THE WIRELESS ENGINEER · Regulations Controlling Receiving Licences-Broadcasting Stations of...

THE

WIRELESSENGINEER

ANDEXPERIMENTAL WIRELESS

NUMBER 115 VOLUME X

A JOURNAL OF

RADIO RESEARCHAND

PROGRESS

APRIL 1933

PUBLISHED BY

ILIFFE & SONS LTD.DORSET HOUSE TUDOR STREET LONDON E.C.4

April, 1933 ii THE WIRELESS ENGINEER &

NOW ON SALE

Complete Constructional Details of

WirelessP,ToarbioWorld

MONODIAL A.C. SUPERA SEVEN -VALVE SUPERHETERODYNE RECEIVER OF OUTSTANDING PERFORMANCE

By W. T. COCKING

This booklet contains in a compact form the original articleswhich appeared in "The Wireless World" some months ago

It describes at length, with full explanations and illustrations,a superheterodyne receiver for A.C. Mains, which is still

unchallenged for faithful reproduction and exclusive selectivity.

Since the Monodial A.C. Super was first described, manythousands have been built, and no single instance of failure ordisappointment with the performance due to difficulties of con-struction, has come to light. The Receiver is an ambitious one,but the design is straightforward, and a simplified method ofconstruction has been adopted.

For quality, range and selectivity, the performance is

unsurpassed by any known receiver.

Crown 4to. 24 pp. In attractive coloured cover.

PRICE1/6 net.

By post 1/8From all booksellers, or direct from the Publishers

ILIFFE & SONS LIMITED, DORSET HOUSE, TUDOR STREET, LONDON, E.C.4

Kindly mention " The Wireless Engineer" when replying to advertisers.

EXPERIMENTAL WIRELESS RI April, 1933

STANDARD -SIGNAL GENERATORSOf EXCELLENT PERFORMANCE : A COMPLETE RANGE.

We introduced the first commercially practicable Standard -Signal Generator to the Radio Industry, early in1928. Since then, all the leading manufacturers of America and Europe have designed and tested their productswith these Generators.Great improvements have been made from time to time. Present designs are of such accuracy and usefulnessthat it is difficult for us to visualise the possibility of stilibetter performance. Our present range have all beenperfected within the past six months. They will be found absolutely acceptable for Receiver Testing, FieldIntensity Surveys, and by Power Companies in their measurements of inductive interference, and Co-ordinationProblems, when measuring radio interference quantitatively.Full details regarding our Signal Generators or any of our other Radio and Electrical Measurement Apparatus(Bridges, Condensers, Oscillators, Wavemeters, etc., etc.), will gladly be sent by return, on application.

APPEARANCE. TYPE. PARTICULAR APPLICATIONS. SPECIAL FEATURES.604 Recent developments have in-

dicated that the radio -frequencytransmission band around 56Megacycles (5 -Metres) holds mostinteresting possibilities, now beingceaselessly explored by up-to-date communications companies.The 6o4 Generator is useful inthis work.

Range 3 to ioo Metres (tooMegacycles to 3 Megacycles).Has Capacity Attenuator, builtin modulation with provision forvariable external modulation.Output impedance .0002 mF:modulation input 5,000 OhmsR. -f. output : (a) Antenna t Voltand (b) at Jack, 5 to Io,000 micro-volts.

603 Extreme Precision and greatFlexibility. For very accuratesensitivity, selectivity and " fidel-ity " measurements of ultra-modern multi -valve receivers.For very accurate field -strengthmeasurements.

Range 12 to 3,000 Metres --- 25 MCto too KC. May be used evenat 12 Metres on unshieldedchassis. Output 0.5 microvoltto t.o volt. As supplied to theAir Ministry, " U.I.R.," etc.

601 Sensitivity measurements, etc.,.on batches of receivers on theproduction testing line. For air-plane receiver and airway main-tenance service. Also useful onfield -intensity work.

Moderate precision combined withextreme portability and a verylow price. Wide range (5o to3,000 Metres). Ample Output.The ideal instrument for thesmaller manufacturers.

600 For extremely rapid check-ing, where the greatest possiblespeed and the highest possibleprecision are absolutely essential.Checks only the lower Broadcastwaveband.

5 -Channel selector, giving sen-sitivity and fidelity tests at600, 800, t,000, 1,2oo and r,400 KC,and selectivity tests at 600,r,000 and 1,400 KC. Output()Jo to 316,000 microvolts (=-26 to +104 DB).

PROMPT DELIVERY : REASONABLE PRICES : ALL ENQUIRIES TO-

&ADE 61f©MS, LTDHead Offices : 76 OLDHALL STREET, LIVERPOOLLondon Offices : 40 BUCKINGHAM GATE, S.W. 1

e413..

