IHNishers' Photo Service, ii. SEPT. · 2019-07-17 · Every neighborhood can use a good part time...

SEPT.1935

IHNishers' Photo Service, ii. Y.

Fie,d units carry on communication by radio, for50 miles or so, using a hand operated generatoras voltage supply, and a snort wave receiver andtransmitter. Aerial can be seen at rear guy pole.Cadets are giving the demonstration at West Point.

DUBILIERCONDENSERS

Are you building your set with parts that will stand up-or are you going to be continuously repairing your set?

OVERONEONE BILLION

CONDENSERS SOLDSINCE 1910

Condensers specified by the UnitedStates Navy must be good. Uncle Samdoesn't use parts that go blooey at themost unexpected times. That's whyC -D DYKANOL, ELECTROLYTIC,PAPER and mica condensers are inmost of the specifications of the radioindustry.They are good - because there is acreed of quality behind each con-denser, whether it is a ten -cent tubu-lar or a thousand -dollar mica unit. Acreed that dates back twenty-five years-that is backed by the greatest engi-neering skill ever assembled in oneorganization, devoted exclusively tothe manufacture of condensers.

You can't afford to substitute.Cornell-Dubilier guarantees youabsolute satisfaction.

Write for your copy ofCatalog 128

THERE IS NO CIRCUIT, THE EFFI-CIENCY OF WHICH CANNOT BE

IMPROVED BY THE USE OFC -D CONDENSERS.

CORNELL-DUBIC OR P OR A T

4343 BRONX BOULEVARD N

IERONW YORK

September, 1935 RADIO WORLD 3

J. E. SMITH, Pros..National Radio Institute

HERE'S

PROOF

$18 a Week in SpareTime

"Although Iam doing onlyspare time Ra-dio work, Ihave averaged$18 a week. Irecommend N.R.I. training.It is certainlya completeCourse." STEPHEN J.DRAPCHATY, 407 Wun-derlich Ave., Barberton, 0.

Nets about $50 a Weekbesides Sales

"I have beengetting alongtine. 1 aver-age ten callsa week, whichnets me about$ 5 0 , notcountingprofits onsales I have

serviced almost every make9f set and have earned morethan I ever expected." BER-NARD COSTA, 150 Frank-lin St., Brooklyn, N. Y.

Get my FREE LESSONon Radio Servicing TipsI'll prove that my trainingis practical, money -makinginformation, that it is easyto understand - that it isJust what you need to mas-ter Radio. My sample les-son text, "Radio ReceivingTroubles - the Cause andRemedy" covers a long listof Radio receiver troublesin A. C., D. C., battery,universal, auto, T. R. F.,super -heterodyne, all -wave,and other types of sets. Anda cross reference systemgives you the probable causeand a quick way to locateand remedy these set trou-bles. A special section isdevoted to receiver check-up, alignment, balanc-ing, neutralizing andtecting. Get this les-son Free. No obligation.Just mall coupon.

U

I WILL TRAIN YOU AT HOMEIN SPARE TIME FOR A

GOOD RADIO JOB !FREE BOOK

TELLS HOW

MAIL COUPON

Act today for better pay.Act today to breakaway from a low pay, no -future job. Act to getaway from having to skimp, scrape to pay yourbills. Mail coupon far my free $4 -page book. Ittells you how I will train you at home in yourspare time to be a Radio Expert; about my train-ing that has doubled and tripled the pay of many.

Many Radio Experts make $40,$60, $75 a Week

Consider these facts-think of the good jobs theystand for. Over 16,000,000 Radio sets in use, over600 broadcasting stations, over 40 large manufac-turers of Radio sets, over 3,000 manufacturers ofparts, over 100 Police Departments Radio equipped,airplanes and airports Radio equipped. Thou-sands of ships touching every seaport of theworld, Radio equipped. Over 35,000 stores sell-ing sets and parts, about 2,000,000 autos Ra-dio equipped and about 20,000,000 unequipped.Loud speaker systems wherever people gather, in-doors and outdoors. Commercial Radio stationsdotting our coast lines. Radio a big industry-isgrowing bigger fast. A few hundred $40, $60, $75a week jobs, have grown to thousands.

Get ready now for Jobs like theseA spare time or full time service shop; installing,maintaining, operating-broadcast, aviation, com-mercial, ship, television and police stations. ARadio retail business of your own. Installing,maintaining, servicing, loud speaker systems. Aservice or sales job with a store or jobber. I'lltrain you for good jobs in every branch of Radio.Many make $5, $10, $1.5 a week ex-

tra in Spare Time almost at onceEvery neighborhood can use a good part time serviceman.I'll start giving you special instruction material, plans,ideas, the day you enroll, for making money in sparetime. Get my book-read how many of my studentsmake $200 to $1,000 its their spare time while learning.

Your money back if not satisfied11 make this agreement with you. If you are not

entirely satisfied with my Lesson and Instruction Serv-ice when you finish. I'll refund your tuition.

Find out what Radio OffersMail the coupon. My book of information onRadio's spare time and full time opportuni-ties is free to any ambitious fellow over 15.Read what Radio offers you. Read about thetraining I offer you. Read letters from grad-uates-what they are doing and making.There's no obligation. Mail coupon in anenvelope or paste it on a postal card-NOW.

J. E. SMITH, PresidentNational Radio Institute, Dept. 51M4

Washington, D. C.

SAVE MONEY --LEARN AT HOMESpecial Equipment Gives You

Practical ExperienceHold your job. No need to leave home and spenda lot of money to be a Radio Expert. 111 train youQuickly and inexpensively right at home In yourspare time. You don't need a high school or col-lege education. Many of my successful graduatesdidn't finish grade school. My practical 50-50method of training-half with lessons, half withRadio equipment-gives you broad practical experi-ence-makes learning at home easy, fascinating,practical and rapid. There is opportunity for youin Radio. Old jobs are becoming more complicated-many need better trained men. New develop-ments are making new jobs. Short waves, loudspeaker systems, police Radio, auto Radio, avia-tion Radio, television-Radio's newest uses arecovered by my training. Here's a field that's grow-ing. It is where you find growth that you findopportunity.

I have doubled

and tripledthe salariesof many

J. E. Smith, PresidentNational Radio Institute, Dept. 51M4Washington, D. C.

Dear Mr. Smith: Without obligation, send me theSample Lesson and your free book about spare timeand full time Radio opportunities, and how I cantrain for them at home in spare time. (Pleasekrint plainly.)

Name tge

Address

City State14X 1

RADIO WORLDThe How -to -Make -It Monthly-Fourteenth Year

ROLAND BURKE HENNESSYEditor

HERBERT E. HAYDENAdvertising Manager

HERMAN BERNARDManaging Editor

Table of Contents for SEPTEMBER, 1935

Television Race Gets HotterStations Building in U. S., Much Activ-

ity AbroadBy Herman Bernard

Farnsworth Demonstrates Prog-ress 11

His Cathode Ray Television System isStriking

By Neal Fitzalan

Elementary Tube Voltmeter 14The 6C6 in a Universal Circuit

By Jack Tully

The Multivibrator 16Audio Oscillator Rich in Harmonics

By Morris N. Beitman

Power Amplifier Taxes IngenuityArtistic Amphitheatre with No Emplace-

ment for OneBy Herbert E. Hayden

Full Scale "On the Nose" 20Necessary for Correct Ohmmeter Prac-

ticeBy Ambrose Bishop

Underwriters' Rules for Trans-mitters 23

The 6B5, a Tube with Built-inCircuit 24Held the Most Interesting Valve Since

the Vari-MuBy K. N. Satterwhite

Installation in a Console 26Some Pointers on Putting Set into Place

By Herbert Irwin

5 Filter Design and ApplicationSimple Circuits for Audio Systems


Small Rectifiers for B Supply 31Handy for Testing and for Powering

Small SetsBy Jack Tully

Experiences with 12 -Tube AllWave Super 33Gain Is Upheld at the High Frequencies

By Samuel Miller

A High Fidelity Power Amplifier 34Multiple Speakers and Fixed Bias

By Roy L. Smith

Diode Amplifiers in T.R.F. Set19 A Way Out of a Common Difficulty

By Henry Bullitt

Directional Transmission 38How It Is Applied to Ultra Waves

By J. E. Anderson

28

37

A Little More Meter Insurance 42Switch's "Lag" Turned to Safety Method

By Edward B. Nelson

A Balanced Detector 44

A 5 Meter Pilgrim's Progress 46Detailed Report of an Adventurous Ham

By Arthur H. Lynch

The New Metal Tubes in Practice 52Higher Mutual Conductance, Smaller

Grid -Plate CapacityBy George Dubuc

Radio Construction University 55

Vol. XXVII, September, 1935. No. 8. Whole No. 684. RADIO WORLD, published monthly by Hennessy RadioPublications Corporation, 145 West 45th Street, New York, N. Y. Editorial and executive offices, 145 West45th Street, New York, N. Y. Executives of RADIO WORLD: Roland Burke Hennessy, editor and businessmanager; Herman Bernard, managing editor; Herbert E. Hayden, advertising manager. Officers of corporation:Roland Burke Hennessy, president and treasurer; Herbert E. Hayden, vice-president; Herman Bernard, secretary.Entered as second-class matter March, 1922, at the Post Office at New York, N. Y., under Act of March 3d, 1829.

SUBSCRIPTION RATEPrice 25 cents per copy. Subscription $2.50 per year (12 issues), postpaid in United States and Possessions;1113.00 per year postpaid in all other countries. Remittance for foreign subscription should be by International

money -order (postal or express) or draft against or discountable at New York bank.

RADIO WORLD is distributed by the American News Company and its Branches


Television Race Gets HotterWorld Trails America, Which is Getting Experimental

Transmitters Ready-Unsolved Problems Knotty

By Herman Bernard

In the background is shown the Yag: transmitting antenna erected by VE9AK, Montreal, Can., fortransmitting television on 6 meters. The system consists of critically placed and selected reflectingantennas and a directing antenna, all in addition to the energized radiator itself. In the foregroundare J. Francis Dusek (left) and J. Lawrence Cassell, engineers of Peck Television of Canada.

ELEVISION is moving at a faster pace1 toward commercialization than at any pre-

vious time in history, and the advance is beingmade in several countries. Sizeable picturesare now being produced with definition andillumination that give promise of real enter-tainment value, the test being whether theselaboratory demonstrations can be duplicated inlarge scale co-ordinated service, or, as the engi-neers say, "in the field."

In the world wide aspect, arrangements havebeen made between television development com-panies in different countries for the exchangeof patent rights. In the domestic field, however,there is great rivalry, not only among differentsystems of accomplishing television, but alsoconcerning the course that television develop-ment should follow. Some, like Radio Corpora-tion of America, hold that each developingagency should be permitted to carry out itsexperimental and field work along its own lines,while others favor Government insistence on

standardized transmissions, so that a separatereceiver will not be necessary to pick up eachsystem's propagation.

The Four Out FrontFrom each country come reports that, in ef-

fect, tend to create the impression that tele-vision there is of the highest quality anywhere.Some "scouts" for investing interests do noth-ing but go about the world, investigating tele-vision development, and, though they are not allAmericans, they report that systems they haveseen here surpass anything they have seen else-where.

In a recent report to the stockholders of hiscorporation David Sarnoff, president of R. C.A., said "the results obtained by R. C. A. inlaboratory experiments go beyond the stand-ards accepted for experimental television ser-vice in Europe."

The four corporations in the United States(Continued on nest page)

6 RADIO WORLD September, 1935

(Continued from preceding page)that are out front in television are R. C. A.and Farnsworth Television, Inc., with electricalsystems without moving parts ; Peck TelevisionCorporation and National Television Corpora-tion, which have mechanical systems.

All have been introducing improvements forseveral years and are showing better and largerpictures. R. C. A. uses a 343 line picture, com-pared to 30 lines of a few years ago when itwas intensely developing mechanical systems,and now has 60 pictures per second, comparedto 12 pictures then. Farnsworth uses a finelineage and high picture frequency, too. Peckand National at present have 60 lines per pic-ture and 24 pictures per second.

Scanning Is Common to AllAll television today is on the basis of scan-

ning, which consists of breaking up the pictureinto elements, and transmitting the equivalentvalues of the elements for recomposition at thereceiving end. The light values are changed atthe transmitter into current values, by usingsome photo electric agency, and the frequenciesof these currents modulate the carrier. In me-chanical systems a disc or drum revolves, re-flecting surf aces at different angles picking upbits of the picture one by one. In the electricalsystem a camera and tube do the work, and thissystem lends itself readily to direct pickup, useof so-called "live subjects," compared to moviefilm or stills.

The two advanced exponents of the electricalsystems feel that theirs are respectively supe-rior to mechanical methods, whereas the in-ventors of mechanical systems believe they canproduce results that the others can not match.As to actual demonstrations, although few ofthese are given, and mostly in private, observerswho say they are impartial find defects in both

advantages in both, tell their adverse

The women are shown at minimum distance fromthe television viewing screen for enjoyment of60 line pictures, the men at the greater dis-tance from which 180 line pictures may beviewed with equal resolution of the lines of thepicture. A wash drawing is on the televisionscreen to emphasize where the picture appears.

Willi a dipole antenna and a receiver thestrength of the transmission from VE9AK is

measured at various distances from the source,up to about 60 miles.. By using greater power,around 600 watts, it is expected that more than100 miles of reliable service will be attained on 6meters. Transmissions are now conducted daily

criticisms of one system to the backers of theother, and confine their praise to the systemof the men to whom they are talking.

No Claims of PerfectionNone of those developing television asserts

that anything approaching perfection has beenachieved. It is also pointed out that since muchhas to be accomplished before a co-ordinatedservice of consequence can be instituted, thatthere will be rapid and costly obsolescence, oneof numerous reasons for not bringing out re-ceivers now, for sale to laymen. However, ex-perimental television stations will be operatingon a high definition basis within a year or soin this country, particularly one that R. C. A.is about to erect, probably in New York City,and experimenters may be able to do some un-invited viewing with apparatus of their ownthat may be changed from time to time tomeet new requirements.

Considerable impetus was given to televisionprogress from a psychological viewpoint by thereport of the British Television Committee tothe Government, advocating immediate inaugu-ration of television. The word "immediate" orthe phrase "at once" has a delayed significancewhen applied to television.

Trying Our System Over ThereWhat has been shown in Britain generally

has been low definition television, usually 30lines, but now electrical systems are being tried,particularly the one developed in R. C. A. -Victor laboratories in this country under Dr.Vladimir K. Zworykin. Electrical Musical In-dustries (known as E. I. M. in Britain) has aworking agreement with R. C. A., while BairdTelevision, Ltd., of Britain, has a similar agree-

That somewhat the same problems of policyexist in all countries is suggested by the decreeof the British Broadcasting Company that theemissions of two rival companies, Baird andElectrical Musical Industries, alternating inpropagation from the Crystal Palace plant, bereceivable on the same set.

The First Television "Election"Evidently the public is to be at least one of

the judges of the results. Thus perhaps in Eng-


ment with Farnsworth. At present Germany isgoing it alone, and seemingly doing better thanEngland, while France is trying to keep upwith Germany, but so far with little successin the television line.

In Germany, as in England, television isGovernment sponsored, whereas in the UnitedStates and Canada the work is undertakensolely by private initiative. Germany, for in-stance, has viewing places, where interestedthrongs flock to look at the short distance tele-vision. The system is one of movie -filming ascene, developing the film in less than a minute,and putting the film through the transmittingscanner. There is no public sale of televisionsets, and the results as yet do not encourage theofficials to permit such sale.

From France very little is heard about tele-vision developmental work, while from Eng-land come rather frequent reports, some ofwhich sound like impossible dreams come true,

d

A reflecting transmitting antenna system used by Telefunken for centimeter waves. Installations forfrom 10 to 90 centimeters have been made. Much more power is radiated when the wave is beamedby this method. At the receiving end, for short enough waves, a parabolic reflector is used (at

right). The receiving aerial is a small wire at the focal point of the parabola.

but the realities turn out to be something lessthan a prudent person would regard as out-standing. The policy in England is to attemptto produce high definition television, as a pub-lic service, as soon as practical, without regardto the nationality source from which the devel-opment springs. Thus foreign systems are wel-come, and are being tried out, invitations toseveral more having been issued.

land there may be a decision in the bout be-tween the R. C. A. and the Farnsworth sys-tems, both American, before there will be sucha determination in the United States. TheFarnsworth Corporation is arranging to put upa transmitter, and expects that several will begoing full blast within a year.

All interests are a. -4 -reed that television trans -(Continued on next page)


mission will have to be on low waves, becausethen the modulation band width is a small per-centage of the carrier frequency, and well ac-commodated. For high definition systems theband width might run into millions of cyclesper second. At present the reliable range forwaves around 6 meters is around 25 miles orless, and with this limitation is grouped thenecessity of providing links for any systemco-ordinated to a large geographical area likethe United States, and a medium for establish-ing such links.

The Coaxial Line

Here another laboratory development enters,the coaxial transmission line of the Bell Tele-phone Laboratories. While the Laboratoriesare primarily concerned with voice transmis-sion as part of the regular service, and thecoaxial lines now readily enable sixteen mes-sages compared to standard practice of fourmessages per carrier, the company also leaseslines for broadcasting purposes, and if itcould develop in the field the use of a systemto include television modulation band it wouldendow television with a real benefit. It is con-fidently expected that the field experiment withthe coaxial lines to be erected between Phila-delphia and New York will confirm the utilityof this device in practice, for in the laboratoryit was a great success. In fact, the coaxial linein the laboratory is far ahead of television inthe laboratory. And the line can carry a 1,000,-000 cycle frequency readily, and even far higher

( Continued an next page)

( Wide6L-urldj

Operators atop motor trucks go about the streetsof Berlin shooting scenes with a movie camera.The exposures are developed in about a minuteand thus made ready for scanning by a televisiontransmitter. The public is permitted to see the

reception results in booths about the city.

11 4k II r1 )

Direct pickup, using live subjects instead of film, is under experiment in Germany, also. A womanwith a contrast), dress is shown before the microphones and the pickup. Contrasts are what are

particularly studied in the experiments.


RCA and Philco Negotiatingfor Coaxial Television Line

(Wide World)

Samples of coaxial transmission line cable as de-veloped by Bell Telephone Laboratories wereexhibited by Dr. Frank B. Jewett, president of theLaboratories, at a hearing before the Federal Com-munications Commission on the application forpermission to install the line between New Yorkand Philadelphia. Left to right, CommissionersIrwin Stewart and Eugene 0. Sykes, and Dr. Jewett.

Washington.Opposition having been withdrawn by the

Western Union Telegraph Company and thePostal Telegraph and Cable Company to theinstitution of an experimental coaxial line byBell Telephone Laboratories between NewYork and Philadelphia, the Federal Communi-

cations Commission announced that the requestfor permission to install the line would begranted. The assurances of the experimentalnature of the installation overcame the objec-tion.

Dr. Frank B. Jewett, president of the Labora-tories, said that the service would be experi-mental for at least a year. It was denied therewould result any disruption of existing com-munications systems, which fear had also ledsome broadcasting stations to offer opposition.Finally all opposition was withdrawn, except-ing that of one lawyer, representing soundmovie interests.

Dr. Jewett said that no attempt would bemade to monopolize the coaxial service, butthat reputable concerns would be welcome torent the service for television that met standardrequirements. Already negotiations have beenstarted with Radio Corporation of Americaand Philco Radio and Television Corporation.It was denied A. T. & T. is perfecting a tele-vision machine.

The cost of the New York -Philadelphia in-stallation is estimated at $580,000.

Quiet operation of the line is attained byintroducing just enough reverse feedback. It isextremely novel to introduce reverse feedbackin a repeater, which is of course an amplifier,but by discarding some of the possible gain thisway the ratio of signal to noise was greatlyincreased.

(Continued from Preceding page)frequencies. Thus the television current varia-tions that are to be mixed with the short wavetelevision carrier-that is, the modulation orvideo frequencies-could be sent by wire just aswell as audio messages. A central studio thencould send television pulsations to various trans-mitters for propagation and thus the problemof local service, due to short range of the low

radio wave, could be solved by the wire inter-vention.

Coaxial Impedance Computable EasilyThe coaxial principle is an old standby in

physics, but the contribution of the Laboratoriesconsists of the utilization of the principle alongpractical lines of the modern need. The prin-

(Continued on next page)

Bell Telephone LaboratoriesThe outer covering of the coaxial line is a lead cable Yg inch in diameter. Inside are two copper tubingsand eight usual type telephone wires, paper insulated. Each of the two copper tubings has insideit a small copper rod, insulating spacers between rod and inside well of the tubing, so that the

dielectric is mostly air. to avoid losses. Cross section view is shown at left.


DAVID SARNOFF, president, Radio Corporation of America: "The resultsattained by RCA in laboratory experiments go beyond the standards accepted forthe experimental television service in Europe. We believe we are further advancedscientifically in this field than any other country in the world."

(Continued front preceding page)ciple is based on the transmission propertiesof two conductive tubings, one inside the other,that is, the two have the same axis.

For short distances the coaxial principle isapplied to a transmission line in the strict senseof conduction from source to destination with-out loss, familiar in radio practice. It is diffi-cult, certainly at high radio frequencies, tomeasure the impedance of a transmission line.More accurately the impedance can be com-puted, and the coaxial feeders of radio practicerepresent the selection on that basis. The cir-cular symmetry makes things easy for themathematician, and since practice is based onthe computation, which is highly accurate, thecoaxial feeders are accepted as being unsur-passed, that is, as lead-ins do not pickup toannoy a receiver or transmit to confound atransmitter.

Where the coaxial line is to run any consid-erable distance, say, 20 miles, it is necessary tohave a repeater.

"The attenuation of copper wire lines alwaysvaries with temperature-overhead lines beingperhaps 10 per cent. higher than average onthe hottest day in Summer and 10 per cent.below average on the coldest day in Winter.On a transcontinental coaxial circuit this varia-tion would reach tremendous proportions, andeven on a circuit from New York to Philadel-phia the variation would be sufficient to makethe conversation impossible unless the trans-mission were regulated in some manner. Sincemost of these repeater stations would be un-attended, their regulation must be automatic."

What applies to telephone conversations alsoapplies to television.

Mr. Strieby continues:"The coaxial line is also well adapted for

transmitting over long distances the extremelybroad frequency bands required for television.Since the frequency range for high grade tele-vision extends from the neighborhood of zeroup to a million or more cycles, special modu-lating apparatus is required at the terminals to

*,Isizzt

13,11 Telephone LaboratoriesTwo types of Western Electric coaxial conductor that have proved very satisfactory under test.

The coaxial line, Bell Laboratories found,lends itself to broad band transmission, and therepeater could amplify the whole band.

Tests were made of built up coaxial lines, ina check on theory, the line consisting of a cop-per tubing 2.5 inches outside diameter withinwhich was mounted a smaller tubing, which inturn contained a small copper wire. Thus twocoaxial circuits were produced, one consistingof outer tubing and next adjacent one, theother of outer tubing and the wire inside thesmaller diameter tubing.

M. E. Strieby, carrier transmission researchengineer, reports in an article entitled "CoaxialConductor Systems," in "Bell LaboratoriesRecord," for July issue (1935) :

lift the entire band to a range which can betransmitted over a coaxial system."

As for voice telephony alone, a million cyclecoaxial line would carry more than 200 voicechannels.

WIRED RADIO TESTS RESUMEDCleveland

Wired Radio, which tried out sets in homes,where the programs were sent over light wires,stopped the experiment, after three months,for a study of the results. More than 700 fami-lies had metered sets in their homes and theaverage use per day turned out to be fourhours. Service then was resumed.


Farnsworth Demonstrates ProgressHis Television Clear, Luminous, in Black and White

By Neal FazeIan

(Wide World)The Farnsworth television studio in action. The direct pickup was transmitted by cable to a roomabout 40 feet away, where the reproductions were viewed and heard in a recent demonstration inPhiladelphia. Musicians played and stage girls danced and sang. Note the powerful light source

at right, announcer at microphone at left rear.

TELEVISION of good quality, with bothdirect pickup and movie film used for the

program, was demonstrated by Television Lab-oratories, Inc., in its studios at Chestnut Hill,Philadelphia, recently. The pictures were ofsmall size, 51/2 x 7 inches when projected, butthe definition and illumination were much betterthan at a previous demonstration of the samegeneral system at Franklin Institute.

The inventor, Philo T. Farnsworth, not onlyused both direct and indirect program sources,but also sent the results of one by radio, forreproduction in a console type receiver, withgreenish lighting, and the other by wire, shownon an external screen in black and white. Soundeffects were reproduced in both instances. Heexplained he can increase the size of the pro-jected picture materially, with adequate lightleft.

The principal objection that some of the ob-servers had to the results was that the picture

would wobble occasionally, but Mr. Farnsworthexplained that this took place during the partof the demonstration dealing with transmissionby radio, that the power used was small, andthat increased power cured that trouble.

He said that though the transmissions wereput on only about 40 feet from the receiver, theclarity would be the same for distances up to30 miles.

The studio was fitted up in approved style.with a small stage, faced by a powerful lightsource. On the stage girls sang and danced,while an orchestra played, and the usual formal-ities were observed, with William Eddy serv-ing as announcer. The group of guests in an-other room in the same building heard and sawthe antics of the human participants, and also,in the case of the indirect pickup demonstration,heard and saw Mickey Mouse in a televisionscreen reproduction of a movie reel.

