the plateof the A.V.C. tube to the cathode.

With the condition of no -signal thereexists a bias of 24 V. and a platepotential of 100 V. Under theseconditions, there is no plate current

LI.

L6

3C1,

ILT

C2

I64L

N,

L419

C4

.05-

M1

Zr

ET

R F

AM

P. 5

8C

714

.355

mm

V

CE

S /

C6V

50t5

IHrA

i&

O'

7 LI

GH

T

71

220

J

CO

NN

EC

T O

N

250

n~ 110V

CoN

HE

C1

710.

1

55

0310

0.00

0

C13

0. 7

.1,7

1T

RECT.

80

50.0

.55,

*

CS

918

MP

C60

10 1

.4F

,

07

-t-4

°,.

(Ks)

137

TT

O.

'C23

..054

411

138

9504

.01

(1.0

CO

IL

FIG

. 67

DO

T.

& 0

5C

`247

C27

50..4

7

C25

C29

G S

AC

E26

1000

.

.40-

220

P14

7

(SA

CS

)47t n

(..<

.,

(44

1*N

IP y

NA

587

C43

IIX

0M

PP

I240

04

vVA

k R

2223

I ME

GC

28 C

3C7C

33.C

36,C

50 4

-20

N.O

F02

9, C

32,

C35

10.

1503

06F

I NE

GC

51 3

00 1

4140

554

125

0 N

NE

037.

225

0 6.

641`

C39

. 254

0 34

.41.

1226

1500

0 n.

73/

11

C57

.05-

64f

V

C

0000

0753

L33

1.34

16. 7

/,.-1

) T

C4S 40

.0 D

E6.

022

40 6

, -18

7 L35

554

003

s

K R2I

2000

n.

071


flowing and the tube is said to be ad-justed to cut-off. Since no plate cur-rent is flowing, there exists no voltagedrop across the plate circuit resistorsand, therefore, there is no bias voltageon the grids of the controlled tubes.The only bias on the R.F., first detector,and I.F. is caused by the respectivevoltage drops across their cathode re-sistors. These resistors are designedto give the most sensitive operatingpoint.

In the case of a received signal, en-ergy passes through the receiver tothe second -detector grid. Here theA.V.C. (automatic volume control) tubegrid, and the second -detector grid, arein parallel. The signal voltage is fedto the grid of the A.V.C. tube througha small fixed condenser, Cll.

It will be seen that during the posi-tive half of the incoming cycle, thepeak voltage of the signal swing sub-tracts from the original bias voltage;which means that the instantaneousbias ,n the tube is less than the orig-inal bias and the tube begins to drawcurrent in its plate circuit. Since thiscurrent flows in the resistors in theplate circuit Of the A.V.C. tube, thereexists a voltage drop across these re-sistors; also, the flow of the electronsis from plate to ground so that theplate becomes negative with respect toground. Now, since the original po-tential of the cathode of the R.F., first -detector, and I.F. tubes is positive withrespect to ground, it follows that if thegrids of the respective tubes are con-

nected to a resistor in the plate circuitof the A.V.C. tube, that any potentialexisting across this resistor is added tothe original bias and makes the gridsmore negative than the original bias bythe amount of the voltage drop acrossthe resistor in the A.V.C. tube plate.

It is at once apparent that the great-er the signal voltage appearing at thegrid of the A.V.C. tube, the more platecurrent will flow in the plate circuit:an increase in plate current means anincrease in bias on the R.F., first -de-tector, and I.F. tubes; an increased biason these tubes means less amplificationand, therefore, less grid swing on thesecond -detector and A.V.C. tube. Thiscycle goes on until a constant voltageis obtained across the second -detectorinput, or, in other words, until a con-dition of equilibrium is reached.

Since R8 is located where the tonecontrol is normally connected, it wasnecessary to relocate the tone con-trol, C13 -R9 -C14. As less resistance isincluded between the two condensers,they become more effective in bypass-ing the higher audio frequencies; atthe same time, they resonate the prim-ary of T2 to a lower audio frequency.

A 1 -tube ConverterTo complete our gallery of types

we may as well show a so-called "con-verter"; that is, a device to be appliedto the input end of a normal broad-cast receiver so that the combinationwill then become a short-wave super -

The Strombger-Carlson model 33 automobile superheterodyne receiver which employs the6A7 and 6B7 tubes.

BLA

CK

?GN

D

PT

FU

SE

OP

ER

AT

ED

BY

TO

NE

CO

NT

RO

L

RE

D &

WH

ITE

!GR

EE

N'4

7

'51

1I

'51

RFI

L2i1

.1,.;

DE

T.1

YE

LLO

W

. SE

C C2

F

4

C9

12

Rs

BR

OW

N R

4ii

Bul

tR

ED

R5

I.F.1

.1'S

iI.F

.T.2

GR

EE

N'2

115

1O

ET

. 2

..110

4

R6

TE

LL01

1

g

Tll

RE

D

.15.

R7

RE

D

BLU

E

GR

EE

N 5

'W

HIT

E

YE

LLO

W, B

LAC

K T

RA

CE

R

C!,3

12!1

2C

I8

[717

.717

_77

7777

:09

YE

LLO

W

GR

EE

N)

RIO

C21

'27

OS

C

gc3.

1 I B

RO

WN

rrLS

R11

CH

WH

ITE

o4

.-R

ED

BLU

E I;

1=-.

C25

C26

C27

,

/- B

LAC

K /

(CO

MM

ON

)

RI5

RIG

BLU

E-%

GR

EE

NG

RE

EN

AN

D V

VH

ITE

R12

R13

C22

AV

C.

2.5V

.

'27 R14

XX

I-C

23

R.F

.C. T

C12 M

D -

BR

OW

N

BLU

E

mm

,C

27rB

LAC

K

*VIO

(3-1

)V

.I M

ET

ER

.