" I f a man write a better book, preach a better sermon, or make a better mouse -trap than his neighbour, though hebuild his house in the woods, the world will make a g eaten path to his door."-RALPH WALDO EMERSON.

Does not this explain the outstanding acceptance and approval of our " G.R." Instruments?

Kindly mention The Wireless Engineer- when replying to advertisers. A

April, 1,)33 R2 THE WIRELESS ENGINEER Fe

. . . . . for everyCONDENSERrequirement

everyCONDENSERneed, consultT.C.C.

Illustrated on ; right, Type 101Condenser, for maximum workingvoltage of 800 peak value (tested1,500 V.D.C.). Price 91-.

With unceasing research work always directed to the improve-ment of the one product-condensers-T.C.C. are specifiedand used with confidence wherever a condenser of unquestionablereliability is demanded.Radio Technicians, Set designers, Broadcast stations-World'sCable Companies and many of the leading industrial concernspin their faith to T.C.C.

The TelegraphCondenser Co. Ltd.,Wales Farm Road,N. Acton, W.3.

T.C.C.ALL -BRITISH

CONDENSERS253,5

Compiled by the Staff of "The Wireless World."

T'WirelessWorld

DIARYfor 1933

This handy little pocket volume is now in its ninth yearof publication. As a convenient source of informationon many aspects of wireless theory and practice it isunique. In addition to the usual Diary pages-oneweek at an opening-there are 78 pages of facts, figuresand explanations to which wireless amateurs makefrequent reference. These include a series of 16 up-to-datecircuit diagrams of complete receivers and accessoriessuch as Band Pass Filters, Volume Controls, Eliminatorsand Rectifiers, and Methods of applying Reaction.

LIST OF CONTENTSRegulations Controlling Receiving Licences-BroadcastingStations of Europe-Principal Short -Wave BroadcastingStations of the World-Useful Formula'-A Series of Abacs-Copper. Wire Tables; Sizes, Weights and Resistance-Con-version Tables- Symbols in Common Use-Symbols Used inWireless Diagrams-Circuit Diagrams and Notes-TracingHum in Mains Receivers-Coil Construction Data-PracticalHints and Tips-Matching Output Valve and Loud Speaker-

Morse Code-Valve Data.

Size n x 41 inches, 192 pages. Bound leather cloth,backloop with pencil, round corners and gilt edges

PRICE 1'6By Post 1/7

From all Booksellers, Stationers and Bookstalls or direct from the Publishers

ILIFFE dc SONS LIMITEDDorset House, Tudor Street, London, E.C.4

Kindly mention " The Wireless Engineer when replying to advertisers

EXPERIMENTAL WIRELESS 1,)3

PRECISIONLABORATORY APPARATUS

by

The Type 11

Variable AirCondenser isfully describedin our BulletinB -231-A.

TYPE 11

VARIABLE

AIR

CONDENSER

CIThe Muirhead Type 11 variableair condenser is a precisioninstrument of particularly

robust construction, capable ofwithstanding constant use withoutchange of calibration.

Quartz insulation ensuresminimum electrical losses andmaximum mechanical rigidity.

Worm drive of special designprovides subdivision to 1 partin 5,000.

MUIRHEADMAKERS of PRECISION INSTRUMENTS for OVER 50 YEARS

ELMERS END KENTKindly mention " The Wireless Engineer" when replying to advertisers. A2

April, 1933 THE \ IZ ELESS ENGINEER,It2

1 HETERODYNE WAVEMETERS OF ALL RANGES AND ACCURACIES

.i.

(GREATLY REDUCED

ii

I i

iI

R.1100 50 - 7000 metres, i025% accuracy,

reduced to £35.

1

I

ALL PRICES 1

CALIBRATION i

R.12)0

in PRICE and may be DELIVERED

R.1050.. .50 -7000 metres,0.1% accuracy,

reduced from£75 to £57.

...INCLUDE...

FROM STOCK

R.1200.... 18 - 300 metres,

0.2% accuracy,

reduced from£70 to £55.

H. W. SULLIVAN, LTD.,WRITE FOR DESCRIPTIVE LEAF LETS Leo St. London, SEAS9

ELECTROSTATIC MICROPHONE FOR MEASURINGPURPOSES.

Practically equal sensitivity from 5o to 12 000 cycles. Not subjectto changes in temperature nor mechanical influences. Amber insula-tion. Sensitivity about 2 millivolts per bar.HIGHLY SENSITIVE MAINS -OPERATED VALVE -

VOLTMETER.Measuring range from .03 to .6 volts r.m.s. Independent of frequencyfrom 5o to to million cycles. Input capacity 20 micro -micro -farads.Enclosed in metal case.