(Continued on next pcsge)


(Continued from preceding page)Farnsworth uses an electrical scanning sys-

tem having no moving parts. He has his owndevelopmental form of cathode ray oscilloscopetube, one with hot cathode for receiving pur-poses, another with cold cathode for economicaloperation of a transmitter. His pickup tube hecalls the "oscillight." In purpose this corres-ponds to the iconoscope used by R.C.A.

Shows Black and White

The direct pickup scene to be shown is picked

(Wide World)

Philo T. Farnsworth holds a small neon lamp nearthe transmission line feeding the cathode ray

tube at his receiver. If the line is not truly a

transmission line, that is, radiation takes place,the field will light the lamp. Purposeful unbal-ance would accomplish this, but the lamp doesnot light, because the line is balanced. The

speech amplifier and the sound reproducer areshown in the bottom compartment.

up on a camera and the equivalent ground glassis scanned by the Farnsworth method and thelight values converted into electrical values formodulation of a carrier, in the case of radio, orsent directly on special wires, in the case ofwired television.

The conducting wires, once a problem, have

been developed to such a point that carriertelevision and wired television results could notbe told apart, except that in this instance theywere shown on different screens, and the fluo-rescence of the receiving tubes was of differentcolor. The greenish results familiar to cathoderay tube users was due to zinc orthoscilicate.The wired television was received on a tubetarget composed of calcium tungstate and othersubstances, the ingredients being held secret,as getting black and white results with a tubeof that kind is considered an achievement.

During one of the stage shows there weresixteen persons in the scene, including chorusgirls, and it was possible to distinguish themclearly. In fact, among the guests were friendsof the stage folk who recognized them in-stantly, and applauded. This was not taken bythe inventor as the compliment that probablywas intended, since television several years agopassed the point when mere recognition ofidentity constituted proof of progress. Butnow there were more numerous participantsthan formerly, and the distinctness with whicheverything could be seen was well appreciatedby the viewers.

24 Pictures, 240 Lines

The modulation band was about 2 mgc.Such a wide band, or close to it at least, isregarded as necessary for high definition, andrequires that if a carrier is used that its fre-quency be high, so that the modulation or elec-trical impulses that constitute the picture ele-ments shall be a small percentage of the carrierfrequency. Hence Farnsworth's television, likethat of others, is necessarily intended for shortwaves, say, not more than 7 meters.

The net effect of the operation of the scan-ning system is that a very fine spot of light ismade to move over the picture area 5,760 timesa second. There are 24 pictures per second, tocreate the illusion of motion without flicker dueto low picture frequency, the same repetitionrate as used in the talkies, and there are 240lines per picture.

How the Spot MovesThe spot of light is moved across the screen,

as if starting at the top left hand corner of thepicture. traveling to the top right hand corneron a straight horizontal line, then being car-ried to the left in an unobserved return sweep,to start going from left to right again, onetwenty-fourth the distance from the top of thepicture, and next time one twelfth the distancedown, etc.. until the picture is scanned verticallythis way by a light spot occupying what may becompared to rung positions on a closely formedladder, and moving from left to right on eachrung, in a top to bottom progression. Therewas no trace of the lines observable, as evenat a distance of seven feet or so the lines are soclose as to be invisible, due to the resolvingpower of the eye, that is, multiples appear as aunit.

The military and naval branches of the Gov-


ernment are following developments at Tele-vision Laboratories, Inc., as they are at R.C.A.and other places, because of the great pos-sibilities that television holds forth as a defen-sive and offensive weapon in warfare.

The goal of being able to see distant objec-tives for directing explosives toward particular

targets has not yet been reached, but that pos-sibility exists at least in theory, and is the primereason for the deep interest the armed forceshave in television in this country, just as thefighting branches of foreign governments exer-cise the same watchfulness over television prog-ress at home. The American Army officers

are cooperating with otherbranches of the governmentin keeping close tabs on tele-vision progress.

Opinions still differ as towhen television will be hereas a commercial propositionof entertainment value.David Sarnoff, president ofRCA, is said to lean to "10years", but Farnsworth saysit will be much sooner thanthat.

(Wide World)Operator is shown at theFarnsworth transmitter.An entirely electrical sys-tem is used for scanning,both at transmitter andreceiver, based on theprinciples of the cathoderay oscilloscope. Hencemoving parts are omitted.

Germany First, Britain Second AbroadA. W. Cruse, chief of the electrical division,

Department of Commerce, has returned to theUnited States with a report to the NationalAssociation of Broadcasters, who had askedhim to make a survey of television conditionsand results in Europe. The following is a sum-mary of the report:

ENGLAND: British Broadcasting Companyoffers low definition television now, half tothree quarters hour, twice a week. Televisionon 261 meters, sound accompaniment on 398meters. Flicker is bad, picture indistinct, butprogram technique excellent. Great variety of"live subjects" scanned, vaudeville and con-densed operas included in offerings. Numberof television sets on British Isles said to beabout 100. British radio manufacturers fightingtelevision as hurtful to set sales, but sales thisyear may exceed those of last year. Some

movie folk see no threat by television to the-atres.

GERMANY: Berlin crowds flock to televisionviewing places. Using 180 lines, Germany laysdown an excellent picture, of high entertainmentvalue. License fees on sets, plus a Governmentappropriation, finance television experiments.Germany moving along at a swift pace, withthorough technique.

FRANCE : Realization exists in that country.that it is behind Germany and efforts are beingmade by the Ministry of Posts and Telegraphsto catch up. Transmissions now being triedirregularly, 60 lines, 25 pictures per second, on17.5 meters. Trying to get 90 lines for earlyexperiments and expect to settle on 180 lines asquickly as they can, but will take more thasix months.


Elementary Tube VoltmeterMeasures A.C. Potentials, 1/4 to 2.8

By Jack Tully

1.0 word.

All o I hers 1.Power trans. sec. <

6C6

0-5 Ma.

0.05m.fd.

III-1 amp..fuse 5w.

ASIMPLE vacuum tube voltmeter is shown,for measurements from 0.25 volt to 2.8

volts. The curve is given, showing a. c. voltsrelated to d. c. plate milliamperes, when a 0-5milliammeter, having a d. c. resistance of 2,160ohms, was used (Readrite No. 305).

A 6C6 tube has screen tied to plate, as themutual conductance is higher that way, alsosuppressor to cathode, for the same reason,although the difference when connecting sup-pressor to C minus is negligible.

A line cord is connected to the a. c. line.There is a resistor of 350 ohms built into theline cord (third lead), to drop the line voltageto the required heater voltage. Grid of thetube, represented by the head, is brought outto a binding post. C minus 3 volts is broughtout to a separate binding post and the connec-tion is made to a similar post through a con-denser of 0.01 mfd. or higher mica capacity,from the aforementioned ground lead.

The Input Connections

Between the grid post (center) and the directconnection to battery are put low voltages fromsecondaries of ungrounded power transformers,or low a. c. voltages from any other source notdirectly related to the a. c. line or actual ground.Between grid and the upper post are put allradio frequency voltages, because many of themwill be taken from receivers, generators, etc..where chassis and coil returns are grounded.The 0.01 mfd. condenser avoids shorting theline, since otherwise the tube voltmeter could

.communicate to grounded chassis the un-grounded side of the line and cause the short.

The plotting of a. c. volts against d. c. platecurrent shows that the curve is linear fromthe minimum, 0.25 volt, to 1.5 volts, or 1.3 to2.35 ma, after which the curvature is obvious.For the linear tart the change is 0.2 milliam-pere in d. c. plate current for 0.25 volt a. c., on

Zmfd.Po per

For measuring smalla.c. voltages this cir-cuit may be used.Ungrounded exter-nal systems, such aspower transformersecondaries, may bemeasured withoutthe input currentflowing through theI mfd. condenser.All other measurements are madethrough the con-

denser.

the grid, or, expressed as a resistance, 125,000ohms, or as a conductance, 400 micromhos.

Across the meter is a paper dielectric bypasscondenser, to keep as much a. c. out of themeter as that capacity will accomplish, althoughat 60 cycles the job is not done completely, and8 mfd. would be preferable. That, however,would be by way of bringing the small deviceinto more serious dimensions.

Needle May Kick Down

A leak is included with switch so that thegrid can have a return, without depending onthe load for d. c. continuity, which in some in-stances it will not have.

Thus the grid may be operated at the condi-tion at minimum plate current, leak switchclosed, except that when an unknown a. c. isconnected other than in the intended direction,the negative alternation drives the grid negativeand decreases the plate current. The curvestarts at more than 0 plate current, in fact, 1.3milliamperes, and if the needle kicks backward,reverse the connections of the unknown, usingthe same posts as before, but now the positivealternation is toward the grid and drives theplate current up. This increase in plate cur-rent is what enables the determination of thea. c. voltage from the curve.

Calibration Holds for All Frequencies

Nearly all systems that will be measured willhave one side grounded, and if not, the a. c.grounding will be accomplished through thetube voltmeter. Thus 0.01 mfd. grounds radiofrequencies sufficiently, and in other examplescited, for the direct connection to C minus,hence to line, the same condition of groundingas to the a. c. exists for the tested circuit. Forpower transformer secondary measurements


#t)

3.5"

r

1-i

0.25 0.5. 1.0

/7.C. Vol_ TS3.0

The plate current is read in the perpendicular column of figures at left, and equals 0.05 milliampere perivision, or, two divisions for each 0.1 milliampe re. The meter used has division bars for each 0.25silliampere, but the readings can be taken sufficiently for the present purpose. Carry plate current

line across until the curve is intersected, and then come down to the a.c. volts line.

(low voltages, of course) the condition is obvi-ous, since the secondary may add to or subtractfrom the grid voltage. Hence the two voltagesshould be in phase. A correction may be intro-duced by reversing the connection to the wallsocket. For radio frequencies the difference infrequencies is so great that the condition willnot likely obtain, though any effects, if noticed,should .be corrected either at the input posts or,if more convenient, at the wall socket.

While the calibration was made at 60 cycles,it is substantially correct at all frequencies,say, to 10 mgc., not being critical to line volt-ages within the variations encountered in largecenters of population. The line voltage was 116

volts when the calibration was run, and at 110volts the 2.5 volts reading might be expectedto be reduced to 2.37 volts, but was nearer 2.47volts.

Leave Leak Switch Open

With connections properly made to input, themeasurement of relative values of radio fre-quency voltages as found in receivers, etc., atlevels above 0.25 volt, to 2.8 volt, may be made,with high accuracy, in respect to establishingthe same voltage in two different circuits, andalso of course absolute measurements, by using



THE MULTIVIBRATORw Construction, Operation and Application

By M. N. Beitman

Basic circuit above, wave pattern at right.

ALTHOUGH a multivibrator is finding ex-tensive application in station monitoring

and laboratory equipment, it is a little knowndevice. In this article the author will attemptto explain the operation of a multivibrator, men-tion some of its more common applications andgive plans for construction of a small multi -vibrator from spare parts found in any radioworkshop.

The multivibrator is one form of a relaxa-tion oscillator. One of its simpler forms is noth-ing more than two capacity -resistance coupledamplifiers, so connected that the output of oneis coupled to the input of the other, while theoutput of the second is coupled to the input ofthe first. This oscillator employs no inductanceand was originally due to Abraham and Blochas described in "Comptes Rendus" of 1919.

The Basic Circuit

The basic circuit of such a simple vibrator isshown in Fig. 1. On the first examination itwould seem that upon connecting the filament

Time

The wave pattern.

and plate supply, the two tubes would drawsteady current. However, this is far from whatactually takes place. The circuit oscillates vio-lently, producing waves that approach a squareshape, as illustrated in Fig. 2. This representssudden rise and fall of current and is exactlywhat might be expected from a closer examina-tion of the circuit in Fig. 1.

Once the circuit is connected a steady currentwill begin flowing through the two vacuumtubes and the associated plate resistors r, and r:.Suppose that through disturbance(no matter how minute) the current throughri is increased. Since the voltage drop acrossthis resistor is equal to the product of the re-sistance by the current, the voltage across riwill increase, because of the increase in current.The condenser C2 will act as a short circuit forthe sudden change of voltage, and a higher neg-ative voltage will be placed on the grid ofVT -2, reducing the current through it.

How the Action Repeats ItselfNow we have seen that the increase of cur-

rent in one tube will decrease the current in the(Continued on next page)

Vacuum Tube Voltmeter(Continued from preceding page)

the plotted curve, although with somewhatlesser accuracy.

It will be noticed that with the 2 mfd. acrossthe meter there is not quite enough capacityto remove the wabble due to residual a. c. fromthe line, which removal larger capacity wouldaccomplish, but it is very easy to estimate withthe eye the average of the small displacement,and read the direct current accordingly. Thatthis condition is also associated with the shapeof the curve seems to be borne out by the factthat no such wabbling occurred where the curvewas straight, but only over the curved portion.

The device does not draw current from themeasured source, with leak in open position, andthis use is recommended provided the testedcircuit affords d. c. continuity to the grid cir-cuit of the tube voltmeter. Where stopping con-densers intervene, as in grid condenser prac-tice in oscillators, there is no d. c. continuityfor the tube voltmeter when the leak is in par-allel with the grid -cathode leakage of the oscil-lator tube, and then the 1 meg. of the tube volt-meter may be cut in. The current drawn by the1 meg. resistor even at maximum of the plottedscale is only 0.0000028 ampere, or 2.8 micro-amperes.


(Continued from preceding page)other, so that the reduction of the plate currentin VT -2 will further increase the current inVT -1 and ri, since this action works both ways.This action repeated again and again will finallyreach a point where, in this case, VT -1 willhave a maximum current going through it lim-ited only by the emission, while VT -2 will havesuch a high negative potential on its grid thatit will be entirely blocked.

What Determines the FrequencyThis action takes place during but a fraction

of the total cycle and accounts for the sharprise of the plate current. At this stage, whenone tube is blocked and the other is passingmaximum current, the circuit is no longer sym-metrical and the condenser begins to discharge.This discharge continues until the blocked tube

The author's multivibratorfor 60 cycles, using fourtubes, including the recti-fier. This is the circuitused in the constructeddevice shown in the pho-tographs on next page.From grid of first 45 toground is a 100,000 ohmresistor. From grid of first45 to plate of second 45 isa condenserBy increasing either orboth the frequency islowered, by decreasingeither or both the fre-quency is increased. For600 cycles use 100,000ohms and 0.01 mfd. For6,000 cycles 50,000 ohms

and .002 mfd.

begins to draw current and then the entireaction repeats in the second tube.

These oscillations by proper design may ap-pear of any frequency between 50 kc and 1/50of a cycle per second. The period is propor-tional to the product of the capacity and resist-ance of the circuit.

The sharp corners of the plate current curve,of course, indicate the presence of a large num-ber of harmonics. As high as the fiftieth har-monic is detected with ease. This presence of agreat number of harmonics and the fact thata multivibrator locks in step with any frequencythat is a multiple of its own approximate fre-quency, makes a multivibrator adaptable tomany applications.

Suppose a multivibrator, natural frequencyabout 10 kc, but not exactly this, is so con-nected that it is influenced by a 100 kc crystalcontrolled exact oscillator. Once this connec-tion is made the frequency of the multivibratorwill be noticed to shift a little, but upon carefulmeasurements will be found to be exactly 10

kc., i.e., exactly within the same relative errorpossible for the crystal oscillator. Changesin the multivibrator circuit elements or voltageswill have no effect upon the frequency withinwide limits. Under this operating condition themultivibrator is under the absolute control ofthe crystal oscillator.

The Decimal Fundamentals

It is necessary for broadcasting stations tokeep their operating frequency within 50 cyclesof the assigned frequency. Usually a carefullycalibrated quartz bar is used to generate someexact frequency (within about 0.003%). Barshaving a natural frequency of about 100 kc arecommonly employed. By utilizing three sepa-rate multivibrators, one generating 100 kc andits harmonics, the second generating 10 kc and

/0,000n

its harmonics, and the third generating 1 kcand its harmonics, any radio frequency com-monly used by broadcasting stations may beobtained. This frequency will be almost asexact as the frequency of the quartz bar.

As a check of the frequency a 1,000 cycleprecision electric clock is run from a 1 kcmultivibrator. The clock's time is checked fre-quently against standard time obtained fromastronomical observations and broadcast bygovernment stations. If the clock's time corre-sponds to the standard time within the allowedlimits, the station may be considered to be oper-ating within permissible frequency variations.On the other hand, a drift of the crystal fre-quency and, thereby the station frequency, willresult in an error difference in time. Theamount of gain or loss of the clock's time willindicate in what direction and how much thestation frequency is off the assigned figure.

This application of multivibrators as stationfrequency monitors suggests a laboratory use



(Continued from preceding page)of one or more multipliers as exact frequencychecking devices. Accuracy may be carried toa satisfactory degree with an ordinary 60 cycleelectric clock and using a good watch as thestandard of time. Fig. 3 and the photographsshow a 60 cycle multivibrator constructed fromspare parts. Other tubes may be used and thecircuit is not critical.

This unit in connection with an ordinary elec-tric clock* may be used to test any frequencybetween about 60 and 2,000 cycles. For exam -

and watch. What is the frequency of thesource?

The nearest multiple of 60 close to 1,000 is17, so that if the source's frequency were 1,020cycles the electric clock would show true time.But since one minute was lost, the frequencymust be less than 1,020 cycles. One minute is1/60 of an hour, the time the clock ran, so thatthe frequency of the source is 1/60 of 1,020 less(17 cycles less). This gives us as the frequencyof the source 1,003 cycles, a figure reliable to

Two views of the multivibrator constructed by the author, using the diagram shown on previous page.ple, suppose the source to be analyzed is about1,000 cycles. The source is connected to theinput terminals and the control is adjusted un-til the clock runs at about the right speed. Aftera period of time has passed, let us say one hourin this case, it is noticed that the electric clockhas lost one minute as compared to the stand-

* If the clock requires considerable power, an am-plifying unit will have to be made.

about the extent that clock scales can be read.The testing method outlined above may be

extended to higher ranges with the aid of oneor more additional multivibrators. Each addi-tional multivibrator will multiply the upper fre-quency limit by about ten. The application of anumber of multivibrators would enable ama-teurs to grind their own crystals with labora-tory precision, for the method outlined is ex-actly the one used in commercial laboratories.

Short Wave Battery Portable CircuitThe 19 will amplify

fairly well, and certainlywill oscillate, at 1.5 voltson the filament, insteadof the standard rating of2 volts. Since it is ad-visable to have an Asupply that will lastsome time, a No. 6 drycell is used, and the rea- 5 Plate -

son for confinement toone is that these cellsare large, and in a port-able short wave circuitthere may not be roomfor the extra cell.

The input lead is aradio frequency chokecoil, value not critical.Ll is the primary in theplate leg of the first tri-ode, L2 is the grid coilfor the second triode, and L3 is the tickle'affecting that grid coil. The large tuningcondenser is used for "rough" setting, the

V/9 5Lj

/40mot&

0 01m./4,1,1,6

1L gLz

.0/ -mid. "mica

C -/.5Y.

Circuit for a short waveportable receiver tocover either one band,or, by using plug incoils, several bands. Cplus goes to A minus.

320 mm14.

nimbi

Phones

sw

51-45Y.

A ASKA- ,B -

smaller one for finer setting, and the trimmeracross the smaller main tuning condenser forclose regeneration control.

Scptcloher, 1935 R AD I ) ) I) 19

Power Amplifier Taxes IngenuityBy Herbert E. Hayden

14-7-44- t- %,

Opportunity exists for the consistence of public address systems with beautiful surroundings. Oftena makeshift installation is necessary if crowds are to hear what is going on.

TI 1,1:F. ark ,,tise fine architectuni! creations, con, ,1\ ed with the idea of creating a beautifulI setting to commemorate some celel ity's achievements. The vogue for public address ampli-

fiers, and boosting the volume of radio programs for enjoyment by a large assemblage, taxes in-genuity in making an harmonious installation.

At right is shown a small part of the Edward M. Shepard Memorial Amphitheatre at LakeGeorge, N. Y., and amid the fluted columns and the landscape gardening there has to be placedsome means of rendering sound audible to large a ssemblies, for whom there are the -naturalized"stone steps serving as seats in the ancient amph theatre manner, as well as benches. The repro-ducer is shown perched on an urn, a contrast to the placid lake behind it, and the rolling hillsof the distance. At upper left is a closeup of the speaker, showing how the urn actually servesas support and container.

At lower left is shown a wooded scene in the same Memorial, with a speaker attached tothe trunk of a forked tree.

The use of amplifiers amid landscaped and architectural creations imposes a requirement ofconforming the electrical installation to the graces of the civic art. Artificial trees for suchpurpose may be used for artistic attainment, and architects will have to consider creating em-placements for speakers of large size and power, that do not suggest an intrusion upon the scene.


Full Scale "On the Nose"Necessary for Correct Ohmmeter Practice

By Ambrose Bishop

gx(Cow 2esistance)

r WvAv/4vAvAv%A-

E lex(Medium Resistance)

Disregarding the dashed lines, the series circuit isshown. The shunt across the meter is represented

by the dashed lines.

INohmmeter practice full scale deflection is

established in the meter, as a preliminarycheck, by closing the open terminals where laterthe unknown is to be connected, and then whenthe unknown is inserted its resistance is readdirectly, or computed, on the basis of the newcurrent reading. For the tests that include theunknown as a series resistance, the accuracy isnot affected when the current through themeter is changed, because this change is cali-brated in ohms. However, for the low resist-ance shunt method of measurement, currentthrough the meter changes also, and if not con-sidered, as sometimes it is not, reduces theaccuracy.

One diagram shows the conventional un-known series resistor circuit. For such servicethe meter scales are direct reading in ohms.There are hardly any commercial meters usingthe shunting method of low resistance with cali-bration in ohms. Since computation is used

often, the correction factor should be intro-duced. If the calibration were on the scale theerror referred to could be eliminated by includ-ing the differential in the calibration. Theformula in its accurate form is given sepa-rately herewith, while on page 49 of the Mayissue was printed a full -page curve for a typicalmeter, 0-1 milliammeter having 30 ohms inter-nal resistance, readings obtainable from about0.3 ohms to 250 ohms.

How the Shunt Method Works

Another diagram herewith shows the metershunting method. The basis of application isthat putting the unknown across the terminalsof the meter itself diverts part of the currentthrough the unknown, the remainder goingthrough the meter.

A calibration can be prepared, on the basisof the highly accurate formula, for the valuesof the unknown in terms of the current throughthe meter.

Incidentally, the needle moves in the oppositedirection when the shunting method is used,compared to the series method, because thehigher the unknown resistance, the less currentthrough the unknown and the more currentthrough the meter, whereas in the series exam-ple, the higher the unknown resistance the lesscurrent through the meter, because the un-known limits the current.

Consider the shunt method to be applied. Thecircuit is checked for full scale deflection.Whether full scale will be attained depends onthe voltage of the source and the resistance inthe circuit.

The Voltage of a CellIn general the voltage source is a dry cell

or battery. Actually the voltage is not knownaccurately. The commercial rating is 1.5 voltsper cell. Tests made of two popular brands ofcells showed that the voltage at no drain washigher than 1.5 volts and averaged just under1.6 volts. These tests may not be controlling asto voltages that will be supplied by cells ingeneral, but they do show that 1.5 can not beaccepted as a sure thing.

Fortunately, the actual voltage is close to therated voltage, and besides full scale deflectionalways can be controlled by resistance, eventhough the source voltage itself is beyond con-trol for any given cell or bank of cells.

The resistance is in three parts.First consider the resistance of the battery or

cell. This is not a constant. The resistance is


low when the cell or battery is fresh. With useand age the battery resistance increases. Sincesome voltage will be dropped in the resistanceof the cell or battery, the voltage reading whencurrent flows will be less as time goes on. Thisis because the reading is taken at the terminals,which now have an equivalent series resistancein circuit. In ohmmeter practice it is well toreplenish the cell or battery when the readingat 1 ma drain is 10 per cent. less than therating. This is approximately an average bat-tery resistance of 150 ohms.

Rheostat Takes Up DifferencesTo take up differences of this sort a series

rheostat is used, and it should be large enoughto enable exact full scale deflection when thecell voltage is higher than expected or rated,and also to afford reduction of resistance, thatis, work in the opposite direction, when the cellresistance rises. It would be satisfactory, forinstance, to have 100 ohms between the fullscale point established when the cell is fresh,and zero resistance terminal of the rheostat inrespect to arm, so that when one had to turnthe arm to that terminal one knew that thecell should be replenished.

The other resistors in the constructed outfitare the fixed series limiting resistor and theinternal resistance of the meter itself. It is notwise to ignore the meter's resistance, whichfor popular makes of 0-1 milliammeter isaround 30 ohms.

If the shunting method of low resistancemeasurement is used, of course a series circuitis present, too, composed of the limiting re-sistor, the cell, and in fact the effective seriesresistance of the meter -shunt circuit.