I.B

RO

WN

R

VO

LUM

EC

ON

TR

OL

'17

'47

PE

NT

-O

DE

--

our-

2.5

VP

taD

IAL

LIG

HT

RI

BLA

CK

, RE

D T

RA

CE

R

R22

R18

/OR

AN

GE

R19

,'BR

OW

N

E.Y

ELL

OW

GR

EE

NR

21

BL

AL

K-"

FIE

LDC

OIL

CO

IL

R9 C14

!VII

BLU

E

TO

NE

CO

NT

RO

L

WH

ITE

°oar

°V

OIC

EC

OIL

.

RE

D \

SP

EA

KE

R

RR

ED

AN

D B

LAC

K-%

R20

Fig

. 69.

The

How

ard

Mod

el 4

5 A

. V.C

. Sup

erhe

tero

dyne

.R

esis

tor

R15

is 4

50 o

hms;

R16

, 300

0;R

17, 3

,750

; R18

, 2,2

50; R

19, 4

50.

Con

dens

ers

C13

-C

14, 0

.1-m

f.; C

20, .

0009

-mf.

T2C

I3


heterodyne. This is admittedly a typeof apparatus which is historical ratherthan current; being now largely re-placed by outright short-wave or com-bination superheterodynes. Commer-cial development has accordingly tend-ed to stop in this field , and the latesttubes do not appear in most of theavailable types. For this reason, thecircuit shown here is not a commercialtype at all, but an adoption of suchtype to a later tube.

The user of a converter does not havecertain advantages which are availableto the factory designer of a completemulti -range receiver. Therefore, it issomewhat advisable for him to attacha converter to a receiver which is un-likely to complicate the operation byintroducing whistles. This is anotherway of saying that the converter willordinarily live more peacefully witha T.R.F. receiver than with a super-heterodyne, since in the latter case weshall have two oscillators (the one inthe converter and the one in the re-ceiver) with a virtual certainty thatthey or their harmonics will engagein controversies. Of course T.R.F. re-ceivers are gradually being pushed in-to the background because of theirlesser performance on a dollar basis, sothat the converter and the T.R.F. setmay disappear into the past at aboutthe same time. Meanwhile, one of themmay be made very simply into a quiteacceptable shin -wave superheterodyneby attaching to it such a converter ashere shown (Fig. 70).

No power supply has been shown,since the drain is small and may usual-ly be "borrowed" from the power sup-ply of the receiver with which the con-verter is used. Should this supply beat other voltages than those listed,some changes may be necessary. Ifthe plate voltage is below 180 V. os-cillation may be uncertain. If so, de-crease the cathode R (150 ohms at 150volts) and also try increasing turnsin L7 and L8. Voltages below 150 areunsatisfactory, and 250 V. are recom-mended. At this voltage the screen -grid may be fed from the plate supplythrough a simple voltage divider;ground the screen -grid through a50,000 ohm resistor and connect to"plus 250" through 35,000 ohms.

The circuit can be made from stand-ard parts, of which only two need mod-ification. The two -gang tuning con-denser must have trimmers, and itscapacity per section should be slightlyabove 200 mmf. Use a broadcast -range 2 -gang condenser and removeabout 1/3 of the plates from either therotors of both sections or the statorsof both sections. It is needless to takeout both rotor and stator plates.

Since double -tuned 900 kc. transform-ers are not usually available, obtain a175 kc. (or other frequency below 900kc.) type which has double tuning andmodify it as follows. First set up thebroadcast receiver which is to be usedand tune in a station at 900 kc., or notover a channel or so on either side. Ifa good station isn't available, set upsome sort of a simple oscillator withA.C. plate supply and tune to that-but set it at 900 kc. first, by compari-son. Next, connect up the convertercomplete except that L7 and L8 aretemporarily short-circuited to preventoscillation and the connection is takenoff the top (control -grid) cap of thetube. Now connect this cap to a shortantenna, and also ground it to the chas-sis through a resistor-value not im-portant, 'but somewhere between .1 -and 2. megohms. We now have an R.F.stage with an untuned input, coupledto the receiver through our I.F. trans-former. Probably neither our 900 kc.station nor our 900 kc. oscillator canbe heard since the I.F. transformer isstill away off tune. Now by slow stages

:2* ' 41.7,..t.:1°A=411,100 \

AWE

TR1

200 WAR

Ct S OR ON MIASRAM. GOOD100. -RATIO

DIAL 1133.1.1.DAL

IF (OAR

CA11. SYR 2

N' GC OR<ME SMARTOR SEPARATE

FIG.70

This short-wave converter is a modernversion of a type that was very popularsome time ago. It uses an I.F. of 900 kc.


remove turns from both windings ofthe I.F. transformer until it is possibleto tune in the 900 kc. signal (stationor our temporary A.C.-plate oscillator)with the I.F. trimmer condensers setmost of the way out. While the rangeof the condensers in I.F. transformersvaries, it will always be necessary toremove at least 1/2 of the turns to getto 900 kc.-usually much more. Insome receivers best response is had byremoving the original antenna coup-ling transformer and connecting thesecondary of the I.F. transformer be-tween the control -grid and chassis.

Finally connect everything per the dia-gram and proceed with short-wave re-ception.

In all of these maneuvers the receiveritself must stay at 900 kc.

The CoilsThough both on 1 in. forms, the os-

cillator and tickler coils must not beon the same piece of tubing. Placethem to minimize coupling; let thepentagrid tube attend to the coupling-but keep the leads short.

The windings are as follows; all closesingle layers.

Range in meters

Switch on 116 to 38

Switch on 235 to 90

Switch on 384 to 212

Input coils used(spaced 14 -in. apart ontube)

oscillator coils used(spaced 1/4 -in. apart ontube)

L1-51/2 T. No. 28 D.S.C. L4-6 T. No. 28 D.S.C.

above plus L2which is13 T. No. 28 Enam.

above plus L5which is11 T. No. 28 Enam.

both above plus L3which is34 T. No. 28 Enam.

both above plus L6which is27 T. No. 28 Enam.