SYNCHRONISABLE STROBOSCOPIC SOURCE OFLIGHT FOR THE INVESTIGATION OF

OSCILLATING PHENOMENA.Frequency variable from 20 to 3000 cycles. Flash duration onlyabout 2% of one complete cycle, therefore great clarity of outline.The apparatus is, for instance, suitable for the investigation ofloudspeakers and film transporting mechanisms.CATHODE-RAY TUBES FOR MEASURING PURPOSES.New design. Reduced prices.

ACCESSORY APPARATUS FOR CATHODE-RAYTUBES.

All -mains apparatus, deflecting potential oscillators, registeringapparatus, stands. Reduced prices.

HIGH FREQUENCY POTENTIAL -DIVIDERS, OSCIL-LATOR -GLOW -LAMPS, SOUND -FILM

RECORDING TUBES.Detailed catalogues with illustrations and particulars on application.

Development and manufacture of special measuring instruments forlaboratories and general use.

MANFRED VON ARDENNE,Research Laboratory,

Berlin-Lichterfelde-Ost, Jungfernstieg 19.Telephone : Lichterfelde G3.o224.

Telegraphic address : Ardolabor, Berlin.

JOURNAL OFSCIENTIFIC INSTRUMENTSproduced by the Institute of Physics withthe co-operation of the National PhysicalLaboratory. The English journal devotedto scientific instruments, their prin-ciples, construction and use. Essentialto designers, makers and users.

CONTENTS OF THE MARCH NUMBER.REMOTE INDICATION AND CONTROL DEVICES.

By G. F. Tagg, B.Sc., A.M.I.E.E.A MAGNET FOR a -RAY SPECTROSCOPY. By J. D.

Cockcroft, Ph.D.DAMPING AND RESONANCE WITH AN EINTHOVEN

STRING GALVANOMETER. By A. H. Bebb, M.Sc.DESIGN OF AN APPARATUS FOR CATHODE SPUT-

TERING. By James A. Derbyshire, M.Sc.NEW INSTRUMENTS

APPARATUS FOR FLUORESCENCE MICROSCOPY.A NEW OPHTHALMOSCOPE AND REFRACTION UN IT

LABORATORY AND WORKSHOP NOTES:MAGNIFICATION SCALES FOR A CONVEX LENS.By E. M. Eden.

VEE PULLEYS. By G. R. Myers.CORRESPONDENCE, NEW BOOKS, CATALOGUES,

NOTES AND COMMENTS, INSTITUTE NOTES.The subscription rate for a yearly volume is £1 10s. net (post

free) payable in advance. Separate parts, 2s. 8d. (post free).Subscriptions should be sent to the publishers.

CAMBRIDGE UNIVERSITY PRESSFetter Lane, London, E.C.4 - England

Kindly mention " The Wireless Engineer" when replying to advertisers.

A5

The

WIRELESS ENGINEER

II

El

and EXPERIMENTAL WIRELESSA Journal of Radio Research & Progress

Editor Technical EditorHUGH S. POCOCK Prof. G. W. 0. HOWE D.Sc., M.I.E.E.

VOL. X No. 115

APRIL 1933CON T EN T S

EDITORIAL 179

IRON - CONTENT CORES FOR HIGH -FREQUENCY . COILSBy Alfred Schneider 183

BROADCASTING ON 7.85 METRES By P. J. H. A. Nordlohne 186

A MAGNETRON OSCILLATOR FOR ULTRA -SHORT WAVE-LENGTHS By E. C. S. Megaw, B.Sc., D.I.C. 197

CORRESPONDENCE 202

THE RESEARCHES OF THE LATE DR. D. W. DYE, ONTHE VIBRATIONS OF QUARTZ 205

ABSTRACTS AND REFERENCES 206

SOME RECENT PATENTS 233

Published Monthly on the first of each monthSUBSCRIPTIONS Home and Abroad : One Year, 32/-, 6 Months, 16/, Single Copies, 2/8 post free

Editorial Offices116-117 FLEET STREET, LONDON, E.C.4 Telephone: City 9472 (5 lines)

Advertising and Publishing OfficesDORSET HOUSE, TUDOR STREET, LONDON, E.C. 4

Telegrams: " Experiwyr Fleet London" Telephone: City 2846 (x7 lines)Branch Offices