The Limiting Resistance

Suppose that a meter of 30 ohms internalresistance is being used in the series leg, thecell actually is 1.5 volts, and a limiting resistorof 1,500 ohms is used. The assumption, falsethough it be, is that all the resistance is in thelimiting leg, but the meter has 0.02 per cent. ofthe resistance of the limiter. But with 1,500ohms limiter, plus the meter resistance, total1,530 ohms, the full scale deflection current isnot established under the stated voltage condi-tion, because only 980 microamperes flow, in-stead of 1,000. Hence the sum of the meterresistance and limiting resistance should be lessthan the total required resistance by an amountsomewhat smaller than the series rheostat'stotal resistance. Since it is desired to have thisrheostat work in both directions, using, say,rheostat center as datum, then the rheostatshould have a total resistance at least twiceas great as previously mentioned.

As has been said, with the proper rheostatand smaller fixed resistor there is no difficultyin establishing full scale deflection exactly.

SMall Shunted Resistors

Taking now the example of 1,500 ohms fixed(Continued on next page)

Highly AccurateResistance Formula

This is the cir-cuit for measur-ing the unknownlow resistanceRx, when fullscale deflectioni s establishedfirst on the sen-sitive d.c. milli -ammeter, due torelationship ofthe cell or bat-tery voltage E,

and the series re-sistors R and Rh,likewise the me-ter's own resis-tance. The desig-nations coincidewith those usedin the formulaset forth below.

EFor highly accurate determination of un-

known low resistances by the meter shuntingmethod the following formula, which considersthe decreasing series resistance when metershunts decrease in resistance, applies :

Rm (Ro Rm)Rx

E

ImRo

in whichRx is the unknown resistance.Rm is the internal resistance of the meter

itself.Ro is the limiting resistor and would include

a rheostat, if used.E is the battery voltage.Im is the current through the meter alone.When the meter is checked for accurate full

scale deflection current the fact that E and Romay not be accurately known does not affectthe result, if E and Ro are taken at standardratings. The meter error in practice would belarger than the calculation error, by far.


(Continued from preceding page)limiting resistor, plus the meter resistance, total1,530 ohms for the specified conditions, and 980microamperes instead of full scale, if we putsome very small unknown resistance across themeter, the current increases almost to full scale,because the forgotten meter resistance is prac-tically eliminated, and 1,500 ohms becomessensibly right.

For instance, suppose the unknown is 0.3ohm, about the lowest ohmage readable by thesystem on usual meters, for reasons of defini-tion of current on the scale, this is almost ashort to the original 30 ohms, and the metercurrent is 996 microamperes. This would readas full scale. A difference of 0.4 per cent. cannot be closely determined on such meters, andbesides the guaranteed accuracy of the meter,applicable to full scale only, is 2 per cent.

The condition, however, does point to the factthat the shunting method changes the equivalentseries resistance.

The change is exponential. By resorting tocomputation the error is eliminated, using theseparate formula. Most persons do not desireto work a formula, preferring direct reading.In the absence of direct reading a chart mightbe prepared and consulted. The point is thatunless full scale deflection is accurately attained,the results as determined from the chart willbe somewhat inaccurate at some settings. So itis very important to establish full scale deflec-

"on the nose." Then, say, if the meterresistance is 30 ohms, the series resistanceadded, including rheostat, is 1,470 ohms, and for1.5 volts, the current is 1,000 microamperes.

Fairly Accurate Accuracy

When the unknown is so small that the meterresistance is practically cancelled out, the theo-retical full scale deflection would be 1,002microamperes. This is established by testingwith 1,440 ohms series resistance, and using the30 ohm meter resistance to total 1,470 ohms,when the case becomes practical rather thantheoretical.

Ignoring the change of series resistance in-troduced by shunting the meter the unknownmay be determined on the basis of this simpleformula :

Rm ImRx

Imax-ImWhere Rx is the unknown, Rm is the resistanceof the meter alone, Im is the current throughthe meter alone, and Imax is full scale deflec-tion current.

Taking the full scale as 1 milliampere andusing its equivalent in microamperes, we mighttest for a few unknowns, say, for (a) 15; (b)30; (c) 50; (d) 100; and (e) 500 microam-peres.

Rx = 30 X 15/985 = 0.45 ohm -(a)Rx = 30 X 30/970 = 0.93 ohm -(b)Rx = 30 X 50/950 = 1.575 ohms-(c)Rx = 30 X 100/900 = 3.33 ohms-(d)Rx = 30 X 500/500 = 30 ohms-(e)

It can be seen that the resistance is notlinearly related to the current read, simply byobserving that the 50 microampere exampleyielded 1.575 ohms, and the 500 microampereexample yielded 30 ohms, not 10 x 1.575 or 15.75ohms. Also, when the shunt resistance exactlyequals the meter resistance, the meter readshalf scale. This fact is used as a basis of ob-taining a shunt resistance, then measuring theresistance of the shunt externally in a highcurrent circuit, the answer being the resistanceof the meter, for it is the same as the resist-ance of the shunt. This was done and shown indetail in the May issue.

Higher Accuracy

If it is desired to have still greater accuracy,although it is a fine point and will not proveof much value to most experimenters, the rheo-stat in the series leg may be set for some defi-nite value required by the cell, etc., for fullscale exactitude, and resistance differences cali-brated on a scale to be affixed to the pointer ofthe knob. These differences would representohmage from the prescribed full scale setting.

Linear Rheostat Used

A rheostat with linear characteristic wouldbe necessary. The scale would have to berotatable, so as to follow the so-called zerosetting, which in respect to the rheostat wouldchange from period to period. The calibrationcould be in steps of 5 ohms from reference point(taken as 0), to 30 ohms, if that is the meterresistance. Then when any very small resist-ance value is being tested, if the reading is intenths of an ohm the rheostat is moved tohigher ohmage to include 30 ohms more, as ifthe meter had been shorted by the unknown.For 1 to 3 ohms the knob could be turned toread at the estimated equal distance between0 and 5 ohms.

As a comparison, the more accurate methodas derived from the separate formula gives (a)as 0.456 ohm and (d) as 3.324 ohms, comparedto 0.45 and 3.33 ohms. For readings from halfscale to higher currents the correction factorbecomes smaller and may be ignored, i. e., thesimpler formula above the tabulation is ample.

Glad to Answer Questionson Who Makes the Parts

IF you want to know who makes anyradio product shown or mentioned in

these columns, a quick reply is yours forthe asking. Address Trade Editor, RADIOWORLD, 145 West 45th St., N. Y. City.


UNDERWRITERS' RULES FOR TRANSMITTERSThe following paragraphs apply to amateur trans-

mitting stations.a. Antenna and counterpoise conductors out-

side buildings shall be kept well away fromall electric light or power wires of any circuitof more than 600 volts, and from railway,trolley or feeder wires, so as to avoid the pos-sibility of contact between the antenna or coun-terpoise and such wires under accidental con-ditions. Antenna and counterpoise conductorswhere placed in proximity to electric light orpower wires of less than 600 volts, or signalwires, shall be constructed and installed in astrong and durable manner, and shall be solocated and provided with suitable clearancesas to prevent accidental contact with such wiresby sagging or swinging.

b. Antenna conductor sizes shall be not lessthan given in the following table :

Material

Stations to whichpower supplied isless than 100watts and wherevoltage of poweris less than 400

volts

Stations to whichpower supplied ismore than 100watts or voltageof power is more

than 400 voltsMaterial

Soft copperMedium - drawn

copperHard -drawn cop-

perBronze or copper

clad steel

14

14

14

14

7

8

10

12

c. Splices and joints in the antenna andcounterpoise span shall be soldered unless madewith approved splicing devices.

d. Lead-in conductors shall be of copper,bronze, approved copper -clad steel or othermetal which will not corrode excessively andin no case shall be smaller than No. 14.

e. Antenna and counterpoise conductors andwires leading therefrom to ground switch,where attached to buildings, shall be firmlymounted five inches clear of the surface of thebuilding, on non -absorptive insulating supportssuch as treated pins or brackets, equipped withinsulators having not less than five inches creep-age and air -gap distance to inflammable or con-ducting material, except that the creepage andair -gap distance for continuous -wave sets of1,000 watts and less input to the transmitter,shall be not less than 3 inches.

f. In passing the antenna or counterpoiselead-in into the building, a tube slanting upwardtoward the inside, or a bushing of non -absorp-tive insulating material shall be used, and shallbe so insulated as to have a creepage and air -gap distance in the case of continuous wavesets of 1,000 watts and less input to the trans-mitter, not less than three inches, and in allother cases not less than five inches. Fragileinsulators shall be protected where exposed tomechanical injury. A drilled window panemay be used in place of a bushing provided thecreepage and air -gap distance, as specified above,are maintained.

g. Adequate lighting protection either in theform of a grounding switch or suitable light-ning arrester shall be provided. The groundingconductor for such protection shall be at leastas large as the lead-in and in no case smallerthan No. 14 copper, bronze, or approved cop-per -clad steel. The protective grounding con-ductor need not have an insulating covering orbe mounted on insulating supports. The pro-tective grounding conductor shall be run in asstraight a line as possible to a good permanentground suitable for the purpose. The protectivegrounding conductor shall be protected whereexposed to mechanical injury.

h. The operating grounding conductor whereused shall be of copper strip not less thaninch wide by 1/32 inch thick, or of copperbronze, or approved copper -clad steel having aperiphery, or girth, of at least g inch, such asa No. 2 wire, and shall be firmly secured inplace throughout its length.

i. The operating grounding conductor shallbe bonded to a good permanent ground. Pref-erence shall be given to water piping. Otherpermissible grounds are grounded steel framesof buildings or other grounded metal work inthe building, and artificial grounding devicessuch as driven pipes, rods, plates, cones, etc.Gas piping shall not be used for the ground.

j. The transmitter shall be inclosed in a metalframe, or grill, or separated from operatingspace by a barrier or other equivalent means,all metallic parts of which are effectually con-nected to ground.

k. All external metallic handles and controlsaccessible to the operating personnel shall beeffectually grounded.

No circuit in excess of 150 volts should have anyparts exposed to direct contact. A complete dead -front type of switchboard is preferred.

1. All access doors shall be provided with in-terlocks which will disconnect all voltages inexcess of 750 volts when any access door isopened.

m. Under the conditions noted in paragraphs1, 2 and 3 below, wiring may be grouped in thesame conduit armored cable, electrical metallictubing, metal raceway, pull -box, junction -boxor cabinet.

1. Power -supply wires are introduced solelyfor supplying power to the equipment to whichthe other wires are connected.

2. Wires other than power -supply wires runin conduit, armored cable, electrical metallictubing, metal raceway, pull -box, junction -boxor cabinet with power -supply wires are in-sulated individually or collectively in groups byinsulation at least equivalent to that on thepower -supply wires or the power and otherwires are separated by a lead sheath or othercontinuous metallic covering.

3. Terminals for the power supply conduc-tors and the other conductors provide spacingsof all other terminals from the power terminalsat least as great as the spacing between powerterminals of opposite polarities and also suitablemeans are provided to guard against connectingother conductors to the power supply terminals.


THE 6B5

B- B

CO Most Interesting TubeSince the Variable Mu

By K. N. Satterwhite

The 605 in a single sided circuit is shown, with theelements identified by subscripts. The input triodeconsists of PI, GI and KI. It can be seen KI is

not accessible externally. The output triode con-sists of P2, G2 and K2. P, G and K refer re-spectively to plate, grid and cathode. The dia-gram does not show heater. Refer to the under-neath view of socket. A load resistor or trans-former secondary may connect to binding posts.

THEmost interesting tube to come out since

the variable mu tube appeared a few yearsago is the 6B5, consisting of two triodes in oneenvelope, one of which is a signal voltage loadsupply tube, the other the output tube. It ispractical to use two of the tubes in push pulland obtain large power output under circum-stances of unusually low distortion.

One of the fascinating facts about the tubeis that the input load on the first tube servesas grid to cathode impedance of the secondtube. This is made possible by tying the gridof the second tube to cathode of the first tube,done in the manufacturing process. Thus directcoupling prevails. The manufacturing processalso establishes the correct phase relationships,so that the tubes work together, and their entireperformance should be regarded as that of asingle tube, although for technical dissectionit is necessary to consider the tubes separately.The intimacy of one triode to the other makesit difficult to consider the final result as that oftwo separate tubes, especially as a strikinglylinear relationship is established because of theunified variation.

Grid Runs PositiveIn fact, the curves show that the tube is

closer to a diode detector, as to linearity, thanany other amplifier tube, so that harmonics dueto curvature of the characteristic are almostabsent, and such harmonics as are present then

For push pull this circuit may be used, if only 300volts are applied. The power output then wouldbe 15 watts. For 20 watts output the B voltageshould be raised to 400 volts, and a bias applied.Either a 140 ohm resistor or a 12 -volt dry batterywould be connected between common cathodeand the joint of the ground lead, battery oliptto cathode. Thus the 6B5 gets into the 2A3 class

for output capabilities

would be due largely. to effects of output loads(for instance, speakers), and working the tubesfar beyond the intended limits.

Another interesting point about the 6B5 isthat it runs a positive grid. This formerly wastaken somewhat in die sense of a patient run-ning a high temperature, but when Class Bamplifiers of consequence came along, equippedwith special purpose tubes, operated at zerobias at no signal input, and running grids in theonly direction possible, the world became curedof positive grid phobia. It is therefore preparedto hear of a tube that not only runs a positivegrid, but has a straighter characteristic on thepositive than on the negative slope, so isworked as Class B would be, in usual practice,at zero operating bias, and with 300 volts onthe plate, too.

What Class? A, B, AB or What?

It is not intended to say that the operationis that of Class B. It is Class-well, it is hardto say just what the class is. What shall bethe criterion? During what part of the cyclethe tube is functioning?

Taking efficiency as a basis, the new tube isequal to a pentode.

Taking power output capabilities as the stand-ard, the performance of the new tube is equalto Class B.



CHARACTERISTICS OF 6B5

Heater Voltage (a.c. or d.c.) 6.3 voltsHeater Current 0 8 ampereBulb ST -14Base Medium 6 pin

Single Class A AmplifierInput Output

Plate Voltage 300 volts 300 volt.;Grid Voltage 0 0Plate Current 6 ma 45 maPlate Resistance 24,100 ohmsAmplification Factor 58Mutual Conductance 2,400 micromhosLoad Resistance 7,000 ohmsTotal Harmonic

DistortionPower Output

5 per cent.4 watts

Push Pull Class A Amplifier

Plate InputPlate Voltage 400 voltsGrid Voltage 13 voltsPlate Current 4.5 maLoad ResistanceTotal Harmonic

DistortionPower Output

Output400 volts

13 volts40 ma

10,000 ohms

5 per cent.20 watts

Portable

6B5Amplifier

The photographs show theconstructional factors of asmall portable amplifier,the diagram revealing thecircuit, with the new 6B5combination tube. As verysmall impulses may have tobe amplified at times, a

driver is used in this in-stance, although in re-ceiver practice it would notbe strictly necessary toprovide a driver for thenew tube. At right, theflap is to be pulled upwhen speaker behind it is

in operation.

The socket connections as seen from the bottom.The symbols correspond to those shown on thediagram of the single sided Class A amplifier.Just where the heater is to be grounded will de-pend on the type of circuit used. For a.c. opera-tion the center of the heater winding is usually

grounded.

(Continued from preceding page)

Taking performance as a comparison the newtube ranks Class AB.

Taking distortion level, the tube is compara-ble to Class A.

From the foregoing it may be assumed thatthe strictly technical classifications are put asidefor the moment, efficiency, power, performanceand distortion being considered, that is, theeffects of the different classes, rather than thecyclic operation that determines the A, B, ABclassification.

Opportunity Exists for Stable OscillatorSince the 6B5 tube is one that may be op-

erated at zero bias, resulting in practically astraight line on either side, that is, for negativeor positive pulses applied to the grid, the tubemay be used as an otherwise unbiased oscilla-

tor, with promise of frequency stability. This istrue because if the amplitude is held low theplate current change will be small and linear,both considerations fully consistent with stabil-ity.

k A I) 1 U \V U R L I) September, 1935

Utilizing an Old ConsoleTuner, Power Amplifier and Speaker Installation

By Herbert ErwinIT was decided to install a high fidelity tuned

radio frequency tuner and amplifier, to-gether with its dynamic speaker, in a consolecabinet that had previously been used to house

SIR

A tuner and separate power supply -amplifier werebuilt to be fitted into a console the author had.The first view is of the side of the amplifier, withthe tubes (57, 2A5 and 80) removed. Next isshown the bottom view. Note how the mediumcapacity Cornell-Dubilier condensers are neatlyarranged to permit of splendid symmetry of wiring.Bottom picture shows tubes in place on amplifier

chassis size 10 x 4 x 1.5 inches.

7a standard set. Since holes had been left in thecabinet by the previous fittings, such as pro-truding metal shafts and dial openings (in thiscase the opening was rectangular) it was de-sirable to construct the new unit so that itcould match up with these holes as nearly aspossible.

This was accomplished by placing a sheet ofcardboard across the face of the cabinet, cover-ing all openings, and with a lead pencil, mark-ing these positions on the cardboard, includingthe dial opening, then transferring this patternto the chassis, and moving the controls over tofit on the same holes. The airplane dial, beingcircular, was centered over the section formerlyoccupied by the rectangular opening, the card-board pattern then being transferred to the cab-inet again, and the surplus wood cut away.

Shelf Inside CabinetAll that remained was to build a little shelf

inside the cabinet to hold the tuning unit, whichwas then slid forward and the controls shovedthrough the holes in the cabinet. The amplifieris connected through a two wire cable whichis sheathed with a metallic covering andgrounded to prevent pickup. The amplifierrests on the floor of the cabinet alongside theRola G-12 speaker.

The combination of the three circuits, highfidelity tuner, amplifier and high fiedity speaker,result in extremely satisfactory performance,although some will prefer a superheterodynefor the tuning if better selectivity is needed.For all practical purposes and actual opera-tion, the combination described is very satis-factory.

It is useless, however, to design a flat tophigh fidelity tuning clement that delivers afaultless signal into an amplifier that is poorlydesigned, and in turn brings forth a jitterysignal, or to put it different, it cannot be ex-pected that a good tuner and a good amplifierwill deliver a final pleasing sound through aloudspeaker that is constructionally bad or elec-trically imperfect.

The speaker can only reproduce the energythat is delivered into it, and no matter hownearly perfect this signal may be if the speakerwill not reproduce it, if the speaker has nutsufficient signal scale response, then the re-sultant output in sound waves can scarcely beconsidered high fidelity.

Three ConsiderationsThree things therefore must be correlated:

the tuner, the amplifier and the speaker. Poweris a later consideration.

A set designed for average home use doesnot need 15 watts output. This has 4 watts.

September, 1935 RADIO WORLD - 27

The two views at left are of the receiver side and bottom. At top right is a wooden caster, withrubber bed just below the polished metal cap at top (the caster is shown upside down). The reasonfor replacing the simple metal caster (lower left of same picture) was that tore was improved, theconsole being freed of certain jarring effects due to acoustical resonance because of the liveness ofthe sprung rubber. Immediately below are shown the a.c. cable outlet for connecting to wall socket,(right) and the lead from amplifier to set with male plug attached, so the set switch also turns the

amplifier on and off.

A template was prepared from cardboard as shown at left, so that the central and airplane dialpositions could be closely located, for communication of markings to the consc e front, preparatoryto drilling. Center view shows templates on set. At right is the finished installat,on, front view.


Filter Design and ApplicationSimple Circuits for Audio Systems


FIG. I

A filter diagram illustrating T section division between dottedlines.

FILTERS are electrical devices that showvaried "shut -out" discrepancy to different

frequencies. In other words, filters changetheir impedance to different frequencies. Byutilizing capacity, inductance and resistance invarious circuit combinations it is possible tovary the amount of suppression of any fre-quency group. By combining a number of simi-lar sets of filters much sharper and more ex-act results may be obtained.

How Filters Are ClassifiedThe use of filters in radio and allied fields

is great and varied. All tuning units are oneform of filter. Bypass condensers across biasresistors, detector radio frequency chokes, andpower chokes and condensers are only a fewmore examples of filters found in radio.

In one manner filters may be divided intofour categories depending upon the functionthey are called to perform. Each one of thesefour classes will be considered as to circuit de-tails, application, and design.

Filters may be low pass, high pass, bandpass, and band elimination relative to the fre-quencies passed or attenuated. Each of thesefilters may be divided into a number of eitherT or r sections. Which division is chosen isimmaterial, for there is a simple group of

FIG. 2A filter diagram illustrating pi section

division between dotted line.

mathematical formulas that enables interchangeat any single frequency of a three element Tfilter with a three element r filter, and viceversa.

The reason that one of these divisions is usedis because the analysis and calculations aregreatly simplified. For example, with the aidof this interchangeable theory, any networkcomposed of linear impedances for any givenfrequency can be simplified to a single T orw section. If we begin with a 7 section it canbe reduced to a T section; two T sections canbe reduced to a r section with extra arms ;and this r section can be further reduced to aT and the arms are added on. In this mannercomplicated networks can be greatly simplified.

The Low Pass Filter

A low pass filter with no resistance lossesin its elements will pass, without attenuation,currents of all frequencies from zero to thecutoff value and will greatly attenuate fre-quencies beyond this value. Of course resistanceis always present in the inductance or con-densers that go to make up the filter. Thisresistance has the effect of making the cutoffpoint less sharp; i.e. frequencies near the cut -


Barkhausen Theory Undergoing RevisionWhen Barkhausen and Kurz announced that

they had made conventional tubes oscillate oncentimeter waves, by making the grid positive,plate either zero or slightly negative, and tun-ing by altering the voltages of the supply,their discovery was hailed as representing anew form of generator. It was considered an"electron" oscillator, different from the usualkind. The circuit as devised by the two Ger-mans was modified by two Englishmen, Gill andMorrell, who used tank circuits. The Bark-hausen oscillator then began to look theoretic-ally somewhat more like any other oscillator.

Now F. B. Llewellyn, noted oscillator author-

ity, writes in the Bell Telephone "Record" thatthe Barkhausen is in fact much like any otheroscillator, and he details the electron path andactivity on which his conclusion is based.Moreover, in other quarters the theory thatthe distance between elements limits the fre-quency, due to the transmit time of the electronmovement between these parameters confiningthe upper frequency, is under fire, if not alreadydisproved. And the limits of oscillationare under other forms of iconoclastic study,so that the early theory of the Barkhausencircuit has been partly upset and is due to meetfurther amendment.


off point but below it are attenuated to a lim-ited extent, and the frequencies slightly abovethe cutoff point are passed with relative free-dom. For this reason the units of a highgrade filter should always possess the least re-sistance possible. For the cutoff frequency

=z-VLC

where L is the inductance in henries and C isthe capacity in farads, of the low pass filtershown in Fig. 3, both as a T and r types.

If the input and output impedances areknown and are equal to Z, the inductance andcapacity of the filter may be computed for thecutoff frequency f:

L = -f

1

f Z

Reason for More SectionsIf need arises to cutoff all frequencies above

1,000 cycles, where Z is 500 ohms, these valuessubstituted in the equations above will give :C = 0.64 mfd., and, L = .159 henries. Com-mercial units close to these values are obtain-able.

If a certain frequency is suppressed a certainamount by a single section, two such sectionswill suppress exactly twice this amount. There-

' fore by utilizing a number of sections any de-gree of attenuation may be attained. By re-ducing the circuit resistance to a low figure,the same results may he obtained with fewersections.

High Pass FilterA high pass filter freely passes all frequencies

above the critical or cutoff frequency andgreatly attenuates the currents below this fre-quency. One high pass filter is shown in Fig.4, -both T and r sections. The capacity C andinductance L are computed from the formulas :

L = henries4 n fo

1

C = farads4ir10Z

Band Pass Filter

A band pass filter passes a given band offrequencies and attenuates those above and be-low this band. The object in designing a goodband pass filter is to have little loss in the bandto be passed, high attenuation in all outside fre-quencies, and a sharp turn from band pass toattenuation regions. These objectives can beaccomplished by reducing the d.c. resistance toa minimum, and using a sufficient number ofwell designed units.

From the discussion of a high pass and a

FIG. 3Low pass filter with a cutoff frequency fo, show.ing attenuation variations with rise in frequency.

low pass filter it becomes evident that if thetwo are combined and if they have cutoff fre-quencies that differ, the band that is passed willbe between the two cutoff frequencies. Forexample, if a low pass filter will attenuate allfrequencies above 5,000 cycles, and a high passfilter will pass frequencies above 4,000 cycles,then if the two are combined only the bandbetween 4,000 and 5,000 cycles will be passed.

In practice, however, the two filter sectionsare combined into one structure as illustratedin Fig. 5. A more complex structure is pos-sible and is used for some special applications.But for most ordinary needs of a band passfilter the type illustrated would serve the needexcellently. At some group of frequencies,corresponding to the band to be passed, CiLIcombination will be resonant and CIL2 com-bination anti -resonant, so that CiLi will offernegligible resistance and C2La will offer in-finite resistance, theoretically. The values ofthe condensers and inductances may be com-puted from the formulas given below, wheref' is the lower limit of the band to be passed,and f" is the upper limit.