CHAPTER 9

I. F. Transformer DesignHAVING determined the minimum

R.F. voltages which must be sup-plied to the grid of the detector to de-liver the maximum power output as ex-plained in Chapter 7, we are in a posi-tion to determine the total gain whichmust be obtained from the I.F. ampli-fier. Figures 71 to 74 show a typicalcircuit and operating conditions.

Modern radio receivers of the super-heterodyne type have an input sensi-tivity of less than 5 microvolts permeter and, with the standard heightof the antenna set at 4 meters, we findthat the absolute sensitivity will beabout 20 microvolts (a microvolt beingone -millionth of a volt). Thus, if wedesire a receiver (as shown in Fig. 71)that will deliver about 4 volts of R.F.signal to the detector from an inputsignal of 20 microvolts, the total volt-age gain of the amplifier will be

R.M.S. volts on grid of detector

R.M.S. volts input from antenna4 volts

or = 200,000 gain..000020 volt

As a certain amount of amplificationcan be, and is, obtained by one or morestages of conventional T.R.F. ahead ofthe modulator tube (first -detector), itis not absolutely necessary that the en-tire burden of amplification be borne bythe I.F. amplifier. If there are twostages of T.R.F. ahead of the modula-tor, then there will be a voltage gain

Elementary circuit of an I.F. amplifier.

of about 1500 (assuming a gain ofabout 40 V. per stage) which must beconsidered in designing the I.F. ampli-fier.

Now, let us see just what the finalfigures will be with the added gain ob-tained in the pre -amplifier.

If the input to the receiver is .00002-V. and the pre -amplifier has a gain of1500, then the input to the first I.F.transformer will be .00002 x 1500 or.03-V. The 4 volts required by the de-tector, divided by the .03-V. input tothe I.F. amplifier, will then be thevoltage gain required by the I.F. am-plifier, which is 133.3 V.

As it will be impossible to obtain again of 133.3 in a single intermediatestage, it will be necessary to use twostages working at a gain of about 65,or three working at 44 per stage.

In the example cited above, the am-plification due to the modulating tubeis ignored, as various conditions devel-op which cause the gain of this portionof the circuit to vary over wide ranges.The sensitivity and output will be af-fected by the strength of the receivedsignal, by the power output of the localoscillator, and by any change in oper-ating potentials which may take placeas the receiver is functioning.

The check for the correctness of thecalculation can be made by multiplyingthe gain in the pre -amplifier by thegain in the I.F. amplifier; thus, 1,500 x133.3 gives a value of 199,950.

Figure 75 shows a skeleton circuitwith the voltages developed in the var-ious circuits. Two stages of I.F. am-plification are shown and, as each stageis not working at the maximum pos-sible gain, the I.F. amplifier will bevery stable and the coils easy to de-sign.

If an actual condition exists wherethe gains and voltages are measuredand found to be as indicated in Fig. 75,the volume control on the pre -amplifier

56


end of the receiver will be full on andthe gain on the I.F. amplifier cut awaydown.

If the pre -modulator amplifier is lim-ited to one stage, it will be necessaryto increase the gain of the I.F. ampli-fier if the same level of sensitivity isto be maintained.

Unlike the conditions which exist inT.R.F. amplifiers (where the limitationsof the minimum and maximum capacityrange of the tuning condenser, plus theunavoidable circuit capacities, define themaximum ratio of the tuning inductanceto its tuning capacity), we find that thetuning circuits of I.F. amplifiers arenot limited as stated above, and theratio of L to C can be any ratio de-sired, within sensible limits. The aboveinformation is a partial repetition ofthe data in Chapter 7, but is necessaryto illustrate the use of the attached in-ductance chart.

Inductance Design

Thus, the inductance of the I.F. trans-

former can be made as large as de-sired; the limitations being defined bythe R.F. resistance and the physical sizeof the coil and associated shield. Asthe frequency of the I.F. amplifier isgenerally lower than the broadcast -band frequencies, the effect of the cir-cuit and coil capacities can be neglectedfor the moment as any calculation whichwe shall make will generally assumethat the signal is fed into the tunedcircuit by induction in the coil itself.In Fig. 76A, we find that the distributedcapacity of the coil shunts the tuningcondenser and is simply added to thecircuit; in Fig. 76B, the signal is inseries with the coil.

Calculation of Load Impedance

To obtain the greatest percentage ofthe "mu" of a vacuum tube, it is neces-sary that the load in the plate circuitbe as large as possible.

The effective impedance of the tunedcircuit at resonance (Fig. 77) is equalto the following mathematical formula:IfX START OF GRID

CURRENT18

.2X v. START OF GRID

CURRENT

viI 32 TUBEScio I

gi 16

r 14 X

ot

28Eb . 174 V.

X o12

c ,,2....

ECG - -6 V.ESG4 67.9V.

B>i 24

1-

43IV 24 TUBES

io Vs 27 TUBES 2°'16

g CURVE-A CURVE -BEb

8i.:. 6

Eb Ecc, 01- 12 12i30

20A 180Y. EL180y.

Elio. 45V. (so.. 75V. ...z 4

INV CURVE A 45Y. -5V.A CURVE B. 90V. -9 V. 8. a rA.

10 Phi PLATE

Eco-4.5V. Ec4,-7.5Y.LOAD MEG.

5 NEVAFEN A

CURVE C180V. -20V.PLATE LOAD

ua wmWAIMg

FE 0

..no

.2- g 2.....0&III

.2-MEG. 4.....

pra FIG.74cz 0 2 3 4

Ramp FREQUENCY INPUT-a FIG.72 VOLTS R.M.S.

0 0 2 4 6 8 10 12k1. RADIO FREQUENCY INPUT-'C FIG.73 VOLTS R.M.S.

0la 0 2 3 4D . RADIO FREQUENCY.4 INPUT VOLTS R.M.S.

FIG. 72-left. Curves showing the possible gain that may be expected from a 24 tube.FIG. 73-center. The same curves as in Fig. 72, only for a 27.

FIG. 74-right. Same as for Figs 72 and 73 but for a 32.

aF I PAZ40 a. 40

E2 .0008Y. E3

oat* FIG 75 1AI I.F.2 DET. Za. 114 .. II+ AUDIO OUTPUT

54 0 032.V. ES 0.352.V. E64.0V.

An elementary circuit illustrating how a signal is increased through an amplifier.


.. r 4,2....4..IS%

" E bl..Ourt0 WE TVII 11- X0 SCAAL

"MINN

. PrAll'LI

. NE IliNr.; ''' " '. t' .4° R40.24.. [,, ]7,FIG. 76. Upper right, effects of distri-buted capacity; FIG. 77-lower left coilresistance; FIG. 78-coil Q computation.

L2 W2Z - - Crr

where L =the inductance of the coil,W=6.28 times the frequency f,r =the series high resistance of

the coil,C=the capacity necessary for

resonance.

It will be noted that the effective im-pedance increases as the square of theinductance; so, provided we keep theR.F. resistance of the coil low, a largeinductance will be superior to a smallone.

In such a tuned circuit, the selectivityS will be proportional to

WLS

rand the width of the resonance curve,Fig. 78, at a point where the responseis .707 times the value at resonance,is related to the ratio

WL frS - - f2 figiving another valid reason for usinga coil as large as possible. A handyrule to use in the design of such cir-cuits is that

WL

r

12 -

so9-e=7-6 -5-4-

3-

2-

11 -

1 -

ii'_

nie-

.

33

W.2

25E

E 30

z

W 40-

50-O

60770

40 -90 -100-

150 :

2

0

I

INDEX LINE

CHART ix

I

22

0

0

2

-i Ia

z

- 20

3

4

5 26

7

10

IS

20

50

40-SO- 60

ao=100

-Ise-- 200.,,

-250_300

- 400- 500 1:4

600 4

7 8001000

1500

2000

3000

4000

A- 30

40-50- 60-- 100

- 150

- ZOO

- .300400

- 500--600

800 ", 1000

1500

2000

3000

40005000

5a

3

41

C.- .00001

-.00002

- .00003

.00005

.00006

r -.0000e=. coot

.0002

- .0001- .0004

.0005

0- .001

.002

FIG.79 CHART I

An automatic coil -condenser calculator. Knowing the value of either a coil or a tuningcondenser, the other may be determined, for any wavelength by reference to the chart.

/0 1.44 *0 01 ° ° 3 &44

0 0 t:1 SO'c

CHAPTER 10

Servicing SuperheterodynesTROUBLE shooting and, servicing a

superheterodyne requires the samesystematic procedure that most read-ers of this book are already familiarwith in connection with T.R.F. receiv-ers. When we realize that the signalfrequency amplifier, the audio amplifierand the power supply in a superhetero-dyne are virtually the same as the cor-responding units of a T.R.F. receiver,it is obvious that the same troubles arelikely to occur in each and the correct-ing of these troubles in each case willbe the same.

Since the superheterodyne has, in ad-dition to the above units, an oscilla-tor, first and second detectors and anintermediate amplifier, and the hetero-dyne method of reception being some-what more complex than the ordinary,additional problems present themselves.When a superheterodyne fails to workor works improperly there are moreplaces where trouble is likely to occur;and the method of finding them mustbe conducted in a systematic manner.We will limit this chapter mainly todiscussing problems inherent only tothe superheterodyne and show how tocorrect them without going into de-tailed instructions for trouble shootingand servicing the power supply, audioand T.R.F. amplifiers.

To properly service a commercial re-ceiver it is advisable, if possible, tothoroughly study the service bulletinspublished by the manufacturer of thereceiver. These will give detailed in-formation about the particular set --which obviously cannot be included ina general treatise like this on the sub-ject. But with a thorough generalknowledge of superheterodynes backedwith a fair amount of practical experi-ence, a service -man can usually tracethe source of trouble and correct it invery short order without the use ofparticular instructions. However, in

59

any event, well calibrated instrumentsare essential. While home-made test-ing instruments may be suitable, werecommend that the serviceman investin good manufactured equipment.

Equipment Required

A modern up-to-date analyzer is ab-solutely essential for accuracy andspeed in servicing a receiver. Thisshould be designed to accommodate thelatest tubes. With its use a quickcheck of the voltages at the varioussockets can be made, as well as platecurrent measurements, etc.; the tubescan be tested and in many cases thesource of the trouble will immediatelybe localized.

Next in importance in superhetero-dyne servicing is an accurately cali-brated R.F. and I.F. oscillator. Thisshould generate modulated oscillationsthroughout the entire broadcast bandof 1,500 to 550 kc. and throughout theintermediate band of 550 to 125 kc.Some commercial oscillators are de-signed to generate a fundamental fre-quency covering the intermediate bandand the harmonics of this fundamentalare used for covering the broadcastband.

An output meter and an ohmmeterare also essential equipment. Thesemay be part of the set analyzer. Inaddition the customary service tools formaking rapid repairs are necessary.

Complete instructions for using theabove equipment are furnished by themakers. For this reason we will notgo into a detailed analysis of their in-ternal design and methods of use.

Harmonics

Unless an alternating current has aperfectly pure sine wave, harmonics ofhigher frequencies will be present.