=(f"-f')

(f" f')

4 7r f' f" Z

(f"-f') ZL2 =

4 r f' f"

henries

farads

henries

1

C2 = faradsr (f" - f') Z

Band Elimination FilterA filter of the band elimination type is de-

signed to pass freely currents of all frequen-cies except those within a given band. A struc-ture such as illustrated in Fig. 6 is commonly



used. The curve in Fig. 7 shows the charac-teristics of this type filter. This type of filteris quite efficient for ordinary application and isquite simple in design. Fig. 6, of course, il-lustrates a single section of an L type filter

IT

FIG. 4A T and pi type high pass

filter.

incorporate in the phonograph line a band elim-ination filter that will eliminate the band slight-ly above and below 4,000 cycles. Let f' = 3,900cycles and f" = 4,100 cycles. Substitutingthese figures in the formulas above we get the

FIG. 5A T type band pass filter.

for this application. The values of the dif-ferent parts used may be computed from theformulas given below, where f' is the lowerfrequency of the band to be eliminated, and f"is the upper limit of the band, since there areonly two limits to be considered for suchsimple. filters.

LI =

=

L2 =

(f"-f') Zft ft,

1

(f"-f') Z

47* (f"-f')

f"-f'C2 =

f" f' Z

henries

farads

henries

farads

Application of Scratch Filter

For example, we may consider the need offiltering out the needle noise of a phonographpick-up. Usually this noise is of one frequency,about 4,000 cycles. In other words, all that isneeded to eliminate this undesirable noise is to

FIG. 6.Band elimination filter. Single L

section.

following values (assuming input and load im-pedance 17,000 ohms)

Li = 0.068 henriesCi = 0.023 mfd.L, = 6.8 henriesC2 = 0.0025 mfd.

Commercial units close to these sizes will ofcourse serve the purpose.

In certain engineering procedure need arisesfor more specialized filters, but the types de-scribed should prove satisfactory for the usualradio and public address practice.

FIG. 7Characteristics of a band elimination filter. Itis intended to eliminate the frequencies between

f' and f".

Clow Indication of Zero BeatUnder some circumstances it is advisable to

get exactly zero beat, or at least closer to zerothan by the usual ear test, and in that instancea visual index fills the need. One oscillatorbeats with another to produce the audio fromthe two radio sources. Therefore if the outputsare united and put into another tube-a de-tector-the output of the detector can be suf-ficient to actuate an indicating device, such asa regular d.c. milliammeter or the new ray in-dicating tube, 6E5. The oscillator being varied

is brought as close to zero as practical, mean-ing the needle will stand still, or indicatingtube stay steadily lit or out, or movement willbe so slow that the number of cycles a seconacan be counted accurately against the secondhand of a synchronous motor clock. Then even"how far off from zero" is known, but not onwhich side. The "side" may be determined byslightly turning the oscillator dial. If themovement increases the former "offness" is tobe subtracted from the true frequency.


Small Rectifiers for B SupplyHandy for Testing and for Powering Small Sets

By Jack Tully

37

Substitution of the IV rectifier tube for doublingthe current drain. The d.c. output voltage is

unchanged, about 105-125 volts.

OFTEN in experimental work the needarises for a source of d.c. voltage up to

around 100 volts or so, and this may be ob-tained from a small rectifier. A diagramshows such a rectifier that has been in usefor several weeks, the appearance being re-vealed in the photographs. In that time onscores of occasions the rectifier has been putto handy service, for supplying the d.c. voltageto small high frequency generating circuits andfor making meter shunts and multipliers andmeasuring resistance. Resistances 2 to 10meg. have been measured, a range not cov-ered by a battery operated ohmmeter on hand.

The circuit shows the use of a 37 tube. Theother tube that may be used in the same socketwithout any change is the 76. The grid is tiedto the plate for this type of rectifier, that is,where the current is relatively high, comparedwith detecting diode use, for up to 50 ma maybe drawn on short occasions, although around25 milliamperes steady drain is about the safelimit, consistent with long tube life.

Precautions Against ShortingIf 50 ma are to be drawn for regular service,

that is, if it is desired approximately todouble the safe current drain, instead of afive hole socket a four hole socket is used, andthe IV rectifier tube is inserted instead of the37 or 76. A diagram shows the connections.

The line cord has a third lead built in, andthis third lead contains the limiting resistor.This lead has to be so placed that during thatalternation when the plate is positive to a.c.the cathode is negative to a.c. and full a.c. isapplied to the plate. During the other alter-nation, since the plate is then negative, thereis no conduction, as these are half wave recti-fiers. Hence the a.c. has to appear betweencathode return and plate, so cathode is neces-sarily connected to the line to provide a return

/V

1 am, lasa

A 37 or 76 tube may be used as rectifier. Thetwo electrolytic condensers, 175 volt peak rating,

are in one cardboard container.

circuit, and any grounded line presents thedanger of a short if connected to a groundedsystem, since "hot" side of the line could beconnected to the grounded side through suchexternal connection. This is a zero potentialdifference or short, hence the line fuse is in-cluded for safety. This may be an automobilecartridge type, supported on an insulated cliptype holder. There are stamped end piecesobtainable which, when insulated from thechassis, permit safe insertion of the fuse. Thechassis itself, if of metal, is never connectedto any part of the circuit.

Usually the "hot" side of the line can bedistinguished by touch. It can be felt as atingle, whereas the other side can not. Nega-tive d.c. output post may be used for the test.

How to Test Line CordThe line cord just mentioned may not follow

familiar color coding. It is common to expectthat black represents ground, red representsplus, and white would be the third connection.The cord used had the resistor tied at one endto the black cord side of the plug. Thereforeat this plug prong black and white (the re-sistor) were common, so to follow the circuitconnect red to cathode and one heater side,black to plate and white to other heater.

The line cord should be given a continuitytest to confirm the correctness of intended con-nections. Laying the cord on the table, con-nect an indicator between the two plug prongs.No reading should obtain. Connect indicatorbetween two cord terminals other than the oneobviously representing the resistor lead, whichis usually white. No reading should obtain.Between one of these leads and the resistorwire no reading should result, while betweenresistor and the other free terminal wire therewill be a reading. The two cord terminals


32 RADIO WORLD September, 1935 1

(Continued from preceding page)that show this reading are to be connected asfollows: resistor lead to either side of heater,other wire to plate. Cathode is to be con-nected to the remaining lead. This was redin the cord used in the rectifier.

Low Efficiency Yet Low Cost

The two diagrams for transformerless opera-tion represent what must be admitted to be anefficient method, since to work the simple rec-tifier around 40 watts power dissipation is re-quired. This would be enough to operate a

Max.*

Min. *

six tube receiver. The trouble lies in thelimiting resistor, where nearly all the poweris spent, mostly in the form of heat for whichthere is no use. Since the resistor wire is inthe a.c. cord covering, the cord will get warm,but not hot. No danger attaches to the factthat this device warms up.

Of course the 40 watt power consumption isnot financially ruinous, simply representing apoor efficiency, about 0.16 per cent. outputcompared to 100 per cent. input, at full work-ing output current, 50 ma, but in cost, at 10cper kilowatt hour, and the rate is less incities, this is only 0.4 cents an hour.

Use of Transformer

A single voltage output is obtained, as thed.c. circuit is brought out to two binding posts,but if a more elaborate rectifier is to be built,it may be done as shown in another diagram,where a small power transformer is included,having 115 volt primary, 350 volt secondary(or lower voltage secondary) and a heaterwinding, 6.3 volts. Not only is the maximumd.c. output voltage higher, which also requireshigher rating filter condensers, but the filtra-tion is made better by inclusion of the choke,Ch, and the d.c. output is freed from dangerof short due to connection of external groundedcircuit to ungrounded side of the line. Hencethe first two diagrams represent a use forohmmeter, shunt, multiplier and other similarpurposes, without danger, but for wider ap-plication the third diagram is better. Also,

the current drain may be safely 50 ma at alltimes, and on occasion may be raised to 100 mafor brief intervals, for some special test re-quirement.

Choice of VoltagesWhile the earlier circuits may be used for

voltages less than maximum by introducing aseries resistor, say a rheostat of a few hundredthousand ohms, and the voltage measured atthe test circuit, the more elaborate rectifiermight well have adjustable resistors built in..One fixed resistor should be included, markedR, so that .under no circumstances could the

amp.fuse

Small rectifier, using atransformer. In this waythe voltage output isgreater, depending on howmuch stepup ratio existsbetween primary at rightand high voltage second-ary; the shorting danger isremoved, because of iso-lated grounding; the B

choke, which may be anaudio transformer winding.improves filtration; and therheostats afford ready volt-

age selection.

rectifier output be completely shorted to d.c.,hence R may be 2,000 ohms, 5 watts. ThenR2 and R3 may be two equal rheostats of afew hundred thousand ohms each, or, if un-equal, R3 may be the larger. The resistors'bypass condensers should all be rated at maxi-mum B voltage, which is about the same as thea.c. voltage on the secondary feeding the plate..

Follow Line Voltage

The transformerless rectifier circuits shownherewith give a d.c. voltage output at 10 mil-liamperes drain or less that is equal approxi-mately to the r.m.s. line voltage, and so maybe used with d.c. meters for the a.c. measure-ment. If the drain is substantially increasedabove 10 ma this condition no longer holds.

All Circuits We Print CanBe Readily Duplicated

ALL parts for circuits described inRADIO WORLD constructionally are

obtainable. Most of them stocked byregular supply sources. We are alwaysglad to identify the trade names of anyparts, and tell where any parts arestocked. Address questions to Trade Edi-tor, RADIO WORLD, 145 West Forty-fifthStreet, New York.

34 R A 1) I O WORLD September, 1935 September, 1935 RADIO WORLD 35

A HIGH FIDELITY POWER AMPLIFIERFOR my personal use

I have constructed theamplifier and power sup-ply illustrated in thephotographs, using thebest parts and standardcircuiting, so that with afaithful reproducer theresult is high fidelity.

Since it may be desir-able to use more thanone speaker for com-mercial use of such anamplifier, provision ismade for powering suchextra dynamics, the neg-ative bias for the ampli-fier output tubes beingadjusted to meet the con-ditions imposed by theextra current drain of anadditional speaker on the80 rectifier circuit Thisbias may be made of suchvalue as to serve Class.1, Class AB or Class Bpurposes. It is derivedfrom a C supply rectifier.

Meters DescribedThe front panel of the

amplifier shows also themeters for the mixingcircuit, which meterswere included because onhand, not being reallynecessary. These are the0-4 voltmeter and the0-25 milliammeter, asshown in the diagram ofthe mixer on the nextpage. The three othermeters are two 0-100milliammeters and a d.c.voltmeter, the latter notin the diagram on page36.

There are five tubes inthe amplifier, a single 56input, push pull 56drivers and push pull2A3 output tubes. Variousoutput tubes have beenused, and if the filamentvoltage and plate volt-ages are suitably selected,the substitution may bemade quickly. The out-put transformer is tappedfor three voice coil im-

Provision for Multiple SpeakersBy ROY L. SMITH

and Fixed Bias

C

The mixer -amplifier is shown at

pedances, 4, 8 and 15 ohms, with 500 ohms as the maximum, for useif desired with a transmission line, a popular impedance for which is500 ohms.

The power supply has two tubes, an 80 for the C bias voltage anda 5Z3 for the B voltage. The photographs at right on this pageshow three sockets that are not used, as the particular chassis hap-pened to be on hand.

The mixer provides for input from a two button carbon micro -

left in three viewstwo views above.

phone, or a high impedance phonograph pickup. If a low im-pedance type pickup is to be used, then, unless the impedance isclose enough to that of the primary intended for the microphone,a special matching transformer would be needed, and the pickupmanufacturers have the proper coupling coil for this purpose. Thespeech or music, or both, may be faded out at will.

The amplifier gain is not sufficient for a crystal microphone,although one may he connected across the grid load of an input

and the dual power supply in

tube without any match-ing devices.

Since the power supplyis rather elaborate, it iswell to point out againthe load across the out-put of the 280 will varyaccording to the numberof speakers used, so the20,000 ohms bleeder(marked "10,000 ohmseach") is tapped to pro-vide C bias of such valueas never to exceed 100volts, preferably not toexceed 80 volts.

One Battery UsedThe small potential

required for carbon mi-crophone excitation isobtained Fronk a C battery, all other voltagesbeing taken from thepower supply.

Diagrams of the mixer.amplifier and power sup-ply are printed on thenext page. There is a

mfd. condenser acrossthe 1,250 ohm biasingresistor in the push pull56 stage. Ordinarily nocondenser is required.because with true bal-

ance there is no signalacross the resistor.However, in service, ifa 56 should get "sour"or go dead, the presenceof the condenser per-mits continued opera:tion without immediatetube replacement, andwith excellent tone pre-served.

For the output stage,separate bias is appliedto each tube, so thatstatic balance can bereadily accomplished.This is of particular im-portance when 2A3tubes are used, becausethey have such a terri-fically high mutual con-ductance that s smallgrid voltage differencemakes a large plate cur-rent difference.

September, 1.935 RADIO WORLD 33

12 Tube Super ExperiencesConstructional Guidance from Designer of a Precise Set

By Samuel Miller

THE12 tube all wave receiver on which I

have been working for more than threemonths was developed along standard lines,but a few specialties had to be introduced toimprove the performance, and where involv-ing constants these are shown. They consistedprincipally of the following:

(1)-Use of minimum bias on the first tubeon the highest frequency band, to reduce noiseand raise sensitivity.

(2)-Medium voltage on the screens of theoutput tubes, to improve the tone, bringing thefunctioning nearer to that of triodes.

Doing Better with the Trimmers(3)-Use of trimmers across secondaries of

selector tuning coils far from their full capacity,e.g., for 35 mmfd. condensers less than 15 mmfd.is ever used, as thus the tension is less, andthe compression trimmers then will stay putquite well.

(4)-Use of a separate oscillator, and coup-ling to modulator made inductive so that incoil design coupling can be controlled and thusaccommodated to the frequencies concerned, thecoupling being different for all bands.

(5)-Use of a light limiting resistor in theoscillator plate leg, bypassed by 0.1 mfd., sothat the frequency stability of the tube isgreatly improved, due mainly to the resistor.

The Oscillator Grid Condenser(6)-Use of a small grid condenser capacity

in the oscillator, to avoid an objectionable timeconstant that otherwise might introduce highpitched modulation (scream) at high radio fre-quencies, yet enabling maintenance of a fairlyhigh leak value, 0.1 meg., for stability aid,sharpness, and good bias control of the oscil-lator.

(7)-Use of semi -fixed bias under audio re-generation conditions elsewhere in the circuitthat dispense with the need of a bypass ca-pacity across the biasing section of the voltagedivider (200 ohms).

(8)-Separate tube for working the tuningmeter, so that closeness of indicating resonanceis independent of the setting of the controlsother than the tuning control.

Where the Peaks Are HighTo the circuit output is connected a high

fidelity speaker. The power output is broughtto considerably higher level than usual by pur-poseful increase of the B voltage and bias.

Therefore two of the usual 500 volt typeelectrolytic condensers are connected in series,

with protecting resistor across each, instead ofcondensers in parallel, or simply one condensernext to the rectifier, where the peaks run high.The circuit can be worked at 15 watts at 7 percent. total harmonic distortion, or at about 10

watts, 5 per cent. total harmonic distortion.The construction of such a receiver is a very

considerable undertaking, Save for the X handcoils (low frequencies) all were wound byhand, the X coils being universal wound (hon-eycombs). The range is from 140 kc to 30 mgc.in five steps, and the dial is frequency cali-brated. The coincidence of dial reading toactual frequency received is always better than1 per cent.

The coil assembly is very interesting and itis planned to present details of the structurein the October issue.

Design for 6B5Portable Amplifier

The amplifier described herewith has an out-put of about 4 watts with one single 6B5 tube,which is plenty of power. Better quality canbe obtained with other and more expensivetypes of amplifiers, but this is another story,since the purpose of this construction was tobuild an entire circuit, all in one package thatcould reasonably be called good fidelity, andat low cost. For adherence to the closer re-quirements of high fidelity a triode here is usedas output.

The 2.5 voltwinding maybe used forheater of thedriver, if the6C6 is to bereplaced by a

57. The 6B5heater voltage

is 6.3 volts.

Re/mid. 685


[These diagrams illustrate article on preceding two pages]The author used a

double button car-bon microphone,and to excite it a

4.5 volt C battery.The meters shownto right were in-cluded largely fordecorative purposes(he had them onhand) and may be

omitted.

.56

Z5nrid.25V

844/,000.,, 4.5a

Mike 0-Z5 Ma

Man Impedancemagnetic /nckap

56

Z5 mid

E"v

,7.,,aT 400 Y.

50.250V

The lower coil witharrow through itrepresents a highimpedance phono-graph pickup. Thedouble potentiome-

i- ter system repre-sents a T pad, whichpresents practically

c_ a constant imp9-4! dance to the cir-

cuits it connects.

40 Ma.- ZA530 Ma, - 45

45-

ZO mid100 Y. %

/0,000n each

40 Ma. -2A550 Ma. -45

ZO mid/00 V

C- 5* 300Y

500n

15.n.

4 .n.

A 25 mfd. condenser is connected across the biasing resistor of the push pull 56 stage so that if oneof the tubes goes dead the circuit will work well. Otherwise the condenser could be omitted. This

is the diagram of the amplifier.

SZ3 250 Ma. 150 Ma. 60 Ma.

f32.E? (32Bf\

+

mild el In fel450V. 450V.

°midZOO Y.

60 Ma

midZOO Y.

60 Ma. 60 Ma.

51-300 V.

80-250 V.

C

F e /cis

,C

One 80 and one 5Z3 are used in the power supply, the 80 for C bias voltage and speaker excitation,and the 5Z3 for the B voltage.

September, 19:55 R AD 10 WORLD 37

Diode -Amplifier Tube in T.R.F. SetAWay Out of the Difficulty Presented by Grounded Rotor

By Henry Bullitt

3

Won't work. Doesn't solve problem.

THE use of a diode detector as found in themultiple tubes presents a problem in con-

nection with tuned radio frequency sets, becausethe tuning condenser rotor is grounded naturally,so return to cathode through a resistor, in near-ly all circuits, prevents grounding the cathode'directly. Also, the load resistor for the rectifieris the biasing resistor for the triode or pentodeamplifier in the same envelope. This resistorhas to be so high, to support the rectifier, thatthe bias is high, and the bias has to be overcomeby the signal before rectification takes place, animpractical attainment, considering the simplet.r.f. set. The first diagram (at left) shows thesituation.

Then we might consider grounding the sec-ondary as well as the condenser and the cathode,and have the circuit as shown in the seconddiagram.

Series Tuning Condenser

This has an objection, in that radio frequen-cies get into the triode grid, indeed that grid isin parallel with the diode anode, so the diodemay be taken out of circuit, for we have gridleak detection anyway, whether we want it ornot.

Another scheme is the third one, with thetuning condenser in series with the coil. Thischange may be made because series and parallelconnections produce the same frequencies, or,rather, the formulas for both are practicallyidentical. But the tuning condenser becomes avariable bypass condenser, and so we have tonecontrol of a pronounced sort, more high audiofrequency attenuation at the higher wavelengths(lower frequencies).

A Practical Case

The fourth diagram, however, offers a prac-ticality, because a choke coil of 10 millihenriesinductance is introduced between the anode and-the control grid of the amplifier, and this keepsout a sufficient percentage of the radio frequen-cies to prevent a feed for grid leak detection.

Not hot, either.

.101n1h

-1=1

0.01

"7.2E '00015

_L

r1715

AM

IOm/h r

The remaining radio frequencies are wiped outby the filter in the plate leg.

The diagram represents diode biasing of thetriode, but is applicable also to self biased ampli-fier, by interposing the biasing resistor betweencommon lead and the ground symbol, by-passing that resistor with 0.1 mfd. up. Connectthe choke to the diode anode through a stoppingcondenser, say, 0.01 mfd., and put a leak of 0.5meg. or more from grid to ground. The bias-ing resistor for audio transformer in the plateleg may be 1,000 ohms, and for a resistor of 0.1to 0.25 meg. in the plate leg, bias with 5,000ohms, bypassed by 0.1 mfd. up.

The point may be considered that the vari-able bypass effect of the tuning condenser ispresent in the fourth example as in the third.However, the tuned coil may have sufficientd.c. resistance to bring the coils' distributedcapacity into play, so the tuning condensermight have little effect on tone.

Selection of Bypass Condensers

The selection of bypass capacities across bias-ing resistors like those discussed may be on thebasis of actual regeneration in the audio chan-nel. Where resistance coupling is used theremust be some regeneration expected, and whenthere are two transformer stages there may beconsiderable regeneration, also. This conditionchanges the requirements of high bypass capaci-ties, because regeneration is at some low audiofrequency, usually, and small condensers takecare of protection of the high audio frequencies,even capacities as low as 0.05 mfd.

Capacity Compensation

The bypass capacities therefore may be in-creased to that point where the low note repro-duction is strong, or, if motorboating or humsets in, then the capacities may be decreased justuntil the trouble.stops. By tuning in an orches-tra the test may be made with sufficient assur-ance of a good standard. Listen for the bassviol and the drums.


DIRECTIONAL TRANSMISSIONHow It Is Applied to Ultra Waves

By J. E. AndersonTHE advantages of directional transmission

of radio waves, especially of those nowcalled ultra short, have been recognized sincethe beginning of radio technique. Hertz, thediscoverer of electric waves in space, producedultra short waves and employed directional de-vices for stuying their properties.

Directional transmission of the longer wavesis not practical because of the large dimensionsrequired of antennas, reflectors, and special an-tenna arrays. For this reason little has beendone about beam transmission until the lastfew years, when technique was developed tothe point where very short waves could be pro-duced simply and efficiently. At present thegreatest progress in radio is being made inultra short waves and their directional trans-mission and reception.

/.3

1.3

L2

,9

x.8

.6

5

.4

.3

.2

00 / .2 .3 .4 .5 .

A A

2

The general idea back of directional trans-mission is to send as much of the availableenergy as possible in the direction in which itis desired, and none at all, if possible, in anyother direction. The transmission may be suchthat the energy is confined in a narrow beam,which is most desired in point to point com-munication, or it may be such that the energyis spread out over an angle of 180 degrees,practically nothing going in the opposite di-rection. In ordinary broadcasting the availableenergy is spread out over the full 360 degreeswith almost equal intensity in all directions.

The simplest device that will send out a beamin one direction is a parabolic reflector. Theradiating antenna is located at the focus ofthe parabola and the energy is sent out in abeam just as light is sent out from an auto -

Basis for Constructing a Parabolic Reflector

= s -----pi HA -iENINIIIIIIIITE-IffilliEN11111 MON 1

It LIIE1110&11E IEEE NM alrilliiii. , 1 1 oi .. -

ANIE.ZEIREP.IIRRIE,..':11r1

IIII 7.:i '': . .......

= 2 i:m....1":.

ssIn

I.I

i. .. LaEllitIRMi:LI.R-ADMEEMEEffiENNINEME-OR

In : -... LP qiiiMINELINE1111HEINIENTLITHRIME-RIIIII i erii if

-li h a hAllikiERIEBNEMIRIBIllINaiiiEiTERMIERMERIE4111MREEIRIERMERILMINE.

MODEM MEE' MEI-ORP.:1""111 Ili

KilliffinliniirligiffiliffiliERRIPMEMU: r1 E:

MEI bliK' ' i a m

'. Min d

IP = ,' PARIRRIEMMESIRERI'MERIEMERIMIfflfflERBOMME MIDI. ": 1: IIIIIM

IiITIMIIMPAIIIIIIIMMINI1.0111111MINIMEIMIIIIREMEMBIN,

MI EMMENIBUSIIIIMILMIINEMIN111111.3

PINCE IR RfilE.IRMRIEFAR: :

' It IIITIMEIRIBMIRMEREiRMEMIEREMMAREllifilliEffi DIHRummitmHrillumormompfigERRIIRRELMIE iftlill1...:1 '' 'WI'

' Iii

:i: RIERIMERINT REIREDERREIRIRERE Mil=Pill' EqI1111110===aillin NI RILMIRREARRIffil RIR EDE HRIMMIIIIMIIIIL OMIT' E ihathil

hi! r`32 IirliriMINIIIIMIHMEINI MOTAIIHIERII HE ' IIII R IMPIRIARIEM 1 i:

: 'AitillillERIRMIRMEINERIRMIPIXT4 HERM RIMMURRIEllia-e-i

..': xiMPINIERPROMPERIMIRMILliffilini 11 IBM INIIIIIMIllmors"F. 'h imilmairoilmamit..t.E.gos um , Mb ! II lin 9u. I .

-Th.n MIMI!! 1' gin "L:i Vii . I. ,,*: HIT ' R'''0;04 11"tt.

=13i 1d i i AMP

_'d t TSMIUMF:

ii .511E.

r .......- -1. ' 3 i 11 11"1111 " 11-

......

inr ' WM lill .:--r

Lin

..

li iii !IN I E. . . ..:

-,iiii 9

1 iii;::....iii::

' i . =ii " iiRIERitiliiirticl iiiiiiilli i 1 i stli 113.L.11:

.. rA. III J : , iffill Iliii.V '.....

: iiii. ...: RNA' I illii a MU 1111 H. hi.i7

' I Miilhi Si ... . .q.4 Ril in ..BRILVE"ViR PE BRIM BRIE'-t

.. 'al nal ' 'n "nhE MINEMEPidiediMill IMMO ilITP.7 .8 .9 10 /l 1.1 /3 1.4 /S /6 /7 /8 1.9 2.0

JA A 3A ZA4

FIG.One half of a parabola having an axial depth of two wavelengths. The focus, F, is one quarterwave from the origin. To get actual distances for the construction of a parabolic reflector multiply

the coordinates by the desired wavelength.


mobile headlight. There is a difference, how-ever. The mirror surface in a searchlight is aparaboloid of revolution. The reflecting sur-face in the beam transmitter is a paraboliccylinder. It is produced by bending a flatsheet of copper, or of other metal of very goodelectrical conductivity, into the form of a para-bola.