Whenever vacuum tubes are used in acircuit the true sine wave is distortedbecause even the straightest portion ofa tube's characteristic curve is slightlybent. For this reason we have manyharmonics in a superheterodyne re-ceiver with which we have to contend.

Harmonics are always multiples ofthe fundamental frequency. The sec-ond harmonic has a frequency of doublethe fundamental; the third harmonic isthree times the fundamental; thefourth, four times, etc. The most pow-erful harmonics are the ones nearest tothe fundamental, such as the / secondand third, although harmonics up be-yond the fifth may cause interferencein a superheterodyne. For example,the fourth harmonic of 175 kc., is fourtimes 175 or 700 kc. This harmonicand all those up to the eighth lie inthe broadcast spectrum and may causetrouble due to heterodyning with othersignal frequencies if the shielding ofthe receiver is defective.

As an example, suppose we are tunedto a broadcast station having a fre-quency of 690 kc. With an I.F. ampli-fier operating at 175 kc., the oscillatorwill be tuned to 865 kc. If for somereason the fifth harmonic of the I.F.frequency, which is 875 kc., gets intothe same circuit with the broadcastcarrier (say the second detector tubecircuit), a 10,000 cycle beat frequencywill be generated. Unless the audioamplifier has a cut-off below 10,000 cy-cles, this current will either produce awhistle or squeal in the loudspeakeror cause serious background noise. Ifthe I.F. amplifier is slightly out of ad-justment, so that the intermediate fre-quency is slightly below 175 kc. due toimproper adjustment, the trouble maybe worse. For example, if the fre-quency is 174 kc., our oscillator will betuned to 864 kc. which may mix withthe fifth harmonic of 174, which is 870kc. and produce a 6,000 cycle hetero-dyne whistle which would be quite ser-ious in the audio amplifier. For thisreason it is important that the inter-mediates be properly adjusted to 175kc. Furthermore, if the oscillator andtuner circuits are accurately aligned totrack 175 kc. apart, the I.F. amplifiermust be adjusted to this frequency orthe sensitivity of the set will be great-ly reduced.

Aligning the Circuits

For a thorough check of the tunedcircuits in a superheterodyne, an ac-curately calibrated modulated oscillatorshould be used. The pre -selector cir-cuits in the signal frequency amplifiershould be perfectly aligned and bymeans of the calibrated oscillator acurve such as is shown in Fig. 31, A,can be made, if desired. To do thisthe output meter or simply a milliam-meter connected in the plate circuit ofthe first detector should be used to in-dicate maximum sensitivity. The os-cillator should be coupled to the inputof the signal frequency amplifier; ifshielded coils are used a couple of turnsof wire inside of the shield can bebrought out to a similar coil placednear the oscillator. The oscillatorshould be placed in operation and thetrimmer condensers on the pre -selectorcircuits adjusted for maximum sensi-tivity.

The above procedure should then berepeated with the oscillator circuit. Thedial should be set to 1,500 kc. and theoscillator tuned to 1,675 kc. by meansof the high frequency trimmer conden-ser. Successive readings can then bemade and the oscillator curve plottedas shown at B, Fig. 31. At the 500 kc.dial setting the oscillator should be ad-justed by means of the low frequencytrimmer to 625 kc. If the curve of theoscillator deviates from the theoreticalcurve it can be corrected by bending theslotted plates usually provided on vari-able condensers.

The next step is to acurately adjustthe tuning condensers of each I.F.transformer so that the I.P. amplifieroperates exactly at a peak of 175 kc.To do this the modulated oscillatorshould be set exactly at 175 kc. and theoutput loosely coupled to the grid ofthe first detector. With an outputmeter to indicate resonance, the tuningcondensers of each transformer shouldbe carefully adjusted until maximumresonance is indicated on the outputmeter.

Some receivers employ a visual tun-ing indicator. With such a receiver anoutput meter is not needed as the pointof maximum resonance will be indicatedby the visual tuning meter.


While the above procedure assumesan intermediate frequency of 175 kc.,it should be definitely understood thatthe same process of aligning the cir-cuits may be followed on sets having adifferent intermediate frequency.

If these instructions have been car-ried out properly, the set should nowbe in perfect alignment at all pointson the dial, and further changes in thetrimmers at any point on the dialshould not be necessary.

Un-uniform Sensitivity

Poor sensitivity on one end of theband, as compared to the other end, oron both ends as compared to the mid-dle, is almost invariably a sign of im-proper tracking, and can be correctedby making the adjustments already de-scribed. Lack of sensitivity all overthe band, provided all other things arecorrect, is usually an indication that theintermediate transformers are nottuned accurately. As already stated,the adjustment of the intermediates toexactly 175 kc. is of extreme import-ance.

Whistles and Squeals

"Birdies"-sounds like a regenerativereceiver passing stations at variouspoints on the band-are caused eitherby the intermediates being tuned tosome frequency other than 175 kc., orby insufficient selectivity in the R.F.tuning circuits. An easy way to findwhich is the cause is to short the os-cillator tuning condenser, and then ro-tate the dial with the volume controlturned well up. Under these conditions,no stations should be heard, in fact thereceiver should be absolutely silent. if

stations are heard at some points,without the oscillator tube operating,it is a certainty that the intermediatesare not tuned properly. If the set issilent without the oscillator working,but whistling "birdies" are heardwhen it is working, the selectivity ofthe R.F. amplifier is insufficient. Thesimplest way of correcting this is touse a much shorter antenna, or to re-move turns from the primary of theantenna coil. A very small condenser,of the order of 500 mmf., (a midgetvariable will do) inserted in the anten-

na lead, will very often eliminate thewhistles without appreciably cuttingdown the sensitivity of the set.

Repeat Points on Dial

Occasionally, on some supers, therewill be found repeat points abiut 350kc. off the proper place for a station.There are two remedies for this-eitherthose already described for "birdies"(which will usually be found on setshaving the repeat points) or by improv-ing the shielding of the set from directpickup; as, for example, mounting a setwhich has the chassis unshielded onthe bottom, on a metal plate, so thatthe bottom will be shielded. Coveringthe top of the chassis with a groundedmetal plate, so as to shield the variablecondenser sections and grid caps is of-ten very helpful.

Microphonic Howls

Microphonic audio howls will befound troublesome in some imperfectsupers, and the builder, naturally at-tributing it to a bad tube, will hunt invain for the tube that is causing thetrouble. Actually, the howl may becaused by vibration in the plates of thevariable condensers. It can usually becured by mounting the entire chassis ona piece of sponge rubber, allowing theentire chassis to vibrate, instead of justthe condenser plates.

Poor Selectivity

Some sets will have ample selectivityso far as music is concerned, but on astation next to a powerful local, theloud notes of the local will "carry over"with a kind of scratching blast. Thisis a sign that the local is modulating aband more than 10 kc. wide, and inas-much as the trouble originates in theair, it cannot be completely eliminated.It can, however, be considerably ameli-orated by the addition of a band-passstage ahead of the tuner. This will re-duce the amount of signal from thelocal that reaches the grid of the firstR.F. amplifier tube, but will not ser-iously affect the strength of the signalfrom the station to which the set istuned.


Dead Spots

Some sets will be found which workvery nicely over a portion of the band,usually the high frequency end, butwhich stop working entirely on otherportions. This is caused by the oscilla-tor tube having incorrect voltages, sothat it stops oscillating in spots. Acheckup of the voltages supplied to theoscillator tube, and the correction ofthese (if incorrect) will usually fix thetrouble. Sets using dynatron oscilla-tors are particularly subject to thistrouble. In this case, trying out sev-eral tubes will result in one being foundwhich will work properly over the wholeband. Many 24 tubes will not oscillateat all as dynatrons, although they willfunction perfectly as detectors; and al-most all tubes, so used, require veryaccurate settings of the screen andplate voltages to oscillate over the en-tire band.

Poor Quality

Occasionally, a set will be foundwhich has perfect quality on full vol-ume, but when reduced, the quality"goes to pieces." If this is the case,examination of the tubes will probablydisclose a 24 in a socket where a 35 or51 should be. Proper placement of thetubes will make this right. This troubleapplies to T.R.F. sets only: the use ofa 24 in an amplifier socket in a setbuilt for the multi -mu tubes will in-variably produce this phenomenom.

No reference has been made here toaccount for poor results due to im-proper connections, wrongly placedparts, or similar troubles which wouldapply to any receiver. It is presumedthat the correct hookup has been fol-lowed throughout, and the receiver isfree from all defects in wiring, parts,or similar mistakes on the builder'spart.


Here you will find other titleswhich are included in the Seriesof the book you are now reading

Presented on this page are the new books of the RADIO -CRAFT' LIBRARY-the most com.plete and authentic set of volumes treating individually, important divisions of radio. Each bookhas been designed to give radio men the opportunity to specialize in one or more of the popularbranches of the industry. The material contained in these books will increase your knowledge:you will find them a real help in your work and they will contribute to your money earningcapacity. Read these books during your spare time at home.The authors of these books are well-known to everybody. Each one is an expert radio man:an authority on the subject-each is thoroughly familiar with the field which he represents.This is perhaps the first real opportunity that you have ever had to build a radio library ofbooks that are authentic, right -up-to-the-minute and written so that they are easily digested.

Book No. I

RADIO SET ANALYZERSAnd How To Use Them

With Full Instructions and De-seriptIons of Set Analyzers, Tube

Checkers, Oscillators, Etc.By L. VAN DER MEL

Book No. 2MODERN VACUUM TUBES

And How They WorkWith Complete Technical Data en

All Standard and Many SpecialTubes

By ROBERT HERTZBERG

Book No. STHE SUPERHETERODYNE BOOK

All About SuperheterodynesHow They Work. How to Build and

How to Service ThemBy CLYDE FITCH

Book No. 4MODERN RADIO HOOK-UPS

The Best Radio CircuitsA Complete Compendium of theMost Important Experimental and

Custom-built ReceiversBy R. D. WASHBURNE

Book No. 5

HOW TO BECOME A RADIOSERVICE MAN

How To Get Started and How ToMake Money In Radio Servicing

By LOUIS MARTIN

Book No. 8BRINGING ELECTRIC SETS

UP TO DATEWith Pentodes, Multi-Mus, Dy-namic Speakers-Complete Inform-ation How to Modernize A.C..

D.C. and Battery OperatedReceivers

By CLIFFORD E. DENTONBook No. 7

RADIO KINKS AND WRINKLESFor Service Men and

ExperimentersA Complete Compendium en the

Latest Radio Short -Cuts andMoney -Savers

By C. W. PALMERBook No. 8

RADIO QUESTIONS ANDANSWERS

A Selection of the Most Importantof 5.000 Questions Submitted by

Radio Men During the Courseof One Year

By R. D. WASHBURNE

Book No. 9

AUTOMOBILE RADIO ANDSERVICING

A Complete Treatise on the SubjectCovering All Phases from Installing

to Servicing and Maintenance

By LOUIS MARTIN

Book No. 10

HOME RECORDING AND ALLABOUT IT

A Complete Treatise on Instan-taneous Recordings. Microphones.Recorders. Amplifiers, Commerelal

Machines. Servicing, etc.By GEORGE J. SALIBA

Book No. IIPOINT-TO-POINT RESISTANCE

MEASUREMENTSThe Modern Method of Servicing

Radio ReceiversBy CLIFFORD DENTON

Book No. 12

PUBLIC ADDRESS INSTALLA-TION AND SERVICE

Modern Methods of Servicing andInstalling Public Address Equipment

By J. T. BERNSLEY

ALL BOOKS UNIFORMThe books in the new RADIO -CRAFT LIBRARY are all strictly up-to-date and written by men who know theirsubjects. The volumes are all uniform size, 6 x 9 inches, and contain on an average from 50 to 120 illustrations.Each book is printed on fine book paper.