The Wave Front ConsideredThe special property of the parabolic re-

flector is that if a plane wave (plane wavefront) enters the parabola so that the wavefront is at right angles with the axis of theparabola, or so that the direction of propaga-tion is parallel with it, all the waves will con-verge at the focus.

Conversely, if a point source of wave motionis located at the focus, the wave front of thewaves emerging will be plane. This means thatthe light, or sound, or other wave motion, willnot attenuate. It will be as strong ten miles

FIG. 2 FIG. 3A cross section of a A parabolic reflec-pare bolic cylinder for terminating athaving an axial depth the focus, a x i a I

of one wavelength depth quarter wave.from the origin to One half of the ra-the termination X. dieted energy is re -The right angle solid flected into a beamline is the same and the other halflength as the axis, is radiated radially,since each half is as shown by diverg-half a wavelength. ent lines.

away from the source as it is close to thatsource, excepting for losses by absorption in thetransmission medium.

This is the great advantage of the parabolicreflector. Without the reflector an electricwave from a vertical antenna would decreasedirectly as the distance. Ten miles away fromthe source it is only half as strong as at fivemiles. Actually it is less than half, due toabsorption.

Construction of a ReflectorThe construction of a parabolic cylinder for

waves of less than one meter is not difficult.The principal material required is a long rec-tangle of sheet copper one-half wavelength wideand at least 1.3 wavelengths long. This sheetof copper is bent over two equal wooden formspreviously cut into the shape of parabolas ofthe proper dimensions.

As an aid in laying out the parabolas on thewooden boards previously to cutting, a parabolaof the proper shape is given in Fig. 1. Only

half of the parabola is given, for it is exactlythe same below the x-axis as it is above it.This half parabola has been carried to a pointwhere x is equal to two wavelengths in caseit is desired to construct it that length. Thefocus of this parabola is at F, on the x-axis,and therefore on the axis of the parabola, andit is located one -quarter wave from the origin.At the focus the distance y is one-half wave,or the total width of the parabola is one wave.

It is not necessary to make the depth in thex direction much greater than one -quarterwave, but a sharper beam will be obtained if itis made deeper. This effect will be discussedlater.

Looks Like a Cradle

When the two wooden parabolas have beenshaped properly the sheet of copper may betacked to the edges of the wood. The finished

FIG. 4The depth of thisreflector is one halfwave. In this case39 per cent of theenergy is radiateddivergently and 61

per cent in a paral-lel beam. Greaterdepth than this doesnot much increasethe parallel beam.

FIG. 5A parabolic reflec-tor may consist ofresonant conductorsspaced less than onequarter wave apartin the parabola. Eachconductor shouldhave the same nat-ural frequency as

the radiating anten-na.

structure will be in the shape of a cradle orround bottom trough. The radiating antenna,which should be one-half wave long, can befastened to the wood pieces at the two foci ofthe wooden parabolas. Provision should bemade for varying the position of the an-tenna in the axial direction by a slight amount.Therefore oblong holes, or slots should be cutabout the foci so that the position of the an-tenna may be varied closer or farther awayfrom the origin than one -quarter wave. Theantenna should occupy that position which givesthe greatest forward radiation, and this canbest be determined by trial. Only a slightshift from the focus, if any, should be required.

It should be pointed out that a half wave an-tenna is never one-half wave long, as the waveis measured in free space. Thus an antenna fora one meter wave will be slightly less than one-half meter. The reason for this is that theelectric wave travels at a slower rate on thewire than it does in free space. The same does


40 R ADIO WORLD September, 1935

(Continued from preceding page)not apply to the quarter wave distance betweenthe focus and the origin, nor to any other di-mension of the parabola, because these distancesare free space. It would apply to the widthof the sheet of copper out of which the reflectoris made, but there is no harm in making this alittle greater than one-half wave. Indeed, it isadvantageous.

Sharpness of BeamAs has been said, the sharpness of the beam

depends on the depth of the parabola, that is,on the distance between the origin and thepoint X in Fig. 2. The waves starting outfrom the antenna at F will be circular but onlythat part of the wave front intercepted by thereflecting surface will be reflected in a parallelbeam in the direction X. That part which isnot intercepted will be attenuated just as ifthere were no parabola present. The deeper

Less than k.

is 61 per cent. Thus doubling the length ofthe reflector has increased the sharpness of thebeam considerably, since only 39 per cent of theenergy is allowed to spread out and attenuate.

If we add another quarter wave to the depthof the reflector we make the reflected portionof the energy 2/3 and the divergent portion 1/3.The improvement in the sharpness of the beamis not proportional to the increase in the depthof the reflector, and we suspect that little isgained by increasing the depth still further.This is borne out by the figures, for if wemake the depth a full wave the direct forwardtransmission is 30 per cent of the total andthe reflected portion 70 per cent. As the re-flector becomes unwieldy as it is made deeper,and also as the increase in sharpness is notrapid, it may be impractical to make the depthgreater than one half wave. Of course, it de-pends on whether the reflector is to be movedor rotated or whether it is to remain in a fixed

-4--L.N -< NI <

FEDCBAB CDEFthe parabola the greater will be the percentageintercepted.

Suppose the parabola is terminated at thefocus. This case is illustrated in Fig. 3. Onlyone-half of the total energy from the antennais intercepted by the reflector and sent out ina straight line beam. The other half is radiatedradially. The first part is represented by theparallel lines and the other part by the linesdiverging from the focus. That half which isreflected is not attenuated but the other halfdecreases directly as the distance increases.Therefore at no great distance from the antennathe reflected portion of the energy will dominate,since that does not decrease. Still, the beamproduced by this transmitter is broad and notsharply defined.

Diminishing Returns

Now suppose we double the depth of the re-flector and make the distance from the originto the end, measured along the axis, equal toone half wavelength. This case is illustratedin Fig. 4. The divergent radiation is now 39per cent of the total and the reflected radiation

FIG. 6A reflector of the typeshown in Fig. 5 can bemade by stretching wiresbetween two supports andthen bending these into theform of a parabola. Or asingle support can be used

at the center.

position. Small reflectors are often rotated sothat the beam will be directed not only in thedesired direction in respect to the meridian butalso in respect to the horizontal.

Relatively Long WavesIt is not necessary to have a solid sheet of

metal for the reflector. The solid sheet may bereplaced by a number of half -wave rods allplaced parallel with the antenna and in theparabola, provided that the distance betweenthe rods, measured along the curve, is less thanone quarter wave. The arrangement is illus-trated in Fig. 5. The half wave rods are in-dicated by A, B, C, and so on, and the distancebetween any two, as between A and B, mustbe less than one quarter wave. As in the caseof the solid reflector, the radiating antenna,R, is placed at the focus.

Each rod in this case should be a duplicateof the radiating antenna, for each must be tunedto the same frequency. Thus the length of anyone rod is slightly less than the length of thewave radiated, as measured in free space.

If the wavelength to be radiated is long, sayseveral meters, it would hardly be practical to


use rods. Wires of the proper length, placed inthe correct positions, would be more practical.

A simple way of constructing a radiator ofthis type for short waves, say less than 2 meters,would be to lay out a parabola on a flat woolstructure like a table top and drill holes atuniform distances along the curve of the properdiameter to hold the rods that are to be used.

Why Rods ReflectOf course, these holes must be drilled with

the tool at right angles to the surface so thatthey will all be perpendicular to the sameplane. As in the case of the solid reflector,provision should be made for moving theradiating antenna by a slight amount in theaxial direction.

The reason why the resonant rods, or wires,reflect the waves is that the waves induce cur-rents in the rods and these currents are suchthat they oppose the wave. That is, they sendout waves of their own in the opposite directionto the incident waves. In this respect they actas the secondaries of transformers. Theresonating reflectors must be spaced less than aquarter of a wave length or some of the energywill pass through the spaces between them.These reflectors are especially effective becausethey are resonant to the incident %%ave.

Flexible Supports UsedFig. 6 shows another way in which the

resonant rod or wire can be constructed for re-flection. The half wave resonant wires arestretched between two flexible supports, thedistance between any two adjacent conductorsbeing less than one quarter wave. When thishas been done the supports are bent into theshape of a parabola. A reflector of this typehas also been made by using a single supportin the middle, the conductor then bent into thecorrect shape. This arrangement perhaps issuperior to the ladder type of construction be-cause when the support is in the middle it isplaced at the current maximum and voltageminimum. Under these conditions the losses inthe dielectric will be a minimum. When thesupports are at the ends of the reflecting rodsthey are at the voltage maximum, and conse-quently dielectric losses will he greater.

LITERATURE WANTEDReaders whose name and addresses are

printed herewith desire trade literature onparts and apparatus for use in radio con-struction. Readers desiring their names

and addresses listed should send theirrequest on postcard or in letter to LiteratureEditor, Radio World, 145 West Forty-fifthStreet, New York, N. Y.

Elwin Marg, 159 19th Ave., San Francisco, Calif.J. Babin, 323 Beaver St., North Adams, Mass.Shailer A. Peterson, Head of Science Dept., University

High School, University of Oregon (186 East 14thSt.), Eugene, Oregon.

John Adams, 340 Rugeles St., Fond du Lac, Wisc.Arthur Lillwitz, 836 N. Cicero Ave., Chicago, Ill.11. J. Greenspan, Radio Scrvice. South Haven, Mich.N. C. May, 1421 W. 77th St., Cleveland, Ohio.S. H. Evans, Esq., Weaversville, No. Car.S. W. Case, 198 Woodlawn Ave., Jersey City, N. J.Chas. H. Bost. 1014 Euclid Ave., N.W., No. 204,

Washington, D.C.William Settles, Legion Bldg., No. 2, Owensboro. Ky.Sr. D. Antonio Lopez, Espronceda II, Linares (Jaen.),

SpainStanley Lewis. 512 Union St., Portsmouth, OhioMilton J. Schreiber, 256 Westfield Ave., Elizabeth,

N. J.J. Ginger, c/o Lattio, 269 West 12th St., New York

CitySylvan Elencweig, 343 McKee Place, Pittsburgh 13,

Penna.Charles Rome, c/o Bonano, R.D. No. 2, Nassau, New

YorkCarl von Herrmann, 1424 - 49th Ave., San Francisco,

Calif.Bernard Croninger, 1062 Divisadero St., San Francisco,Calif.

Afsar Mooraj & Co., Tai Villa Souter Street, 8, Bornbay, British India

C. A. Coolidge, Kentfield, Calif.Irwin Creedon, Box 271, Route No. 1, Arlington, Wash.Carl S. Taylor, Dubois, Penna.S. S. Bradley. Ayden, No. Car.Luther M. Cope, Box 505, Connellsville, Penna.Allen T. Simmons. Station WADC, Tallmadge, OhioEverett Clutter, RR. No. 2, Blacklick, Ohio\V. W. Cook, 2313 East 60th St., Kansas City, Mo.M. R. Purlee. Seymour, Ind.C. V. Patterson. 2348 Tipperary Road, Kalamazoo,

Mich.J. M. Corey. Bellaire, Mich.T. H. Cox. 522 - 9th Ave., So., Columbus, Miss.Charles Walter Greenley, Station KGCA, Decorah,

IowaT. F. S. Cushine, 349 Worthington St., Springfield,

Mass.D. C. Akers. 181 Greenwood Ave., E. Orange, N. J.

Separate B Supply for Short WavesIn quite a few short wave sets it will

be noticed that the power supply and thespeaker are external. One important reason forthis is the elimination of hum. Short wave re-ceivers are much more sensitive to hum thanbroadcast receivers, and for that reason extraor-dinary precautions must be taken against it.Perhaps the greater sensitivity of short wavereceivers to hum is in large measure due to thehigher gain necessary in order to bring in theshort waves, since most of them are weak. Butit is also due to the fact that short waves strayaround more than longer waves. They will getthrough a small capacity where a long wave

would not get through. Likewise they stray byvirtue of the greater effectiveness of inductionfields for the higher frequencies. Also, if humfrequencies are present in the set, grids in theleads to which there is a small condenser and ahigh grid resistance will pick up the hum.Thereafter it will be amplified as a modulationon the high frequency signal, or directly as anaudio signal. In many instances it has beenfound that by moving the power supply a fewfeet away from the set all hum disappears.whereas if the supply is built into the samechassis with the tuner and audio amplifier thehum may be intolerably loud.


A Little More Meter InsuranceCan't Fuse Sensitive Device, So Watch the Switch

By Edward B. NelsonLow ger/stance

I N the construction of volt -ohm -ammeters theexperimenter may select more current ranges

than voltage ranges, the total number of rangesbeing governed in general by the number ofswitch positions. Switches ordinarily obtainablehave nine positions maximum. So three volt-age ranges, five current ranges and a high re-sistance range may be selected. If it is desiredto have low resistance measurements to lessthan an ohm, then the meter shunting methodmay be employed. This does not require anextra switch position, as the meter terminalssimply are brought out to extra binding postsor jacks. So, too, the meter is accessible forits maximum sensitivity on current measure-ments.

Computation of Voltage Drop

The reason for more current ranges thanvoltage ranges is that often a current conditionis critical, whereas seldom is a d.c. voltage socritical. Since current in modern receivers,particularly small current, flows through highresistances, a small difference in current maymean a large difference in effective voltage.Sometimes the current is so small the meter oneis likely to have does not enable the measure-ment of effective voltage, so B plus to cathodevoltage is measured, then the current throughthe plate leg resistor, and the effective platevoltage computed, it being the difference be-tween the applied voltage and the voltage dropin the load.

That drop is determined by Ohm's law, volt-age in volts equals current in amperes times re -

A three deck switch is usedin this volt -ohm -ammeter,so the faster connectingtab on deck Db or Dc maybe connected to the meter,to insure the shunts beingin place before the outputbinding post is picked up.This prevents exposing the"raw" meter to high cur-rent. Rm I etc. are volt-age multiplier resistors andRsl etc. are shunt resistors.If the battery voltage E isproportioned to one of theresistors (Rm2 above) aseparate ohmmeter resist-

ance is not required.

sistance in ohms. The voltage and resistanceare units, but the current is a fraction repre-sented decimally by milliamperes, that is thou-sandths of an ampere, or microamperes, ormillionths of an ampere. The values 1 milliam-pere and 1 microampere are written, respective-ly, 0.001 and 0.000001 in ampere terms.

Commercial devices usually have the samenumber of current ranges as voltage ranges, be-cause the two are linear and the same calibra-tion may be taken for volts or current purposes,simply by selection of the proper accurate multi-plier resistors for voltage and shunts for cur-rent, and applying decimal or other integralscale multipliers.

Meters are very strongly built these days,and while it must not be taken as a recommen-dation to be careless, they often stand terrificoverloads. An example is that of a 0-500 micro-ampere galvanometer, used in constructing thevolt -ohm -ammeter as diagrammed. Accident-ally shorting of an external resistor used forvoltage dropping purposes caused 100 volts d.c.to be momentarily connected across the meter.The resistance of the meter is 16 ohms, so thecurrent for that brief moment was at least 6amperes, yet there was no burnout.

Hit the End Hard

The needle hit the end stop so hard that thepointer was bent, and the whole needle wasjammed against the scale and would not move.So the meter seal had to be broken to removea screw at top of the case side wall, one of threescrews holding meter to case, and then the


needle was straightened on the removed meter.After that everything was just as fine as before.

While such a condition is not to be expectedin ordinary practice, it is nevertheless possiblewith a switching arrangement to have the inputbinding posts intended to complete a series cir-cuit for current measurement, and then whenone moves the meter connections from one rangeto another in this condition, expose the meteralone to the current, which may .be 100 timesthe current for which the meter is rated, adangerous condition, and moreover always in-troducing the risk of bending the needle.

Reason for Extra Deck

The danger is present when a two deck switchis used, as then one side of the meter is per-manently connected to an input binding post.The switch action may be such, and with cheapswitches this is not uncommon, that the ridersdo not leave the tabs at exactly the same time.The difference may ruin the meter, or damagethe needle at least, because one side of themeter is connected to one binding post, henceone side of a shunt, so one side of the metergoes to the current source, while the other side,though contacting the other meter terminal, maynot quickly enough pick up the shunt. The re-sult is that for a brief period at least, and ifthe condition is not watched the period mightbe ruinously long, large current is passedthrough an unshunted meter.

An extra deck on the switch will enable oneto free the meter from permanent connectionto an input binding post. The diagram showshow this is done. Unfortunately, a schematicdiagram, not being literal, 'fails to convey thecomplete picture of the operation. Consider thethree decks, Da, Db and Dc. Da is unchanged.Db contacts the other side of the meter to acommon branch, the voltage positions, 1, 2 and3, not requiring service now. For the currentfive positions of the meter are shown pickingup the shunts, one at a time, while Dc communi-cates the meter to the otherwise open side of theshunts.

Speed Is Apportioned

However, if there is an opening it is in theexternal circuit. The method reduces the dangerof unshunted meter one-third, as there stillmight be the same condition as before, but if sothe delayed contact lug can be spotted, and theslower lug made to pick up the meter insteadof picking up the binding post.

This is possible because the two switch decks,Db and Dc, are tied to the common meter ter-minal, but one is a shunt pickup and the otheran input binding post pickup. If the shunt ispicked up before the binding post is, then theproblem is solved. Of course, another answerwould be to get a better switch. The better typeof switch for this purpose lists at $7.50.

Mention has been made of the schematic dia-gram not quite telling the full story. Thus aline is drawn from switch tab to switch tab andseems to be a continuity to the moving arm.Of course this is not so. The arm contacts only

the tabs, and not the full stretch of the wireconnecting the tabs, hence arm can be out ofcircuit completely, that is, simply in space.Hence the long line between switch tabs denotesthe tabs themselves are connected together with-out reference to whether the rider really makescontact.

Meter Shunting Method Simplified

In the May issue was a detailed article on"Measuring Low Resistance," using the metershunting method, and in the present issue thereis a discussion of a sidelight on accuracy. Sincethe same method is used in the diagram here-with, measurement made by putting the un-known across the meter posts, a simplified cal-culation is proposed for those who would notcare to prepare and consult a chart. The meterresistance must be known. The manufacturer'scatalog usually discloses this resistance value.

The meter is set for full scale deflection, doneat the high resistance setting of the switch byclosing the input terminals at right. If the un-known across the meter terminals at top is ofless resistance than the meter the needle willswing to less than half scale. If the unknownis higher than the meter resistance the needlewill not move as far as midway. Hence if theneedle is midscale the unknown is equal to themeter resistance. In general, the lowest re-sistance is easily determined, due to needledefinitely above zero, is say 0.5 ohm, and themaximum 200 ohms, so a general idea of the un-known is determined this way. For a 30 ohmmeter, as most of the 0-1 milliamperes are, thenan unknown is found to be between half an ohmand 30 ohms or between 30 ohms and 200 ohms.

Computation from Scale

To get the value closely enough, get full scaledeflection, note whether the needle is moved toless than half scale or above half scale whenthe unknown is shunted across the meter, andget the proportion the unknown bears to themeter resistance by using the fraction developedfrom the current calibration on both sides ofthe needle indication. Thus, if the needle reads250 microamperes for a 0-1 milliammeter(0-1,000 microammeter), 250 is the numeratorand the difference between 1,000 and 250, or 750,is the denominator. So the unknown is 250-750or 1-3 of 30 ohms, equals 10 ohms. If the un-known is greater than the meter resistance theneedle will rest on the high side of midway, andthe fraction will have to be inverted. Thus, ifthe reading is 750 microamperes, the unknownis 750-250 of 30 ohms, or 3 x 30 ohms, equals90 ohms. Nearly all computations can be doneby mental arithmetic.

The high resistances will have to be computedeach time, too, unless the meter has a resistancescale, which it may have, or such a scale can beput on, that has the other readings, too. How-ever, there are no such replacement scales forthe shunting method of low resistance measure-ment, nor is this method used commercially, sofar as the author knows.


A New Balanced DetectorUsing the 6H6 in Nonreactive Circuit

By Edward C. Sturgis

Lower left, positive grid on right push pull tube, so that'sout; center, stopping condensers are reactive; right, some-

thing to think about.

THERE has been a strong desire amongexperimenters for direct coupling and pref-

erably nonreactive direct coupling. Whenevertwo circuits are united by a single load thecoupling is direct, and when the coupling me-dium has a constant impedance, despite varia-tions in frequency, hence behaves as a pureresistance, the coupling is said to be non -reactive. It is inductance that largely intro-duces reaction in audio amplifiers, for stoppingcondensers may be sufficiently large to makethe reaction relatively small, but there areleakage and phase shift conditions that haveto be considered with condensers and it is there-fore especially interesting to follow the path ofresistance, though not necessarily least resist-ance.

Why Push Pull Is IncludedSince the object of a nonreactive amplifier is

strictly one of quality, to attain which certainsacrifices of gain are made, it will be assumedthat a diode detector is to be used, as thishas the closest to a straight line characteristic,meaning that the changes in current throughthe load resistor are directly proportionate tothe applied voltage.

Push pull is selected for quality, too, becauseit frees the output from the even order har-monics, second, fourth, etc. Since second har-monic distortion is the strongest, as a rule, thepush-pull circuit is favored for output.

So we combine the requirements of a non -reactive circuit with push-pull output.

571For several years we have brought up now

and again for discussion the first and secondcircuits illustrated, the first of which does notwork, while the second does, though none toowell. They are shown now to set forth thetheory. The interconnection of both filamentleads in the output tubes is simply a habit whendealing with symbolic rather than construc-tional circuits.

A modulated r.f. voltage (at left) is de-veloped across the tuned circuit AB, one sideof which circuit goes direct to anode of adiode rectifier, the other side to cathode throughthe load resistor BD. Ignore C and groundingfor the while. When A is actuated by the posi-tive alternation of the cycle, rectification takesplace, because only when the anode is positiveto a. c. will current flow. Note that the cathodemay be floated. The tube's action is unilateral,or the tube is like a one-way street. During thealternation when A is negatively voltaged bya. c. there is no rectification. Nothing happens.No current flows through BD. The tube, then,is a single wave rectifier, and the alternationthat it rectifies is the positive one.

The Grounded CenterWhen the tube conducts there is direct cur-

rent flowing through the load resistor BD.The potential relationship never changes. B isalways negative in a d. c. sense, and D may beregarded as positive, or zero, depending on theviewpoint.

Now take the center of BD and call it C.


If C is used as the datum or reference point,then B is always negative in respect to C, andD is always positive, else there is no voltagedifference between them, i. e., the tube is notconducting, or nothing is being put into thetube and nothing put out.

If we now connect B and D to the grids ofintended push-pull tubes, we introduce to thegrids a theoretically symmetrical circuit, forthe voltages at B and D, hence at grids, areequal in value but opposite in polarity at anyinstant, and this is a push-pull requirement.

Unfortunately, we encounter bias trouble orunbalance right away. The tube at lower left inthe first diagram has grid at zero (no rectifica-tion) or negative (rectification), but the tubeat right has grid at zero or positive. So oneof the tubes runs a positive grid and we runinto very serious distortion right away. Thecircuit therefore may be said not to work. Youhear something, of course, but you would notwant to hear it long.

Cathode May Be FloatedCompromise might be provided by inserting

a biasing battery, but even if the battery is notobjectionable, the precise bias required for bal-ance would be hard to find, that is, there is dif-ficulty in ascertaining the identical operatingpoints for the two tubes. The tube curve itselfwould have to be linear, that is, or possess aspecial symmetry.

Aside from the fact that the circuit is nogood, it does consist of direct coupling, and ifone fails to look too seriously into the outputtubes, and forgets the speaker load, the wordnonreactive might be applied.

The circuit is shown because it demonstratesthe practicability of diode detection with folat-ing cathode, that is, cathode independent of anyconnection anywhere else in circuit save to loadresistor. Afterward use of a ground interposeshalf of BD in each grid circuit and in this waycathode is returned to ground.

R. F. Driver NeededAnother point is that there is no audio ampli-

fication, except as is present in the output tubes,and the detector thus is moved as near to thespeaker as is practical. There is no presentequipment for performing the detecting workin one tube or push pull pair and still gettingenough power out to drive a speaker.

The circuit therefore suggests that, due tothe absence of the customary audio amplifier,it is necessary to supply a strong voltage, atgood power, to the detector tube. The diodeas previously found, even consisting of a triodewith plate and cathode interconnected, will notyield a straight line on high voltage, and willnot stand the current. So an r. f. driver andmore capable detector are needed.

Is It a "Converter"?Next we come to the second illustration,

which provides a solution to the bias objectionraised against the first diagram, by includingstopping condensers. Now BC is the diode load

resistor, the detecting work being done thistime the same way as before, according toexactly the same rules, but the stopping con-densers become charged by the audio voltageequally and oppositely. Only direct currentflows through BC, except for some residualradio frequency which may be ignored in thepresent consideration, but this d. c. is pulsatingat the rate dictated by the original modulationof the carrier. The discharge is through theload resistors RL in the grid circuits that fol-low. The condenser thus acts somewhat likea converter, fed by pulsating d. c. and deliver-ing at its output actual a. c. of single alterna-tion, the d. c. "line" across BC being the zeroaxis. This "converter" explanation, by the way,is not the orthodox one.

This One Not So Hot, EitherTo the power tubes in the second circuit we

may apply a negative bias, as the circumstancesare exactly the same in this respect as in anyother push-pull output.

The circuit does not work very well, as pre-viously stated, and the reason may be duemostly to the capacity unbalance. The locationof the cathode in respect to the grounded heaterof the diode, and the coupling between heaterand cathode by thermal radiation, account forpart of the unbalance. Of course, it might besuggested that this could be counter -balancedcapacitatively, but again, as with bias, it is atricky feat, and the very capacities may change,for instance the cathode capacity effect may bea variable, not balanceable by any fixed capacityin the other branch.

This Is DifferentIt is assumed the load resistors RL in the

grid circuits are equal. They are grounded atcenter for grid return, to effectuate the bias,through the resistor Rb. There are probablyother reasons for unbalance, judging from re-sults, hence the circuit is deemed too tricky tobe worth while.

Something to think about is the third circuit,which represents a theory only. One of themetal tubes, the 6H6, is a double diode, withseparate cathode for each anode, symmetricallyconstructed, and offering the opportunity of abalanced detector.

It will be observed that the modulated radiofrequency is fed to a pair of secondaries, eachsecondary a separate winding, that is, no centertap. It is assumed that the windings are equal,are in such direction, or connections so made,that the phases are right for the diodes, that is,the instant that one diode is positively sup-plied by voltage E from the source the otherdiode is negatively voltaged by the samepotential E, of opposite sign, of course. Alsothe coupling from primary is equal in respectto both secondaries. Hence the diodes taketurns at rectifying and we have full waverectification. But what else happens?

[Readers are invited to send in solutions ofthe problem and an analysis of the third circuit.More data next month. -Frump.]


A 5 Meter Pilgrim's ProgressAdventurous Experimenter Who Risked Life for

"Cause" Reports ResultsBy Arthur H. Lynch (W2DKJ)

THE ultra high frequency activity of theGarden City Radio Club during the past

year and a half has been productive of suchinteresting results that there seems to be somereal reason for a more or less complete andaccurate record.

The club itself is unique in that there are noofficers, no dues, no by-laws, no regular meet-ing place and no regular meetings. When ageneral meeting is called, arrangements aregenerally made over the air and the meetingtakes place in the home of one of the members.The discussion of any radio subject is lookedupon with great disfavor. All this seems ratherfoolish, doesn't it?

A great deal of the incentive to really ac-complish things which are a bit out of theordinary is brought about by such members as"Doc" Dunn, W2CLA, former American RadioRelay League director for the Hudson Di-vision, and Al Williams, the aviator, who haslong been using radio in connection with privateflying. Frank Hawks, the aviator, is a mem-ber of the club by virtue of his operating outof Roosevelt Field, which is in Garden City,N. Y.

The club's work on the really high frequen-

A silhouette of the authoradjusting the rotary beamatop the Hotel New Yorker,some 600 feet above thestreet, where an earlier ex-perimental station waserected. This beam was ar-ranged so that it could berotated without difficulty,but the arrangement had tobe abandoned because ofthe high winds encountered.Moreover, the author hadto be warned by personalfriends not to risk life andlimb in his enthusiasm forradio experimental ac-

tivities.

cies-waves just a few centimeters long-hasbeen under the direction of Ed Glaser, W2BRB,also a former director of the League for theHudson Division.

When the ultra high frequency bands werejust in their infancy there were just four ofthe club's members who were seriously in-terested. They were Dunn ; S. P. McMinn,W2WD ; Jim Tynan, W2BRI, and Glaser. Allof them got rigs working and the neighborhoodgossip which passed during their QSO's wasthe subject of much comment among five metereavesdroppers. Later Dick Depew, W2SB,made himself a transceiver which was usedat his ground station as well as in several ofhis airplanes. The present author was theninduced to get going on "five" with some kindof an outfit. A National transceiver was putto work as a ground station as well as in theair. We had permitted our license to lapseand had to take another examination. TheFederal Communications Commission issued thenew license with our old call. In the meantimeit was necessary to have some other ham aroundwhen we were on the air, either at home or in a'plane. One of the most accommodating, par-ticularly in connection with 'plane tests, was


Kenneth Stanford, W2KE, and our littletransceiver made quite a name for itself inconnection with several test flights.

Airplanes Work DuplexSome time later a regular rig was made up

for us. It was designed by \Ai 2\VD and wasbuilt by Ed Ruth, W2GYL. The original ar-rangement is shown in an accompanying il-lustration. In the club's air tests, at RooseveltField, it was operated as the ground controlstation by Harry Steenberg, W2AOL. Twoairplanes took part in the test, Dunn pilotingone and "Sonny" Tronck piloting the other.McMinn operated the five meter rig in Dunn'sship and Depew operated the outfit in the othership. Very satisfactory two way communica-tion was maintained between the ground andboth of the ships, during the entire flight, whichlasted about an hour. Communication betweenthe two ships was not quite so successful, al-though they did manage to work each otheron voice, when about twenty-five miles apart.

Last winter \V2BRI, Garden City, was heardin Cincinnati, on five meters. And last summerwe were invited to address the ProvincetownRadio Club, of Cape Cod, Mass., on the sub-ject of noise reducing antennas for the regu-lar ham bands. Among those who attendedthe hamfest were K. B. Warner, editor of"QST," and Ross Hull, his associate editor, aswell as G. W. Bailey, W1KH, the A.R.R.L.director for the First District. A paper wasdelivered by one of the engineers of the tele-phone company which described the ultra highfrequency radio telephone link betweenProvincetown and Green Harbor, across Massa-chusetts Ray. The whole thing seemed so sat-isfactory that a group went up to visit the sta-tion and witness a demonstration. It workedbeautifully. \Varner, Hull and the present au-thor breakfasted together the next morning andthey decided right there to get some real workgoing among the league members on the ultrashort wave bands. On the way home fromProvincetown we dropped off to see James Mil-len, W1HRX, Malden, Mass., and discussedthe entire matter with him.

Station at Hotel New YorkerWhat Millen and Hull have done on the

high frequency bands since that time is nowhistory. The great success which accompaniedtheir two way tests, between Hartford and\ falden, was directly responsible for our owninterest in experimenting in the same field. Allthe members of the Garden City Radio Clubjumped into the thing with great enthusiasm.

Through the good auspices of Eli M. Lurie.W2DLG, engineer in charge of radio for allthe hotels under the management of Ralph Hitz.permission was obtained to set up a station onthe roof of the Hotel New Yorker, in the heartof New York City. A room was provided onthe 46th floor and several different antennaswere put upon various parts of the roof, somehundred and twenty-five feet higher.

Our rig was moved from Garden City to thehotel. Just about that time George Shuart.W2A MN, published the information on his

"lung lines" oscillator and the idea was in-corporated in our outfit.

40 Mile Service AreaThe way in which W2DLG "got out" soon

became the talk of the entire metropolitan area.We were getting reports of R9 and R9 plus-whatever that is-from nearly all the stationswe worked and our normal coverage was some-thing like forty miles. Our difficulty in coveringgreater distances was not in getting out butin hearing the stations on the other end. The

Richard Depew, Jr., W2SB, and his home-madetransceiver. The National transceiver on fivemeters is about half the size of the ordinary toolboxes used to house the dry cells and B batteries,

required for power.

evolution of our transmitter is shown in ac-companying illustrations.

Naturally, after finding out how the circuitset up by Millen and Hull operated and actuallyholding a rather long QSO with Hull, fromMillen's station one day last summer, with R8signals at each end of the 125 mile circuit, wewere of the impression that all we needed todo was to set up a suitable beam antenna on theNew Yorker and that we would have notrouble in squirting a good signal into Hartford,about 85 miles away.

Even though the hotel management was mostpatient and offered no complaint about the"spinach" with which we adorned the roof, weran into some unforeseen difficulties of an en-tirely physical nature. At an altitude of some600 to 700 feet we met with an entirely dif-ferent wind condition than one meets on terra



(Continued from preceding page)firma. Then, too, there are certain physicallimitations to roofs 'on high buildings.

But it was "Hartford or Bust." We put upas much of a sky wire as safety would warrantand we aimed it at Hartford. In all this workwe had the hearty co-operation of anotherformer A.R.R.I.. director for the Hudson Di-vision, "Doc" Walsh, \V2B\V. He proved him-self to be quite an aerialist.

Five Meter Chain IdeaWell, when everything was in readiness, we

telegraphed Hull, at Hartford and made aschedule with him. Nearly all last winter someof our gang was on the job several nights aweek, in an attempt to hear from and be heard

The aerial at the ground station at Roosevelt Field,where first tests conducted by the Garden CityRadio Club, between the ground station and twoairplanes, were conducted on five meters. Theantenna actually was attached to one side of awooden ladder, which in turn was fastened to thesteps leading to the roof by means of a couple of

trunk straps.

at Hartford. Many other New York stationsjoined us in the attenmpt and while several ofthem got through, they were never able to doso with regularity and the entire matter wasthought to be a flop. We were never heard atHartford nor did we ever hear Hartford.

We had had visions of a five meter chain ofstations, joining Boston and Washington,through Hartford, New York, Philadelphia andBaltimore. Our dream seemed shot to pieces.We continued to do very well locally but thethrill was practically gone. We cut out thebeam antenna and went back to the verticalhalf wave radiator with a matched impedancetransmission line. We saw Hull and he was ofthe opinion that our efforts had been unsuc-cessful due to his station being behind a ratherlarge hill with relation to New York, while thatcondition did not exist with relation to illen'sstation, some 25 miles north of Boston.

It was thought that there was little use in at-

tempting the stretch to Philadelphia, since theHartford link had proved to be such a failure.We were about to dismantle the station andbring it hack to Garden City, when Herb Gor-don, W 11 B Y, popped in on us one day with theinformation that we were being heard more orless regularly by W3AZG, of Riverton, N. J.,which is just across the river from Philadelphia.That, indeed, was good news.

According to WHBY, the five meter gang inPhiladelphia wanted us to organize the crowdin the metropolitan New York area and put ona series of tests. We agreed to do so and themanner of going about it was this.

Five Meter Radio AssociationIn New Jersey some of the old timers in the

ham ranks get together now and then in thename of the Bloomfield Radio Club. Some ofthe more progressive in that outfit were ratheractive on the ultra high frequency bands. They.in conjunction with some other fellows in thesame area do their stuff under the caption of theFive Meter Radio Association of Northern NewJersey. Frank Lester, W2AMJ, is president ofthe outfit. They hold meetings on the air everyTuesday night. The meetings are conducted justas if the members were actually attending agroup gathering in a single place instead of be-ing spread out all over the State.

At W2DLG we have listened to these meet-ings and, on occasion, because all the memberscould hear our station satisfactorily and thehotel was more convenient for out of town visi-tors than some ham's Jersey location, we havehad the pleasure of introducing speakers whohave had a message for the Jersey gang.

Among those who have addressed them, fromW2DLG, are Ken Hill, W2AHC, the presentdirector of the Hudson Division; Arthur A.Herbert. \VIES, from A. R. R. L. headquarters.at Hartford; F. Edward Handy, W1BDT, alsoof the A. R. R. L. headquarters staff and Capt.Horace L. Hall, the retired sea captain, who is.at the same time, famous short wave broadcastlistener.

Schedule Is Set UpSo, at one of the weekly meetings we passed

the news along to the club and received theirsuggestions concerning a schedule. Theschedule was set up and the information passedalong to the fellows in Philadelphia.

On the appointed night the tests came off.They began at midnight. Earlier in the eveningwe had been receiving reports from stations wecontacted more or less regularly, to the effectthat we were not putting out a signal quite aswell as usual. Static was rather had and whenDoc Walsh and Percy Collison left us we hadcome to the conclusion that there was but littlechance of getting through, but we'd keep the"sked" anyway.

At the end of our first transmitting period.which lasted ten minutes. we listened. Sureenough, there was W3AZG, on interrputed con-tinuous wave. What a kick ! By the end ofour listening period we had been able to hearhim on voice, as well as nn the i.c.w. Well, wewent hack at him with great gusto. On the


next receiving period we didn't hear a tap fromhim.

However, Herbert Snow, W2ETU, who wasoperating a portable outfit, in his car, at thehighest point on Staten Island, called us to saythat he was picking up Anthony Repici,W3FGN, City Hall, Philadelphia. who in turnwas hearing us and reported our signals as R6.W2ETU told us where to look for him and wetuned our head off with no success. So ourfirst Q S 0 with Philadelphia was accomplishedby talking directly to W3FGN and getting hisreplies through W2ETU. If the informationwe have is correct, and we believe it is,W2DLG was the only station heard in Philadel-phia during that test. Also, our signals were

The complete transmitterand receiver, stalled in one

hangars at Roose-velt Field, being operatedby Harry Steenberg, whowas able to maintain con-tact with both airplanes forthe entire time that they

were in the air.

reported by air as well as by post from quite anumber of ham stations in the Philadelphia area.

Two Way With Philadelphia Direct

Well, from that rather humble beginning wewent ahead, encouraged by our first real DXcontact. We put up a new beam, at risk to lifeand limb. We tuned it accurately and we puton another test with the gang in Philadelphia.Before the test we drove over and had a littlechat with W3FGN and made a definite schedulewith him. We phoned W2ETU and he agreedto stand by, on Staten Island, as he had doneduring our first QSO. Everything was all setand we went to work, with but a few of thelocal crowd knowing what we were about.

At the appointed time we called Philadelphiaand sure enough we picked up W3FGN, duringour first listening period. From that point wewent right into a two way conversation.W3FGN told us we were being heard at R9 atW3AZG and that his own conservative reporton our signal was R8. There was as much kickin that 100 mile QSO as we would have derivedfrom a chat with an Aussie on any of the other

bands. While W3FGN was heard by StanleyOehmen, W2HG, in Brooklyn, N. Y., and severalfellows in Jersey, we were the only station onManhattan Island which could pull him through.We had been experimenting with all manner ofreceivers and all manner of aerials and had justabout decided that when we were in for a seri-ous job we could always depend upon our Na-tional SRR ultra high frequency receiver. Itdid the trick for us on this as well as manyother occasions.

The one important point brought to our at-tention by this contact was the great improve-ment in the signal we could lay down in Phila-delphia as a result of the installation of ourbeam. Incidentally, the beam we were using

was a very simple affair, in comparison to theswell arrays used by Millen and Hull.

Strong Local QRM

While W2DLG was generally known to putout a "swell" signal, in the metropolitan NewYork area, nearly all the local hams creditedmost of the Garden City Radio Club's success tothe height of the antenna above ground. Wehad a hunch that this was not entirely so. Weput the transmitter in a car and took it to thehome location. Garden City is a few milesfurther from the Bronx than the Hotel NewYorker is. We got a report from W2APV, inthe Bronx, that we were a little better, at hisstation than we had been from the hotel. Weworked W2AZB and W2DX, both located atSummit, N. J., and they gave us an R8 report,while we were R9 with them from the hotel,which was about one-third the distance to Gar-den City.

As we have said, we were putting out such awhale of a signal that a few of the fellows whocalled us and got no response thought we were


50 RA 1) 1 () WORLD September, 1935

(Continued from preceding pah.)not anxious to work with them. They did nutknow the adverse receiving conditions we hadto overcome. There was local QRM of a sortnot common in amateur stations. This difficultycould not be realized by anyone who did notactually visit our station at the New Yorker.Some of the fellows who did come up to see usthrew up their hands and said they knew thatwe were in a tough spot from what we had saidabout it on the air, hut they had no idea that itcould be as had as it really was.

Working Under Difficulties

Our radio shack was on the same floor as allthe hotel's elevator control anparatt.. It was

automobile and motor bus interference seems tobe totally absent at such a height.

Three Pickups From Philadelphia

W2AOE, Dana Griffin, of the Leeds RadioCompany, helped us to move the gear from theNew Yorker to our new New York headquar-ters. Then he helped us to get it set up and thefirst night we were on the air, we worked thirtydifferent stations and three of them were in thePhiladelphia area. One of them, W3DRA, islocated in Strafford, Pa., twenty-five miles westof Philadelphia, or 115 miles from our own sta-tion. When he is not on the air W3DRA keepsbusy as a Justice of the Peace.

When he can tear himself away from his five

The Depew transceiver and accessories. The hand microphone at the left is of the "close talking"type. This type is used to avoid picking up the noises from the motor when the transceiver is usedin open cockpit airplanes. In the center are key and buzzer. At left is National Company transceiver.

an ideal spot to experiment with noise reducingantenna systems. The current in the buildingis d.c. and juice for our outfit had to be pipedup by a shielded line from a motor generatoroutfit, located four floors below. The results ofour receiving tests with various types of an-tenna systems and the manner in which welicked the local interference are summarized inthe sketches and captions. We were of theopinion that we would be able to do much betterif we could get a location which would be freefrom the electrical interference set up by theelevator control apparatus. Incidentally, theacoustic racket which this type of machinerykicks up is just about as bad as is the electricalnoise.

Through the auspices of a good friend wewere able to lease the tower of the ManhattanCompany building, at 40 Wall Street, NewYork City. We now operate there under thecall of W2DKJ, portable. We are more than900 feet above the street and our antenna sys-tems are entirely in the clear with respect toPhiladelphia, on the one side, and Hartford andBoston, on the other. There is no acoustic noiseand practically no electrical interference. Even

meter rig, W3FGN spends his time keeping theradio equipment of the Philadelphia Police in-tact. His boss, M. Gault, who has charge ofPhiladelphia's Electrical Bureau, told us that heis much interested in the application of ultrahigh frequency radio to police work. He hasauthorized the erection of a beam, on the cityhall, which, it is thought, will remove the neces-sity for "Billy" Penn talking through his hat,which is what he has been doing for the past fewmonths. The antenna, which sticks up from hishat now, is 547 feet above the street.

So, Now It's ChicagoAt a recent hamfest held at Syracuse, N. Y.,

we ran into Willard Wilson, W3DQ, of Wilm-ington, Del. He told us that the night weopened up at 40 Wall Street he heard both sidesof our QSO with W3FGN. Recently we re-ceived the following telegram from W9CVO,Chicago:

W1XG, W2AMJ, W2HG, W2KG, W2HHWand about 12 other five meter phone stationshave been heard in Chicago during the time be-tween 10:00 a.m. and 12 noon Central Daylighttime stop At 10 a.m. Chicago five meter stations


Frank Lester, showing hecan tune the Lafayette su-perheterodyne 5 meter re-ceiver with his left hand,without even looking at thedial. Though he is the de-signer, he says anybodyelse can tune it the sameway. The receiver wasused at W2DKJ, Arthur H.Lynch's station, nearly 1,000feet above Wall Street,where the photograph wastaken. Other sets were

used, too.

will call the East Coast for 10 minutes stop At10:10 a.m. will listen for 10 minutes. etc.. until 12noon stop W9RLA has been heard in Philadel-phia so there is a chance for someone to have a900 mile five meter contact

W9CVO and Fritz FrankeThe schedule was kept and more than .a

hundred five meter outfits in the New York areapegged away at Chicago and then listened theirheads off. The net result was zero success.

Because the tests with Chicago have been un-successful, up to now, is no reason for discour-agement. Many stations in the East have beenheard there, even though ours was not amongthem. Worse luck ! As the telegram fromW9CVO indicates, several have been heard andsince that time, we heard W2HG, of Brooklyn,N. Y., carry on a twenty minute QSO withW9CYE, at Dayton, Ohio.

We can't say that we are too optimistic overthe possibility of a regular contact with Chi-cago.

We are of the opinion that the changing condi-tion of the atmosphere has contributed much tothe five meter long distance QSO's which havebeen made recently, but something must also besaid for the real improvement which has beenmade in receiver design as well as the greatstrides which have been made in the design andthe use of beam aerials.

Buffalo First?

At the Syracuse hamfest we ran into someof the five meter hams from Buffalo andRochester. Dr. Burton Simpson has done note-worthy work on this service.

At Syracuse we had a session with EdwardA. Roberts, W8CH, A.R.R.L., director for theCentral Division. He and Irving Cassidy,W8MAN, told us they would get the gang intheir vicinity busy in an attempt to contact us.They have the towers of a discontinued broad-

casting station at their disposal. Their presentplan is to set up a beam on one tower, headedtoward Chicago and another beam on the other,headed toward Buffalo and New York.

Up in New England, in addition to the beamarray which Millen keeps aimed at the NewYork area and which is being fed by a pair of852's, we have obtained the enthusiastic cooper-ation of A.R.R.L. Director Bailey, W1KH, whois going to instal a beam on two and a halfmeters and aim it in our general direction. Onthe other hand we will do the same thing onour end and trust that we will be able to con-tact him direct, or, at least via Hartford.

Plenty of OpportunityThe accompanying map indicates the routes

over which regular communication has been es-tablished and the dotted lines indicate the routeswhich should offer no insurmountable difficultyfor regular working during the coming winteron 5 meters. In these plans we are not layingstress upon the more or less undependable con-tacts which have been made over distances inexcess of 200 miles, although we are not proneto discount them entirely and they may becomequite regular as we learn more about using theultra high frequencies. We are of the beliefthat so much has been said and written by "au-thorities," which has been take as orthodox andwhich has later been found to be unsound, thatthere is still plenty of room for real experimentand research. Few fields offer so .muchromance.

[The foregoing completes the narrative todate of the 5 meter activities of Arthur H.Lynch and cronies. Some experimental findingslikewise hold interest and will be discussed nextmonth, when circuits used by the author will beshown. A coast to coast 5 meter relay leagueis being formed and prospective members willbe particularly interested in the transmittingcircuits.-EDTTOR.

PHILO T. FARNSWORTH, inventor of the Farnsworth television system: "Television should comein less than a year. Receiving sets can be put on the market at any time. The obstacle is the erec-

tion of transmitting stations. Some work along that line is now going on in this country."


The Metal Tubes in PracticeGive Big Lift to Short Waves-Some Problems Met

By George DubucComet Radio Company

Circuit used by the author in the construction of histuned radio frequency set, using the metal tubes.

EVEN persons who have only a mild interestin the technical side of radio are asking

about the new metal tubes. Are they betterthan the glass tubes? Can they be put into ex-isting sets instead of glass tubes already there?

The metal tubes obviously are intended to bebetter than the glass tubes, otherwise therewould be no point in bringing out the metaltubes. They are smaller than the glass ones,thus providing space economy, also they areexternally more durable, but these would beless than vital reasons for bringing out newtubes.

Since there is a growing interest in shortwaves, it was believed important to producetubes that, while functioning very well on thestandard broadcast band, also would carry overmuch of that efficiency to short waves, wherecircuit losses accumulate at such a fast pacethat every bit of advantage is appreciated andvaluable. To that end the tubes were madewith smaller grid -to -plate capacities. The out-put capacities in general still run around 12mmfd. in all tubes. Especially with multielec-trode tubes it is difficult to keep down the out-put capacity, as it is the capacity of the plateto all other electrodes.

Case of Mutual ConductanceAlso the superheterodyne circuit is practi-

cally the exclusive one in commercial produc-tion, so the tubes for oscillation were made withhigher mutual conductance than their glassequivalents. The main example is a comparison

of the companion 76 and the 6C5, with respec-tive mutual conductances of 1,450 and 2,000micromhos. Where oscillation is a likely usethe mutual conductance will be found higher,but for the general run of tubes it is about thesame as for the glass models, and in some in-stances just a bit lower.

It is the mutual conductance that is the figureof merit of the tube as to how well it willoscillate, and indirectly as to how high thefrequency may be with oscillation still sup-ported. It is notorious of tubes with low mutualconductance, like 700 micromhos, that they stoposcillating too soon. Perhaps this was the rea-son that the battery glass tube 1A6 was sup-planted with the glass 106, which has twicethe emission of its predecessor, and supportsoscillation to 25 mgc., whereas the other washardly good for much beyond 10 mgc. Thus,too, as a makeshift in the absence of substitu-tion, the addition of an extra tube in parallelwith the oscillator for the higher frequencieswas recommended, this being the equivalent ofincreasing the mutual conductance of the netsingle tube in circuit.

Shorter Wiring LengthAlso, the amplification factor was raised for

the 6C5, so that there would be larger outputvoltage, the comparison being 13.8 factor forthe 76 and 20 factor for the 6C5. Also, if thetube is used as an audio amplifier the circuitgain becomes higher, due to the greater ampli-fication factor, and the higher mutual conduc-


Top and front views of the receiver. A precision 12 tube superheterodyne, the frequency calibrateddial of which tracks beter than I per cent. Rear and bottom views of author's t. r. f. set.

tance in general permits larger input signal, sothere is improvement in both directions.

The smaller size of the tubes enables shorterwiring lengths and less stray coupling, im-portant on short waves, while the much smallervital elemental capacity, between control gridand plate, enables loading the tube so as touphold gain, without running prematurely intounwanted oscillation that ruins reception.

Elemental Capacities ComparedThe following tabulation gives a comparison

of the interelectrode capacities in mmfd

PentagridMixer

6A8 6A7G4 -P .03 .3G4 -G2 .10 .15G4-Gl .09 .15G1 -G2 .8 1.0Mod. Input 12.0 8.5Osc. Output 5.0 5.5Osc. Input. 6.5 7.0

G1 -P ...G3 -P .