Big Discount OfferedIn order to make it possible for everyone to buy these books, the fifty (50) cents price hasbeen made uniform for all volumes. You can buy these books separately, but you should takeadvantage of our special offer:

WHEN FIVE (5) BOOKS OR MORE ARE ORDERED DEDUCT 20%FROM YOUR REMITTANCE

Simply fill in the coupon below, and mail it to us together with your remittance. Checks, stampsor money orders accepted.

Clip Coupon and Mail TODAY

GERNSBACH PUBLICATIONS, INC. 96-98 Park Place, New York, N. Y.I have circled below the numbers of books in the RADIO -CRAFT LIBRARY, which you

are to send me, and have deducted 20% for ordering five (5) books or more. I haveincluded my remittance in full, at the price of 50c each, when less than five books areordered.The amount of my remittance is (Stamps, checks or money orders accepted.)Circle numbers wanted: 1 2 3 4 5 6 7 8 9 10 11 12

Name AddressCity State SII


'win trainyoa athome/iliaa

106.°1Ady

RadiojobJ. E. Smith Preeident. Na-tional RadioInstitute, the Marlwho has directed the HomoStudy Training of more menfor the Radio induotry thanany other man in America.

Get my big FREE book on the opportunities inRadio. Read how quickly you can learn at homein your spare time to be a Radio Expert-whatgood jobs my graduates have been getting-realjobs with real futures.

Many Radio Experts Make $50 to $100 a WeekOver 300,000 jobs have been created by Radio'sgrowth, and thousands more will be created by itscontinued development. Many men and young menwith the right training-the kind of training Igive you in the N. R. I. course-have stepped intoRadio at two and three times their former salaries.

Many N.R.I. Men Have made $200 to S1000in spare time while learning

The day you eillon wan me 1 send you material whichyou should master quickly for doing 28 jobs, commonin most every neighborhood, for sparetime money. I giveyou the plans and ideas that have made $200 to $1,000for N. R. I. students in their spare time while study-ing. My course is famous as the course that paysfor itself. h

Aircraft Radio Included and tripled thesalaries of many.Find out aboutthis tested wayto BIGGER

Talking Movies, Television,

Special training in Talking Movies Televisionand home Television experiments, Radio's useits Aviation, Servicing and Merchandising Sets,Broadcasting, Commercial and Ship Stations areincluded. I am so sure that N. It. I. can trainyou satisfactorily that I will agree in writingto refund every penny of your tuition if you -are not satisfied with my Lessons and Instruc-tion Service upon completion.

64 -page Book of InformationFREE

Get your copy today. It tells you whereRadio's good jobs are, what they pay,tells you about my course, what otherswho have taken it are doing and making.Find out what Radio offers you withoutthe slightest obligation. ACT NOW!

.. NEW Radio Equipmentfor Broad Practical ExperienceGiven Without Extra ChargeWith this equipment you work out many of thethings you read in textbooks. From it you getthe experience that tells an expert from a be-ginner. In a short time you have learned whatwould take years to learn in the field. Youmake experiments illustrating the importantprinciples in the 25 best known receiving sets.

ave doubled

P.

J. E. SMITH, PresidentNational Radio Institute, Dept. 2ED7Washington, D. C.

Dear Mr. Smith: Send me your freebook. I understand this request does notobligate me and that no salesman will call.

NAME

ADDRESS

CITY STATF

111111 lIZINIVIZE411111:11R1)Si?

A GOLDEN OPPORTUNITYFOR ALERT RADIO MEN IN

THE NEXT GREAT INDUSTRY

Official

T HE idea of electricians, radio service men and otherinechanieally inclined men, servicing Air Condition-

ing and Refrigeration Units is self-evident and thethought has occurred to some untold thousands eversince air conditioning equipment has been installed inpublic auditoriums, theatres, studios, department store..office bulidi"gs and manufacturing plants. The tre-mendously broad possibilities in this new industry arehound to give employment and success to men tar -sightedenough to see its advancement and development. Wequote an excerpt from Mr. Hugo Gernsbark's editorialwhich recently appeared in Everyday Science and Mech-anics magazine.

"I advise young and progressive men to go

into the air-conditioning business during thenext few years; because, this, without a doubt.is the coming industry In this country. Thous-ands of small firms will spring up. undertakingto air-condition private houses. small businessoffices. factories. etc. We are not going to

tear down every building In the United Statesimmediately. It will be a gradual growth; yetsmall Installation firms will air-condition smallhouses, and even single offices In -small buildlogs."

This le only partial proof of the . ertein success ofthis new field. Further assurance is that engineeringschools have already added many important courses onsir conditioning to their regular curriculum. Architectsand building contractors are giving considerable thoughtto installation of this equipment in structures whichare now being planned and built. Tie beginning ofthis bus:Inc a will probably be similar to the autoand radio industries, but in a few short years It will-urpass these two great fields.

Air Conditioning Service Manual

352 Pages

Over 600

Illustrations

9x12 Inches

Looseleat

Leatherette Cover

$5.00 List

The OCI- is i V. Alit taiNli IT loN 1St; sElt V ICE MANI Al. is being edited lir

L. K. Wri ,.iii.who is an expert and a leading authority on air conditioning andrefrigeration. I le is a member of the American Society of Refrigerating En-gineers, Amer lean Society of Mechanical Engineers, National Association ofPractical Refrigerating Engineers; also author of the OFFICIAL REFRIGERATIONSERVICE MANUAL and other volumes.