G1 -G3G1 -A0G3 -A0Output

Mixer -Amplifier6L7.0005 ...25 ...12 ..

8.0 ..11.5 ..12.5 ..

TriodeMod. Output 12.5 9.0 6C5 76

G -PInput

Amplifier617 6D6.005 .01

7.0 4.7

G -PInputOutput

1.8 2.84.0 3.5

13.0 2.5Power

Output .. 12.0 6.5 Triode6D5 45

Detector G -P 3.5 76K7 6C6 Input 4.0 4

G -P 0.005 0.01 Output ... 7.5 3

Input 7.0 5.0 [.Vote: 6D5 and 45 notOutput 12.0 6.5 close duplicates.]

Effect of Shell on HeatThe foregoing classification shows that the

grid to plate capacity is maintained lower inevery cited instance, compared to the equivalentglass envelope tube, but that in general theoutput capacity is the same, and the input ca-pacity may be higher. The input capacity is of

small importance at the frequencies considered,say, to 25 mgc, because much more capacity isadded, both intentionally by trimming, and un-intentionally, due to the distributed capacity ofthe coil, the minimum capacity of the tuningcondenser, and the capacity of the wiring.

Enclosure in the metal housing, with no per-forations, causes the heating problem to becomeserious at times. For the rectifier tube there isa perforated shell.

The effect of the heat on frequency stabilityhas been noted. Even with the glass tubes it isno secret that frequency stability has not yetbeen achieved in commercial receivers for highfrequencies where this stability is of importance.That is, the oscillator changes its frequencyslowly, or drifts.

Effect of Sharp I.F.

For a sharply tuned intermediate amplifier,oscillator drift causes a form of fading, becausesmall change in frequency of the oscillator pro-duces large change in output. Also there arephase shifts that injure tone. It is not sur-prising therefore that the same problems arisewith metal tubes, but correction for drifting ispractical, and some means probably will beadopted other than the makeshift of toleratingthe drift and widening the band passed by thei.f. to hold the station.

A few suggestions may not be amiss. Theproblem of frequency constancy arises elsewhereand solutions found in comparable fields may betried. One of them is to use a saturated pen-tode as a plate load resistor. Thus a pentodetube would be so voltaged that if the platevoltage were increased there would be no in-crease in plate current, for saturation has beenattained, that is, all the electrons originating atthe cathode already have been drawn to theplate, so increasing the plate voltage has noeffect. The behavior of the circuit becomes thesame as that of a pure resistance of constant



(Continued from preceding page)value. The method is used for producing alinear sweep in the cathode ray oscilloscope.

Another approach to the problem lies in

Comparative size of the metal tubes .s shown offagainst the coils

establishing the local oscillator as a I. ,Lle one,worked on the straight portion of ,ts charac-teristic curve. An amplifier tube c..nM be in-cluded to increase the oscillation voltage, andalso buff the local oscillator from the modulator.

Frequency change is related to the violenceof the oscillation, because of the harmonics in-troduced when the grid voltage swing hits theends of the curve particularly. So operation atthe center of the straight portion of the curvetends to improve stability.

Part of the trouble no doubt lies in the factthat the change causes a corresponding expan-sion of the metals of the electrodes, and thusincreases the capacity, or, for metals that workthe other way, the change is in the oppositedirection. Bimetallic trimming and similarplans offer an opportunity for curing that aspect01 the ditticuity.

The new metal tubes are intended for cir-cuits designed for them, as coupling and otherfactors are selected on the basis of the some-what different performance of these comparedto the glass companions. Therefore the intro-duction of the new tubes into old sets, by useof adapters, is not being pressed, and so far noannouncement has been received of any suchadapter in existence or in preparation.

The formal uses to which the new tubes ar.1., !)e put are represented by the receivers nos,on the market, using these tubes. The tir,tsuch commercial receivers put on the mart,ctcomprised the General Electric line, from thelaboratory %%here the tubes were designed,closely foil, r.s1 Icy nearly all the other manu-facturers, \,, IA notably Philco, who printedadvertiseinc! t, wicing confidence in glass tubes,because of 1: long manufacturing experiencebehind them, and something less than confi-dence in the metal tubes of the hour. Alsothe fact that metal tubes were used abroadwith To rt t ,,lutionary advantages was cited.

1,ris,r.t line of metal tubes differs fromthe foreign metal tubes, one significant com-parison being that the American tubes use themetal shell as a sheild, whereas in the foreigntubes the shell is used as an active element, andis even the anode (plate).

It usually takes about a year for the tech-nique on a new tube to become settled, it istrue of all tubes, and there never would beany new ones if there were hesitancy aboutfacing the consumer proving period. Recentlyin that period the technique has been changedbut little, proving that tubes as produced atfirst, and circuits to go with them, are closeto consumer requirements.

A notable advantage in a transmission linefor a leadin is that it does not pick up an>noisy or unwanted signals on the v.ay dosAnfron the antenna to the receiver. If it isthe r. ncentric type it can run through all kind.,

cctrical noise without picking up any of it.This is a boon to those who live in large apart -

Went houses where the antenna must pass sev-eral floors, each floor contributing a bedlam ofcrackles, before it reaches the receiver. Theantenna itself can be placed high up, awayabove the sources of the n,

Rt ar and bottom viLws of author's t. r. f.


RADIO CONSTRUCTIONUNIVERSITY

Answers to Questions by Readers on the Building of Radio andAllied Devices. Readers Should Address Questions to Radio Construc-tion University, Radio World, 145 West 45th Street, New York, N. Y.

0.000/0.25 Meg. 76

50 T1-1C

1ZFC

The circuit at left oscillates at frequencies much lower than, wavelengths much higher than, thequestioner expected. The circuit at right may be used for 5 and 10 meter work, being constantlymodulated by the line frequency (hum). The tuned coil consists of 10 turns of No. 16 bare copperwire wound on 3/8 inch outside diameter, winding length 3/4 inch, with tap at 4 turns from groundedend. The radio frequency choke consists of 15 turns of No. 30 insulated copper wire, wound on1/4 inch outside diameter. The output coupling winding may consist of two turns, slightly spaced, thinch from the grounded end of the tuned winding. With 25 mmfd. the wavelengths are 5 to 3

meters, approximately.

Right Track But Wrong DirectionRECENTLY I became interested in 5 and 10

meter work and realized that some mea-surement device would be necessary, so that Iwould know the frequencies or wavelengths. Asyet I have not calibrated the wavelengths, but Ihave built an oscillator that seems to workviolently. I put a neon tube across the circuitas shown and it lights from radio frequencies,so the r.f. must exceed 60 -odd volts. This factis checked by stoppage of illumination if thecoil system is touched with the finger. I re-duced the tuned inductance one might say almostto zero and still there was oscillation. Do youthink this circuit is satisfactory for a generatorand should I go ahead calibrating it? Diagramis enclosed.-I. B.

If the supposed inductance that you considerin the tuned circuit were shorted entirely,oscillation would still endure, as you are notworking at 5 or 10 meters but at a much higherwavelength, the tuning coils being the 50 turnhoneycombs. If you omit these coils, or evenone of them, there will be no oscillation, whichshould prove the assertion. Omit the so-calledtuned circuit windings (as you suppose them to

be) and put them in the position now occupiedby the chokes. Oscillation may or may not bepresent, but the circuit could be made to oscil-late, that is, if the voltages are high enoughand the coupling large enough. The modifiedHartley circuit shown herewith will be foundsatisfactory and will oscillate under the dia-grammed conditions at 5 and 10 meters, withsuitable inductances. A tuning capacity of 25mmfd. is ample. You might calibrate a fewpoints on your present oscillator just as verifi-cation of the higher wavelengths actually cov-ered, say 30 meters or so, rather than 5 meterswith one coil system and 10 with another. Thenrig up the Hartley and try it with 5 and 10meter coils.

* * *

Carrier and Sideboard SuppressionWILL you please explain transmission ofcarrier and both sidebands, transmission

of carrier and one sideband, and transmission ofone sideband only, carrier and other sidebandsuppressed?-I. K. M.

When an oscillator is producing a constant(Continued on next page)


(Continued from preceding page)radio frequency there may be impressed on itaudio frequencies, to change the amplitude ofthe radio frequency at the audio frequency rateand intensity. This effect of modulation ap-pears on both alternations, and in both it consistsof the upper sideband and the lower sideband.If the carrier is to be maintained, a filter maybe inserted so that one sideband is suppressed,just as a single sideband filter of the crystalvariety is now popular in communication typereceivers. Then the other sideband and thecarrier are propagated. The receiver is of theusual type for both examples. However, if thecarrier as well as one sideband, is to be sup-pressed, an oscillator is set up and coupled toa balanced modulator, a push pull affair, whichdevelops no r.f. voltage across the output wind-ing when no modulation is introduced, but whenthe modulation appears across the input circuitit is a total voltage across grids and upsets thebalance, causing only sidebands to be propa-gated. If one of the sidebands is suppressedthen only the other is sent out. Otherwisecarrier is suppressed and the two sidebandstransmitted. All carrier suppression systemsrequire an oscillating receiver to supply thecarrier at the set, and the constancy of thisoscillator must be of an exceedingly high order.The tuning must be to within 50 cycles.

* * *

Extra Shielding for Metal Tubes

ARE the metal tubes sufficiently shielded bytheir metal cover for all purposes, and if

not, what remedy could be applied ?-W. S.The metal tubes are sufficiently shielded for

nearly all purposes. The exception exists inthe case of some extremely sensitive circuit,where one or two tubes might have to be givena little extra attention. This consists of shield-ing the top cap of the critical tubes. Also, thelead from circuit to the cap may be shieldedand shield grounded. Receiver stability may beincreased by these extra precautions, but, asstated, the necessity seldom arises. It is notnecessary to enclose the entire tube to provideshielding for the small top cap. A thimbleshaped shield with the proper diameter to gripthe dome of the tube and with sufficient lengthto house the top cap and its terminals meetsthe requirements. Since there is paint on thetube, a part of the top cap shield has flangebent in, so that the protrusion will scrape offthe paint and thus make suitable ground con-nection.

* * *

Measuring an Harmonic Frequency

WILL you please give me a simple methodof determining higher frequencies than

the fundamentals of my direct frequency read-ing signal generator? I understand that har-monics are confusing but that you have amethod that eliminates this confusion.-H. T.

The method to be given is one that enablesthe determination of the frequency of the un-known (receiver) when that frequency is higher

than the fundamentals of the signal generator,but does not enable presetting the receiver tosome desired frequency by use of the method,without additional calculation. The method isas follows : Read a frequency on the signalgenerator that creates a response in the re-ceiver, then slowly turn the generator dial, butdo not molest the receiver dial, until the nextconsecutive generator setting is reached thatalso causes a response with receiver. Havingread the two generator frequencies, subtract toobtain the difference, divide this difference intoeach of the two read frequencies, and multiplythe difference by the two quotients. The an-swer is the unknown. It is a very simplemethod though the statement of it in wordssounds a little complicated. For instance: Sup-pose one read frequency is 100 kc and the otheris 120 kc. The difference is 20 kc. Divide 20into 100, get 5. Divide 20 into 120, get 6. Theunknown is the original difference, 20, times 5times 6, or 600 kc.

* * *

Value Seen in Metal Tubes

D0 the metal tubes have any real advantageover the glass tubes, or are they just some-

thing new, to stimulate sales?-W. D. A.The metal tubes represent a real develop-

ment. What you have possibly heard is thatthere is trouble with some of the tubes. How-ever, nothing is produced without some trouble,and it is foregone that any difficulties will beironed out. Really tubes are tested more incustomers' homes than in factories and labora-tories, for the tube makers have to be gov-erned by public acceptance. There is every in-dication that the metal tubes will be extremelypopular and worthwhile. With one exceptionthe list so far announced is not exclusive as togeneral characteristics, and two glass tubeswill do what that other metal tube does, butthis is scarcely a criterion. The tubes seem tobe largely intended to produce an extra needof service, particularly on short waves, for theinterelectrode capacities are very small, com-pared to those of glass tubes, except the outputcapacities, which are about the same in bothinstances. More gain at the high frequenciesmay be expected. The fact that dozens of setmanufacturers are out with metal tube setsshows that they were convinced of a valuepresent in the tubes beyond mere novelty.

* * *Ordinary Tubes for Below I MeterwHAT is the accurate statement about

the highest frequencies in general thatnormal tubes will tune to? I understand thatsuch tubes are scarcely much good for any-thing below five meters, and in fact the fivemeter oscillators as found in transceivers donot seem to be particularly good.-K. C.

The statement about 5 meters or a bit lowerbeing the limit is a conservative one indeed, asit has been possible to use normal tubes in astraight oscillator circuit (no superregenera-tion) to approximately the conservatively ratedlimit of the acorn tubes, 70 centimeters. The

September, 1935 _RADIO WORLD 57

76 has been used in our laboratories in a 2.5meter signal generator, and also frequencies sohigh as to yield no harmonic or response froma 1 meter test source have been obtained. Thetechnique of the use of normal tubes for veryhigh frequencies, somewhere in the centimeterrange perhaps, has not been fully developedalong conventional circuit lines. The methodusually introduced is that contributed by twoGermans, Barkhausen and Kurz, with positivegrid and negative or zero plate, the wavelengthbeing voltaged controlled, or an amendment tothis scheme, where a tank circuit actually ispresent. We do not refer to these methods butto Hartley and other familiar oscillators ator below 1 meter.

* * *

Accuracy DesiredCAN you suggest some accurate method of

measuring frequencies? I am interested inboth audio and radio, but it seems difficult tomake any progress along precision lines with-out very expensive equipment that I can notafford. I could rig up circuits if you wouldoutline some procedure.-W. E. C.

You should build a tuned radio frequencyreceiver, using plug in coils or otherwise, orfixed tuned to three selectable frequencies, sothat you can pick up the frequency standardssent out by the National Bureau of Standards,through WWV. These standard frequenciesare 5, 10 and 15 megacycles. In conjunctionwith this receiver you should have a modulated-unmodulated test oscillator (signal generator)to cover the bands of radio frequencies youwill desire. By approximately calibrating thisoignal generator for these various bands, say,50 to 15,000 kc, you can check against thestandard transmissions up to generator fre-quencies equal to the standard. From the ap-proximate calibration it is simple to get theaccurate results, by tuning in the standard onthe receiver, combining signal generator andstandard in a single detector, and putting abypassed d.c. plate milliammeter in the de-tector, to watch the needle movement. Thebeats when close to zero will represent mediummovement of the needle, too far from zero willbe shown by fast movement, and sometimes youcan make the needle move so slowly that youcan count the beat frequency, or difference. Ifyou can count the beats, say, four a second,denoting frequency of difference of 4, you arelucky, because your accuracy is splendid. Thegenerator's true frequencies then are referredto the standard on the basis of dividing thestandard by whole numbers, e.g., if 5,000 kcis being received, generator at 50 kc will beatits 100th harmonic with the standard, thehigher generator frequencies will be repre-sented by lower harmonic orders. 5,000/99,taken as 50.5 kc, next 5,000/98, taken as 51,

etc. As the frequencies of the generator be-come higher the points are less numerous andyou would have to wait for a higher standardfrequency, and thus complete your calibration,which may be checked regularly against thestandard. If now you will set another low r.f.generator going, and calibrate that, you canuse closer figures for the audio tones produced

Bo

Neon tube output indicators.

by the difference in a detector tube into whichthe two radio frequencies are fed. Actuallytwo of the frequencies mentioned above are50.505 and 51.0204, and an audio beat oscillatormay be fed into another detector along withthe audio tone from the two low frequency r.f.generators. The difference frequency is 515.4cycles for zero beat, which is not hard to estab-lish in this instance, because the resultants arelow frequencies. If there is a difference themeter will enable one to tell what it is but notthe direction of difference, e.g., whether 5

cycles too high or too low, to register zerobeat certainly, although indirectly even thiscould be checked. The beat between 50 and50.0204 yields 20.4 cycles, as low as you couldcalibrate the audio device, probably. By thesemethods you can proceed along the whole gamutand also refer both radio and audio systems toa standard of frequencies accurate to 5 partsin a million.

* * *Uses for Neon Lamp

CAN the neon tribe be used so that onlymodulation will light a lamp attached to a

set's output tube? What about the direct cur-rent through the primary winding ?-I. L.

The tube may be used in either of the waysshown. At left the lamp is NL, and its limit-ing resistor is R, usually built in, and 100,000ohms. The voltage drop due to the d.c. throughthe primary will not be nearly enough to lightthe lamp. However, assume that it is enough,more than 65 volts, purposely accomplished bygrounding. Put a condenser in series with thelamp, as Cat right, and no d.c. will pass throughthe lamp, only a.c. The condenser should be ofthe paper dielectric type, however, otherwisethere might be sufficient leakage to nullify theintended result.

* * *

"Straightening" the 6E5Pr HE 6E5 electron ray indicator tube, as II understand it, has an illuminated screen, and

the width of the illumination is controlled bythe bias voltage, the more negative that volt-age, the narrower the shadow angle. Since theaction is not quite linear, due to an amplifiercharacteristic of the triode part of the tube,could not some external impedance be used tomake the curve linear, so that the differencebetween 6 and 7 volts, or even 7 and 8 volts,



NOISE REDUCINGTWINS

LYNCH HI-FIAssembled AntennaAll ready to hang. Saves90% of installation time.Any novice can install it-Quickly, easily, correctly.Makes every set performbetter. Results are amaz-ing.

LYNCH HI -Fl e6.75ASSEMBLED P

ANTENNA list

The ONLY TWO Companion Devicesfor Reducing Noise on ALL Waves

LYNCHFILTERADIOCuts out objection-able noise from thelight line. Makesinterference from cir-cuit breakers, motorgenerators, etc., prac-tically negligible.Easy to install; sim-ple to adjust. Onetype serves ec gmall purposes. 01,,vu

listBe sure and see these NOISE -REDUC-ING TWINS at your jobber's or dealer's.If they cannot supply you, order direct,

or write for folder.ARTHUR H. LYNCH, Inc., 227 Fulton St., New York

Pioneer of Noise -Reducing Aerials

LEARN THE FASCINATING

PRINCIPLES ON WHICH

TELEVISION IS BASEDIt is by no means difficult to learn

the fundamentals of television, to geta real insight into the processes andmethods, for they are revealed, simplybut expertly, in "A B C of Television,or Seeing by Radio."

Raymond Francis Yates, author of the book,is a famous interpreter of scientific topics, andmakes everything perfectly clear.

The book is very practical and gives con-structional data for experimental televisionarrangements of the low fidelity type, excel-lent on which to learn, and reveals facts whichmay be used as the basis of construction ofhigh-class receivers and scanners. With tele-vision on the air in Canada, and television sta-tions now building in the United States, every-one should be prepared to get in on the groundfloor, and knowledge is the magic key. Thisbook enables attainment of that knowledge.210 pages. Freely illustrated. Price: $3.00,postpaid. Prompt shipment.

BOOK DEPARTMENT

RADIO WORLD145 West 45th Street, New York, N. Y.

(Continued from preceding page)would be as wide as that between 1 and 2volts?-I. B.

The triode has been loaded with a very highresistance as it is (1 meg. recommended, seeAugust issue), although principally for pur-poses of limiting the current through the tri-ode plate, and there does not seem to be anyready way to straighten this out. Whetherthere is much need for linearity is debatable, asfrom 0.5 to 6 volts there is fairly even division,and the limit is 8 volts, with crowding between6 and 8. The tube is intended for visual tuningpurposes mostly, but other applications, such astube voltmeter and the like, will have to bemade to subscribe to the performance enabledby the geometry of the tube. If it is desired toshift the spreadout from the low to the highvoltage extremes, this may be done by biasingthe tube 8 volts negative, thus reversing thesignificance of the change of shadow angle,that is, zero voltage would be closest angle,and 8 volts would be maximum shadow angle.By a switching arrangement it would be prac-tical to use the tube one way, then the other,to shift the spreadout from one extreme to theother. It should be remembered that it is inthe measurement of small voltages that tubevoltmeters are most important.

* * *

Reversed Tickler OscillatesI S IT possible that the tickler that I con-

nected in reverse, that is, in the oppositedirection to that which it should be for oscilla-tion, nevertheless will cause oscillation? Iseem to have had that experience and waswondering if there was anything within yourown experience to confirm this ?-W. D.

Yes, the condition you mention is not anoddity at all. The oscillation or feedback isthe result of the capacity coupling betweenprimary and secondary. Usually the oscillationprevails for the higher frequencies in tuningany particular band, and then the oscillatorgoes dead. The phenomenon, if such it maybe called, is particularly present on high fre-quencies (short waves). In fact, when thefrequencies become high enough there may bealmost a complete phase shift in the grid cir-cuit, requiring for the support of oscillationall over the band that the tickler be connectedseemingly the "wrong" way, but actually ofcourse it is the right and only way. Withregenerative receivers this condition may benoted indirectly, in that the throttle condenser(if feedback is capacity controlled) has to be'turned the opposite direction than for lowerfrequency bands to increase regeneration. Itis also true that as frequencies become veryhigh the nature of the oscillator changes, thatis, a Hartley turns into a tuned grid oscillator,or other such unintentional circuit swappingtakes place.

* * *Amateurs on 5 and 10

WHAT methods do the amateurs use to besure that they are inside the bands on 5

and 10 meters, as I have not seen any infor-mation about devices for measuring such fre-


(Continued from preceding page)quencies, for ham purposes, only laboratory in-struments of considerable cost.-J. R. D.

We do not know what methcc they use.Many no doubt depend on other amateurs asguide posts, by tuning them in, but on whomthese guides depend we don't know. Frommeasurements we have made we would assumethat much of the frequency selection is theresult of hit or miss, even allowing for con-siderable bootlegging of call letters by poach-ers who wouldn't care what the frequency was.The requirement for adherence to frequencypertains in this band as to others and no doubtin due course ready methods will be perfectedfor enabling hams and others accurately todetermine such frequencies.

HIGH-CLASS PRINTINGPREPAID IN UNITED STATES

1,000 Business Cards, fine vellum, free case $1.351,000 Letterheads, standard size, 20 lb. bond $3.651,000 Envelopes, Statements or Billheads $2.751,000 Veri Cards (six lines maximum) $1.35

Wonderful values. Trial order will convince you.Let us quote on your other printing requirements.

MAIL ORDER PUBLISHERS954 8th AVENUE NEW YORK

Radio Noises and Their CureComplete information on distribution,

interference and suppression.288 diagrams - 76 pages.

Price $1.00 while they last.GETZ, 127 Brick Street, Dickson, Pa.

PROTECT YOUR

VALUABLE KEYS!A Polished Metal Key Check with your name and addressstamped on will identify and usually insure prompt returnof keys if lost. Send 25 cents today and Key Cheek witha Key Ring will be sent to you postpaid. Agents wanted.

IRVIN BUSCHKOWSKY508-N Oak Street Chaska, Minnesota

METAL CHASSISWe make Chassis and Metal Cabinetsfor any make receiver. Experimentersand others who want neat metal basesor housings for assembled apparatus,will find our workmanship superb. In-quiries should be accompanied with cleardrawings, blue prints and other detailsconcerning size and bends in metal ifany, enabling as to give quick service asto prices and delivery.Dealers and Jobbers: Write for special prices.

KORROL MFG. CO., Inc., Dept.R.W.232 GREENWICH ST. NEW YORK CITY

DOCTORS!We will gladly send youReprints and Literature on

Ultra -Short -Wave Therapywith details of the new Lepel

Ultra -Short -Wave MachineIf you will write to

Dept. R.W.

LEPEL HIGH FREQUENCY LABORATORIESInc.

39 West 60th Street, New York City

Royal Air -Cooled Pipe"Smoke with a Breath of Air"

AIRVENT

DRIER

MIXING CHAMBER

Get the Full Flavor ofYour Favorite Tobacco

Three good reasons why the Royal Air -Cooled Pipe has gained so rapidly in popu-larity:

1. No more burning bite, because it is cooledby the exclusive Royal air feature.

2. No more unpleasant dampness or wet heel,because the perforated disc placed justabove the vent allows the tobacco moistureto drop into the mixing chamber where theair mixing with the smoke dries it away.

3. No more disagreeable odor to cling to yourclothes, because this beautiful Algerianbriar pipe styled by experts is ventilated!

If your loco; dealer cannot supply you, writefor illustrations of models, or send $1.50 to

ROYAL PIPE COMPANY, Inc.Chrysler Bldg., New York City


New HIGH FREQUENCYSOCKETS by

HAMM ARLUND

Hammerlund originated low -loss wafer -type sockets, andthey have never been excelled for high -frequency work.Only Grade "B-100," the lowest, strongest Isolantitematerial is used-glazed on top and sides-and "Ceresin"-treated underneath to prevent moisture absorption.Long leakage paths between new -type, positive side -gripping contacts. Due to a square inset anchorage,Hammarlund contacts cannot twist, loosen or shift po-sition, with changes in temperature and humidity. Thenew circular Guide -Groove makes insertion easier. For4, 5, 6 or 7 prongs -60c each, list.

Write for RW-9 for Complete Catalog

HAMMARLUND MFG. CO., INC.424 WEST 33RD STREET NEW YORK, N. Y.