- In this Air Con 'Luning Service Manual nearly every page is illustrated;every modern installation and individual part carefully explai-ed; diagrams fur-nished of all known equipment; special care given to the servicing and installationend. The tools needed are Illu [rated and explained; there are plenty of

charts and page after page of servhe data.Remember there is a big opportunity in this new field and plenty of money

to he ma e In the servicing end. There are thousands of firms selling installationsand parts every day and this equipment must be eared for frequently. Eventalallyair co ditioning systems will be as common as radio: and refrigerators in heroes,

offices and industrial plants. Why not start now-increase your earnings With a.full- or spare -time service business.

Here are sense of the chapter heads of the OFFICIAL AIR CONDITIONINGSERVICE MANUAL: CONTENTS IN BRIEF

History of Air Conditioning; Fundamental Laws; Methods of Refrigeration;Ejector System of Refrigs ration: Compression System of Refrigeration; Refrig-erants; Lubricating Oils; Liquid Throttle Devices: Servicing Expansion and FloaiValves; Servicing Refrigerating Systems; Control Devices: Thermodynamics of Air

Conditioning; Weather in the United States; The Field of Air Co afitioning;Insulating Material:; Heat Transmission Through Walls; Complete Air Condition-ing Systems; Estimating Requirements for the Home, &nail Store, Restaurant;Layout of Duet Systems; Starting Up a System: Operating and Servicing AirConditioning System: Air Filtration, Ventilating and Noise Eliminating DevicesPortable Electric Humidifiers and Room Coolers; Automatic Humidifiers, AirConditioning Units for Radiator System and Warm Air Systems; Central Con-ditioning Units, etc.

Send remittance of $5.00 in form of check or money orderfor your copy of the OFFICIAL AIR CONDITIONING SER-VICE MANUAL. Register letter if it contains cash or

-nrrenry. THE MANUAL IS SENT TO YOU POSTAGEPREPAID.

GERNSBACK PUBLICATIONS, Inc,96-9SSH PARK PLACENEW YORK, N. Y.

$350Reg. List Price

OfficialRadio Service

Manual.

bomplete Directori

i93.3.1934 Radiolteceivers

Full Radio Service Guide

400 Pages

Over 2,000 Illustrations

9 x 12 Inches

Flexible, LooseleafLeatherette Cover

There's plenty of Servicing Material

in the NEW 1934 ManualTHE necessity of GERNSBACK Manuals in

the radio field has been shown by the factthat the total sales of the first three OF-

TICIAL RADIO SERVICE MANUALS, in-the new CONSOLIDATED EDITION,

now exceed 80,000 copies. Radio Service Menand others engaged in various blanches ofradio know the importance of such gooks, andbow they must depend upon them for reliableinformation. Whether for public-adthess work,tube information or a circuit diagram, thematerial needed is certain to be found in oned the OFFICIAL RADIO SERVICE MAN-UALS.

In preparing this new edition many of theoutstanding problems of the Service Men havebeen considered --methods of servicing, the newequipment constantly needed to cope with newtibes and sets, and the other fields of radio,rich as public-address systems, shcrt waves,aAto radio and others.

The illustrations in the 1934 Manual aremore explicit than before; inasmuch as thed agrams are not limited to the schematic cir-cuit, but other illustrations show ;he partslayout, positions of trimme:s, neutralizers, etc.There are hundreds of new circuits included,aid not one from any previous editicns of then-anuals has been repeated. This we uneon-d-tionally guarantee.

It is quite evident that the 1934 Edition ofthe OFFICIAL RADIO SERVICE-. MAN UALig a decided improvement over previous vol-umes. The new book will p...ove itself to beit -valuable as those volumes for previous yea: s.

Contents of the 1934 Manualin Brief

A Diagrams and sei .ice notes, more completethan ever before in ar.y MANUAL. Net mei eiythe schematic hook-ups will be found, but chas-sis drawings showing parts layouts, positions oftrimmere, neutralizers, etc.

a 'Voltage readings for practically all sets,as an aid In checking tubes and wiring.

All values of intermediate -frequency trans-formers used in superheterodynes. with the man-ufacturers' own suggestions as to correct bal-ancing.IP Detailed trouble-shoofirg suggestions andProcedure as outlined by the manufacturers' owne gineers-in other words, authentic "dope"right from headquarters.

Values of all parts indicated directly onall diagrams. A.G.- D.C. el garbox midgets. Publk-address amplifiers.A Short-wave re. rivers.A 13( mote -control systems. A complete compilation of radio tube data.covering Loth the old and the many new types.A Section devoted to test equipment, analy-zers. etc.. with full diagrams and other valu-able information.ON A complete list of American broadcast sta-tions with their frequencies In kilocycles; ex.(rem ly useful In calibrating and checking testoscillators and In calibrating receivers. Free Question and Answer Service, thesame as In our last two Manuals.A No theory; only service information inquickly accessible form.A Absolutely no duplication of any diagrams;nothing igat appeared in any of the previousManuals will appear In the 1934 MANUAL.This we unconditionally guarantee.A A handy. easily -consulted master indexmaking it easy for you to Rad almo.t anythingPertaining to your service problem instantly.This index Includes all the diagrams publishedin all the previous C ERNSBAC K Manuals, aswelt as the 1934 diagrams. A big convenienceand time saver!

Send remittance of $3.50 In form of check or money order for your copy of the[1934 OFFICIAL RA D 10 SERVICE MANUAL. Register letter if it contains cashor currency. THE MANUAL IS SE NT TO YOU POSTAGE f' R EPA ID.

GERNSBACK PUBLICATIONS, Inc. 96-98814 PARK PLACENEW YORK, N. Y.

Documents

SUPERHETERODYNE BOOK...Chevrolet Motor Company Crowley Radio Corp. Delco Air lichee Corp. Emerson Electric Mfg. Co. Federated Purchaser, Inc. Foda Radio & Elec. Corp. Ford -Majestic