PEAKRegenerative

Pre -SelectorIMPROVE your

S. W. Receptionwith PEAKProducts.

The Peak P-11 isthe only Pre -selec-tor guaranteed to give unconditionalfaction to the most critical S. W. listeners.Tremendous increase in signal strength.Absolute rejection of image.Increased selectivity.Decrease of noise level to signal ratio.PEAK Q-5-the last word in design of 5 meter Super -

het receivers.PEAK X -4 -10 -watt, 5 -meter x-mitter. A quality trans-

mitter at a price within the reach of all, nowavailable.

For further data on PEAK Productssee your dealer or write to Dept. R.W.

satis-

Eastern Radio Specialty Co.Manufacturers of PEAK Products

1845 BROADWAY NEW YORK, N. Y.

QUICK -ACTION

CLASSIFIED

ADVERTISEMENTS7 cents a word. $1.00 minimum.

FILMS DEVELOPED-Any size, 25c coin, including2 enlargements. Century Photo Service, Box 829,La Crosse, Wisconsin.

"RADIO TROUBLE SHOOTING," Second Edition,by E. R. Haan. Contains the latest on A.C. receivers,dynamic speakers and television. A practical bookfor practical men. Contains a special chart showingall possible radio troubles and the way to detect them.Size, 6 x 9 inches, 361 pages, 300 illustrations. Flexiblebinding. Price, $3.00. RADIO WORLD, 145 W. 45thSt., New York City.

WANTED: ORIGINAL POEMS, SONGS for imme-diate consideration. MMM Publishers, Studio Bldg.,Portland, Ore.

"WILLCOX AUTO ELECTRICIAN'S WIRINGMANUAL"-Complete wiring diagrams for all U. S.cars made in 1928-1929-1930-1931-1932. Complete dataon ignition systems, generators, starting motors, bat-teries and lighting. Printed on tough, wear -resistingpaper that can be washed with gasoline without injuryto print or paper. Loose-leaf style in sturdy covers.Size 11 x 12". 300 wiring diagrams. Price $10.00.RADIO WORLD, 145 W. 45th St., New York City.

"RADIO TROUBLE SHOOTING," Second Edi-tion, by E. R. Haan. Contains the latest on A.C.receivers, dynamic speakers and television. Apractical book for practical men. Contains aspecial chart showing all possible radio troublesand the way to detect them. Size 6 x 9 inches.361 pages, 300 illustrations. Flexible binding.Price $3.00. RADIO WORLD, 145 W. 45th St.,New York City.

RADIO WORLD AND POPULAR MECHANICSMAGAZINE - Radio World is $2.50 a year, andPopular Mechanics Magazine is $2.50 a year. Popu-lar Mechanics Magazine does not cut rates, butRadio World will send both publications to you forone year for $3.75. Radio World, 145 West 45th St.,New York City.

THE ART OF INVENTING AND WHAT TOINVENT-By Raymond F. Yates, an inventor ofnumerous devices. 297 pages and 68 illustrations.Invention is not the exclusive property of genius.Often it is the property of the ordinary tinker-infact, it has been said that every man has in himat least one useful invention. This book is intendedfor just ordinary people who do not know how toturn their creative ideas into profit. It assists themby imparting fundamental scientific and technicalknowledge, by providing suggestive lists of neededinventions, and by describing how to patent inven-tions and what inventions are unpatentable or un-profitable. Sent postpaid for $3.00. Book Dept.,Radio World, 145 W. 45th St., New York City.

HAVE YOU A SOUND, PRACTICAL INVENTIONFOR SALE, patented or unpatented? If so, writeChartered Institute, Dept. 70, Barrister Building,Washington, D. C.

KNOW STAMPS-"Stamp collecting" in ten easylessons, to approval applicants, 2k. Havert StampCompany, Dept. D, Milford, Kansas.


Pep on 17mgc.

THE Raco All -Wave Signal Generator, using Raco's own decimal repeating vernier airplane dial,covers from 54 kilocycles to 17 megacycles, a frequency ratio of 315 -to -1, and all on fundamentals.Moreover, wavelengths are read directly from the dial scale, and so are frequencies, of course.And on the highest frequency, 17 mgc. in a region where some other signal generators lose

their strength, the Raco Signal Generator has plenty of PEP.The five wave bands are selected by a low -resistance, low -capacity panel switch. The bands are:(1)-5.400 to 17,000 kc (5.4 to 17 mgc). (2)-1,700 to 5,400 kc (1.7 to 5.4 mgc). (3)-540 to 1,700

kc. (4)-170 to 540 kc. (5)-54 to 170 kc.By close attention to exacting requirements we are able to produce this Signal Generator, and

track the scale to 1 per cent. accuracy, without the use of compression type trimmers, which notoriouslydo not stay put. In fact, no external trimming capacity is used at all, the required adjustment beingtaken care of in the coil winding, so that the capacity between plate and grid windings is properlyestablished for each band. This capacity is held to very close tolerance.

The Raco 339 Signal Generator works on a.c. or d.c., 90-125 volts, iselectron coupled as to output and also as to audio tone, has a separatormodulator tube to create that tone, and is equipped with attenuator, on -off a.c.switch, and on -off modulation switch. Model 339, wired, calibrated, adjusted,complete with 6D6, 37 and neon tubes; instructions (shipping weight, 8lbs.)

$16"FOR INTERMEDIATE FREQUENCIES ONLY

THOSE whose work does notrequire that they have an

all -wave Signal Generator mayuse the Raco Model 341 fordirect reading of intermediatefrequencies, 109 kc to 500 kc.As is our invariable practice,the frequencies and wave-lengths are directly read fromthe dial scale. This model isuniversal, also (works on 90-125 volts a.c. or d.c.). On a.c.the line hum constitutes themodulation. On d.c. there isno modulation.

MODEL 341 is housed in ablack crinkle -finish box and

is a compact handful thatmeets the requirements ofthose otherwise at loss forproper means of peaking anintermediate amplifier. Thismodel requires a 37 tube (notsupplied). Otherwise every-thing is complete-wiring, ad-justment, calibration. Thoughthe price is exceedingly low,the service is high, and theinstrument's life practicallyunlimited. Model 341, $4.50.

(All prices quoted are net.)

RADIO CONSTRUCTORS LABORATORIES136 LIBERTY STREET Dept. W-208 NEW YORK, N. Y.


1/2% AccuracyAttainable in Build -Your -Own

SIGNAL GENERATORLABORATORY perfection of a new type precision

inductance meter has enabled us to hold oursignal generator coils to even a higher accuracy

than formerly, so that they are literally precisioncoils. Used with a particular tuning condenser of406 mmfd., in series with which is to be put a 0.0065mfd. fixed condenser, the tracking of our direct fre-quency reading dial enables coincidence to be within;42 per cent., the limit of accuracy of reading de-termined by mechanical conditions. That is, theelectrical circuit is super -plus accuracy, and only thedifficulty of reading the calibration closer than 14per cent. limits the accuracy to that quantity.

And you can build at low cost a battery operatedsignal generator (as the one diagramed on page 59 of the August issue of Radio World) or a universalsignal generator, operating on a.c. or d.c. (as diagramed on page 24 ot the same issue). We have coilsfor three bands, to be shifted by switching. The frequencies are: 140.500 kc; 540-1,600 kc; 1,400-5,000 kc.

THE 333 SERIESFOUNDATION KIT 3332-N consists of etched metal dial scale, escutcheon, two Isolantite coils forintermediate and broadcast frequencies, tuning condenser (remove trimmer) and precision fixed $2.90series condenser 0.00065 mfd. (Shipping weight, 2 lbs.)

FOUNDATION KIT 3333-N consists of dial scale, escutcheon, two Isolantite coils for intermediates

0\111111 II 1p\\\*%11111 itir /

*

is\

./.7 *(4 914%. oilgniaitl 0'\-tsWiflitsf121%*. ss

Scale and Escutcheons for 333 Series

and broadcasts, one intermediate short wave solenoid coil, tuning condenser, series condenser.(Shipping weight, 2 lbs.) $2.75FOUNDATION KIT 3334-N consists of the same as 3333-N, except that a 0.0008 mfd. precisionfixed condenser is supp'ied, and two escutcheons instead of one, so that by using a three deck switch,or extra switch to cut this fixed condenser in parallel with the intermediate coil, frequencies 83-100 kcare read, and cutting it across the 1,400-5,000 kc coil, frequencies 830-1,000 kc are read, both withenormously wide spreadout, and with wavelength readings for these bandspreads present as well asfrequency readings. The low frequency use enables checking old sets with i.f. around 92 kc, as Victoreen and Magnaformer. The bandspread at the middle of the broadcast band enables closer usedirectly or by harmonics for very high frequency determinations. The bandspread method requires theextra escutcheon to reveal the other half of the dial on which the 83-100 kc or 3,010-3,600 metercalibrations are imprinted. Trimmer to be removed from tuning condenser (Shipping weight. $3.552 lbs.)FOUNDATION KIT 333$-N consists of 3334-N plus the shield cabinet for building either universal orbattery model (room inside for batteries), and two chrome finished face plates, one for output attenua-tion (volume), other for 0-100 for percentage modulation, as generators have separate modulators.And both are electron coupled and stable. Remove trimmer from tuning condenser. (Shipping $5.59weight, 7 lbs.)

INTERMEDIATE FREQUENCY SIGNAL GENERATORSPECIAL Single coil, escutcheon, tuning condenser (do not remove

trimmer, as adjustment is to be made with it for this use)and (wool mfd. fixed condenser to go across tuning condenser, for covering109-220 kc fundamentally, all important higher intermediate frequencies alsodirect reading, with harmonic confusion elimination by two spot loca-tion of these important i.f. Cat. PR -6-K. (Shipping- weight, 2 lbs.) $1.89

PRECISION COILS

SHIELDCABINET

FOR"SPECIAL"

(Cat. PR -6 -SC)

92c

Our precision coils of the 333 series may be used by those who desire to tune them with a 0.00035mfd. single gang condenser, and calibrate the frequencies themselves. The ranges will be close to140-500, 540-1,600, 1,400-5,000. No series condenser would be needed, because of independent calibration.Intermediate Frequency Isolantite Coil (Cat. L-140) 65c.Broadcast Frequency Isolantite Coil (Cat. L-540) 60c.Intermediate Short Wave Solenoid Coil (Cat. L-1400) 55c.Output Coupling Coil, for use with signal generators having an ampli-fier stage. Connect large winding in plate circuit, small windingas output. Frees generators from molesting the sen-sitivity of receivers, and enables measurements onsets using doublet antenna posts (Cat. L-QP) 50c.

Precision - QualityCUSTOM SET BUILDERS SUPPLY CO.

135 LIBERTY STREET, NEW YORK, N. Y.Telephone: REctor 2-6650

When ordering, Please request the particular circuit diagram you want.whether a.c., universal or batteries. Any individual features you prefershould be specifically mentioned and will be included in a special diagramat no cost to customer.


Servicemen, Experimentersand Set Builders !

Build better new receivers,make old receivers betterwith new coils of the latestand most advanced types.These improved coils makesets more selective, moresensitive, and quieter in

operation.Line filter that reduces line noise and eliminates

need for antenna

RADIO DESIGN LINE FILTER AND ANTENNA ELIMINATOR (illustrated)List price, $3,00. Special introductory price, $1.50.

I.F. TRANSFORMERS, 465, 456, and 175 kc, doubly tuned by high grade micatrimmers -3 -pie windings with Litz wire-High Q-The very latest de-velopment in coils List price, $2.50. Our price, $1.00

ALL -WAVE SUPERHETERODYNE COIL KIT, for three -gang variable con-denser, tuning bands, 16-50 and 66-200 meters, 540-1, 500 kc, consisting of 6high frequency coils and 2 Litz -wound I.F. transformers. Circuit diagramincluded List price, $8.00. Our price, $4.00

TWO -BAND SUPERHETERODYNE TUNING KIT for B.C. and foreign short-wave bands (530-1, 600 kc and 16-53 meters), consisting of 4 coils and 2 Litz -wound special I.F. transformers, for two -gang condenser. Complete withcoil switch and circuit diagram Price, $3.00

T.R.F. BROADCAST COIL KIT for 3 or 4 -gangThree coils for three -gang condenser, four forincluded.

SUPERHETERODYNE COIL KIT for broadcastB.C. coils and two special Litz -wound 456 kc,gang variable condenser. Diagram included

condensers. Shielded coils.four -gang. Circuit diagram

Price each coil, $0.50

receiver, consisting of twoI.F. transformers. For two -

Price, $2.50

We can supply wide choice of radio parts at lowest prices.

RADIO DESIGN COMPANY98 PARK PLACE, NEW YORK, N. Y

Free technical advice given by J. E. Anderson, if self-addressed and stampedenvelope is included with questions.


SCI"go4e'

4416,11^Nat:11111

lay.elkiM

-11111ratmdr"

126 LIBERTY ST 419RAT NEW YORK CITY

SOUND SYSTEMSPRESENTING A COMPLETE LINEOF AMPLIFIERS READY FORIMMEDIATE OPERATION

MODEL 6-A

6 -Watt Hi -Fidelity Portable AmplifierINCORPORATING 3 Tubes - 1-57, 1-2136, I -5Z3 - High ImpedanceInput-Gain 85.5 dcbls.-Fader Mixer-Microphone-Microphone Battery-

Unit built into a strong, durable carrying case.

Complete-F. 0. B. New York $29.50 Less Tubes

MODEL 12-A (Illustrated)12 WATT HI -FIDELITY PORTABLE AM-

PLIFIER.Incorporating -5 tubes -1-57, 1-53, 2-2A5s,1-5Z3-Phase inversion-Gain 105 dcbls.-Highimpedance input-Fader mixer-microphoneand full sized stand-microphone battery-Two 11'/.z in. hi -fidelity speakers, each with25 foot cable and plug-Built in speaker fieldsupply-Entire unit built into a sturdy carry-ing case 15"x15"x181,5".

Complete-F.O.B. New York-.$55.00 less tubes

MODEL 30-A30 WATT HI -FIDELITY STATIONARY AM-

PLIFIER.Incorporating -8 tubes, 1-57, 2-53s, 4-45s,I-5Z3-Phase inversion-Gain 94 dcbls.-Highimpedance input-Fader mixer-Microphoneand full sized stand-Microphone battery-Volume, and Tone control-Built in outputtransformer-Two heavy duty Hi -Fidelityspeakers.

Complete-F.O.B. New York-$59.50 less tubes

RADIO SCIENCE LABORATORIES126 Liberty St. New York City


P. A. S.5 Tube - High Fidelity - 15 WATTSUses one 57, one 56, two 2A5, one 83V.

Sold complete with case, speaker, microphone, stand, input stage, phono pickup,phono motor and connecting cables - ready to operate, nothing else to buy.

$42.50

167 GREENWICH ST. %Ft NEW YORK CITY

P. A.PRINCIPLES of

PUBLIC ADDRESSSYSTEMS

A practical handbook, full of useful information,fully i:lustrated, written by M. N. Beitman, anengineer and noted authority on audio amplifiers.

Considers in detail two -button and condenser,microphones, radio and phono input, acoustic feedback, mixing and volume controls, use of vacuumtubes. laterstage coupling, power amplifiers, output coupling.loud speaker placement, the decibel, P.A measurements,power level, and other points of Importance, including actualcircuit diagrams of P. A. systems. (Size. 8'k a 11 In.)

Either book, price

MATHEMATICSof Radio Servicing

Introduces and explains the use of arithmeticand elementary algebra in connection with units,color code, meter scales, Ohm's Law, alternatingcurrents, ohmmeter testing, wattage rating, seriesand parallel connections, capacity, inductance,mixed circuits. vacuum tubes, curves, the decibel,etc., etc., and has numerous examples.

Plainly written and easy to understand. Onlyuseful data included. M. N. Beitman, author.Size, 13%x11 in.

only 50c, post free.

SUPREME PUBLICATIONS 3727 WEST 13th STREETCHICAGO, ILLINOIS

Coil Winding Made EasyThe biggest help anyone can get who desires to windradio -frequency coils for any frequency from just be-low the audio range to the fringe of ultra frequenciesis to have a book that tells just what inductance isrequired for the condenser one possesses, and justhow many turns of any kind of wire on any sensiblediameter are needed to produce that inductance."The Inductance Authority," by Edward M. Shiepe,gives you all that information, to an accuracy of0.1 per cent. Send $2.00 and book will be mailed youpostpaid; or send $5.00 for a two-year subscriptionfor Radio World and this valuable book will besent free.

RADIO WORLD - 145 WEST 45TH STREET, NEW YORK CITY


GENERAL SHORT-WAVEAND

PUBLIC ADDRESS MANUALBound in cardinal red leatherette, this man-

ual includes articles on the construction ofshort-wave receivers from one to nine tubes,inclusive, and all values between, supers andt -r -f, with the clearest imaginable illustra-tions, both photographic and draughting.Besides the usual plain circuit diagram thereis a pictorial diagram for each circuit. Andall the photographs are informative. Portableshort-wave sets, design and winding of coilsfor short waves and broadcasts (intimatelyillustrated).

List of short-wave stations, with metersand kilocycles and hours on the air ; trouble-shooting, and forty other topics, all donewell.

The public address section contains dataon different systems, how to use them, andoffers opportunities to turn public addresswork to profit. Besides, there are articles ontesting and servicing not encompassed by thetitle of the manual-signal generators, broad-cast home and portable sets, analyzers,formulas, capacity data. Everything plainlytold, simple language, from microphones tospeakers.

PARTIAL LIST OF CONTENTSAmplification Factors 95Amplifiers 115Analyzer Unit, How to Make 129Antenna Design 80Antenna Lead-in Cable 86Antenna Wire 85Antenna, 5 -Meter 71Broadcast Receivers 88Coil Design 23, 24, 25Construction of Short -Wave Receivers 22Condenser Data Sheet 13

Condensers, Proper Replacement 12

Converting Meters to Kilocycles 19Condenser Microphone 110Crystal Microphone 114Converters:

"Regent" Two Tube A.C.-D.0 64"Paramount' Two Tube A.C-D.0 65"Adam's Four Tube A.C. Superhet 66

Curves, Explanation 100Hum Elimination 98Interference Elimination 79Morse Code 75Modern Microphone Trends 109Oscillator, Battery -Operated 134Oscillator, A.C.-D.C. 138Phonograph Combination 128Power Transformer Design 105Public Address Installation 94Servicing Formulae and Data 103Short -Wave Tuning Instructions and Data 20, 21Short -Wave Station Log 14 to 18Short -Wave Receivers starting page 26Short -Wave Power Pack 46Transceiver, Federal 70Transceiver, Powertone 68Transmitter 141Tuner, Three Tube T.R.F 125Tuner, Four Tube T.R.F 126Tuner, Superhet 127Velocity Microphones 111

Volume Controls, Electrad Replacement 132

PRICE SO CENTSSend stamp for technical book catalog.

RADIO WORLD NEWESTYORK CIYORK CITY

CONTENTS OFMONTHLY ISSUES OF

RADIO WORLDMAY NO., 1935

Metal Shell Tubes-By Conrad L. Morrow.More Gain from New Tubes-By Harry Miller.Try a Whisk Broom on Your Soldering Iron-

By Jack Tully.Plain Talk on Aerials-By Connie Andover.Boondoggling.Alignment, All Levels-By Martin A Gobin.Getting the Harmonic Habit-By Albert J. Pierce.Armchair Portable-By J. C. Wall.Tuned Radio Frequency Set with Pentode or

Triode Output-By Henry L. Maks.Portable for All Waves-By Tom Bidwell.What Air Condensers Do-By John Braden.What Pep-By Jack SullivanNeon Gadgets-By Will Ellington.S-u-z-z-e-r-By Leon Wolf.Organizing A Meter-By B. M. L. Ferris.Fundamentals Most Useful-By Herman Bernard.A Filter for High Fidelity-By Sidney Wald, E.E.

Radio University.

JUNE NO., 1935Illus. features in that issue: Coil Switch Assembly-Ultra Frequency Measure with Photocell andLamp-Seventh Metal Shell Tube-Pin Numberingon Octal Bases-Finding CX Easily-Equal Gainon T.R.F.-A Beat Oscillator-High Fidelity Tun-ing-It's the Set, not the Car-New Police Radio-Car Sets on Peak Road-False Alarms with Uni-versal Sets-Dual Range Super-Wave Trap Doesthe Trick - Measuring Low Resistance - ThreeTube S.W. Set-An Effective Transceiver-A Bat-tery Signal Generator-War on Static Succeeding-Circuits We Print Readily Duplicated-StraightFrequency Liner-The Velocity Microphone, etc.,etc.

JULY NO., 1935Illus. features in that issue: Television in

Natural Colors-Technique for 'Phone Tip Solder-ing-T.R.F. Tuner Made High Grain-A 5 MeterSending Aerial-Getting Things Into Places WhereThey Don't Fit-A Four Tube Portable-A Trans-mitter Class $750-The Saturated Pentode-A Ser-vice Wand for Light-A Marine Radio Compass-VTVM Capacity Meter-Pretuned I. F. Amplifiers-Compact Four Tube Set-Neon Tube's AverageCurrent, etc. Radio University Dept.,-HarmonicPractice for Calibrating or Checking a Generator-Articles by Capt. V. O'Rourke, Herman Bernard,J. E. Anderson, Louis Pouy, Jack Goldstein,Harvey E. Sampson, B. Herbert Russ, EarleBranker, etc., etc.

AUGUST NO., 1935Illus. features in that issue: An Electron Ray

Indicator, Ultra Wave in Wall Street; GraphicHelp to Constructor; Shield Effect on Inductance;All Hail the Grid Leak; Receiver Detuning Avoided;Short Wave Converter Factors; How to Use theNew Tubes; An Experimenter's Supe 4; HighFidelity R.F. Tuner; Television on Six Meters;The Thyratron Tube; Radio Construction Univer-sity: Articles by Neal Fitzalan, J. E. Anderson,Herman Bernard, Lionel Rivers, Arthur H. Lynch,Morris N. Beitman, L. M. Loughboro, S. G.Bradly, Maxwell M. Hauben, etc. etc.

25c A COPY. $2.50 YEARLY IN U. S.Include as part of New Subscription if you wish.

RADIO WORLD145 West 45th Street New York City

THE MICROMETER DIAL. The dial on the PW (n Concenser is of arentirely new tyoe, Leing direct reading to I part in 500. t is read exact'y like arordinary dial. Fifty division lines are engraved on the rim of the dial shell, andeach tcrth divisior is numbered in the usual way. However, the dial revolves tertimes in covering the tuning range. Each revolution the numbers change automati-cally, giving continuous readings to 500. The numbers turn with th( division lines,and mange rap;dIr at the bottom of the dial when they art, out of the operator'sline of sight.

THE PRECISION CONDENSER. The condenser is of extrem .ly rigid con-structicrs, with four bearings on the rotor shaft. Inc drive is through a preloadedworm gear of great accaracy. Each rotor is individually inst.lated from '^e frameby bakelite; each stator is individually insclated by low -loss Steatite. End bearingsare insclated for quietness. The condenser is mounted rigidly to the chassis by bosseson the base of the gear ious ng. Neither the frame nor -he dial toucies the chassisor panel a+ any point, making the assembly immune to berding of the chassis.

RANGES. The F W condensers are available in I, 2, 3 or 4 sec -ions with 150

or 225 mm= maximum per section. A new single section model now in preparationwill provide capac -ies up to 500 mmf. The plate shape of 3'1 rang is designedto give straight line frequency tuning when the frequency range is Exactly 2 to I.

NATIONAL COMPANY, INC. MALDEN, MASS.

cAre,w--cgale ed----e410,

51800NET

7ODEALER

Model No. 430

FEATURES:Tests both metal and glass tubes.

2. Applies proper load values to tube under test.3, Tests for slightest leakages and all shorts.

4 But four simple operations required.

5 Uses Triplett Model 221 meter with GOOD -

BAD scale.

6. Meter especially protected against damage.7, Sherew type A.C. meilr for line volts ad-

justment. .

Has double grid cap for metal and glass tubes.

9. Sloping, all -metal lithographed panel.

10. Furnished in handsome quartered oak casewhich you must see to really appreciate.

EVERY serviceman will want this new Readrite Model 430 be-cause it speeds up jobs and makes money on every service

trip. Ruggedly built, and designed especially to withstand theharshest treatment and all weather conditions.

Don't be caught short because of inadequate test equipment. Besure you have the Readrite Model 430 to meet all tube testingrequirements. Be prepared! See the Readrite Model 430 at yourjobber's, and write immediately for catalog describing this andother outstanding Readrite equipment.

WATCH FOR OTHER NEW ITEMSREADRITE METER WORKS

180 COLLEGE AVENUE BLUFFTON, OHIO, U. S. A.

RUSH COUPON FOR DETAILS

See1

READRITE METER WORKS_._____________________________.___

180 College Avenue, Bluffton, OhioIGentlemen:

I1936 /Please rush immediately catalog describing the new,Your ,

,,1

Readrite line.)i

Name i

Jobber i

Street1

Town Statei

.1.-.=o =0,1=0 ..1 4y,11..r fry

IHNishers' Photo Service, ii. SEPT. · 2019-07-17 · Every neighborhood can use a good part time...

Documents

Transcript of IHNishers' Photo Service, ii. SEPT. · 2019-07-17 · Every neighborhood can use a good part time...