COVENTRY BIRMINGHAM MANCHESTER GLASGOW19 Hertford Street Guildhall Bldgs., Navigation St., 2 26o Deansgate, 3 26a Renfield Street, C.2

Telegrams: "Cyclist, Coventry" Telegrams: " Autopress. Birmingham" Telegrams: [litre. Manchester" Telegram.: Glasgow"Telephone: 5210 Coventry Telephone: Midland 2970 (9 lines) Telephone: Black! riare 4912 (4 line4) Telephone: Central 4857

The Editor invites the submission of articles with a view to publication. Contributions which arenot exclusive should be so described when submitted. MSS. should be addressed to the Editor,' The Wireless Engineer and Experimental Wireless," Dorset House, Tudor Street, London, E.C.4Especial care should be taken as to the legibility of MSS. including mathematical work.

April, 5933

BUY THERESISTANCE

OFPROVED

PERFORMANCE

f}6 THE WIRELESS ENGINEER

LOADCHARACTERISTICS

1.0 MEGCHM

0

Js

When tested withthe rated loadapplied for 16hours each day,the resistance )value alwayscomes buck to theinitial Value after 1°cooling. The"hot"resistance L.5also remains con-stant over longperiods.

DUBILIERCONDENSER

Co. (1925) Ltd.Ducon Works,Victoria Road,

N. Acton,London, W.3.

100.000 0 this

a 1.

You can it afford to incor-porate inferior Resistancesin your set! Besides, it'snot worth the risk when youcan buy Dubilier MetallizedResistances and be certainof getting infinitely betterresults ; results that leaveno room for complaints orregrets, because DubilierResistances are subject tothe most exacting tests be-fore being released for sale.There are Resistances tomeet your every need in theDubilier range. See yourlocaldealer about these corn-ponents-not only will heconfirm our claims, but willbe enthusiastic about them.

DUBILIERMETALLIZEDRESISTANCES

A

NEW

VOLUME

CONTROL

Volume controls have had a badname for being unreliable-andmany of them have deserved it,but the new Rotor -ohm is onethat you can fit and forget.The safety factor (dissipation).is much above average workingconditions ; the special rubbingcontact will give unlimited wearand is dead silent in operation.

Prices are: Wirewound,all values to 2.5,000ohms 3/6, 50,000 ohms4/-. Synthetic, 2,000ohms to 5 meg-ohms 3/6.Switch models II- extra.Ganging 9d. a unit extra.

ROTOR ELECTRICLIMITED

Spencer House, South Place,E.C.2.

Telephone : National 5734.

ROTOR OHM VOLUMECONTROL

Manufactured in the British Commonwealth, produced in England.

PHOTO -ELECTRIC CELLSOF EXTREME SENSITIVITY,PLUS GREAT STABILITY.

SALIENT DATA:-' These Cells are the logical result of over eight

years intensive research and specialisation in aFactory producing nothing else than all types ofPhoto -Cells. Our illustration shows our mostpopular style -Type 59-A. This is a genuine mona-tomic Caesium, gasfilled Cell having a sensitivityrunning as high as 150 MicroampslLumen. It hasan extremely long life, is very stable, and has avery good frequency response. Seepage is preventedby paying great attention to sealing. The insertedmetal plate construction is a great improvementover those types in which the light-sensitivesurface is deposited on the inner walls of the glassenvelope.

Complete Data on our entire line on request.Request Bulletin No. I19 -B. Prices are extremelylow and large stocks 'are available at all times.Special terms on contracts, or for Export.

CLAUDE LYONS, LIMITED.(FOR ADDRESSES SEE ADVERT. PAGE I)

Kindly mention The Wireless Engineer- when replying to advertisers.

THE

WIRELESSENGINEER


Vol,. X. APRIL, 1933. No. J.15

Editorial

The Principles of ElectromagnetismIN our last number we published a letter

from Professor Fortescue dealing withsome of the points which were discussed

in the January Editorial, and in the presentnumber we publish letters from Mr. C. R.Cosens and Prof. Cramp dealing with thesame subject. From these and from otherletters which we have received, it is evidentthat the subject is one in which many peopleare deeply interested, and we therefore feeljustified in returning to it.

We are pleased to note that Prof. Fortescueagrees with our criticism of the oft -quotedformula f = m1m2/µr2 and we must confessthat if the unit pole is not misused in thisway, but is used as a rigorously defined con-cept in air, we have no objection to it. Itcertainly lends itself to a simple treatmentof many magnetic field problems, and weuse it ourselves for this purpose. Ourobjection to the meaningless statement thatthe above formula applies when the poles aretransferred from one medium to another isemphasised by Mr. Cosens in his letter.

A Unit Pole without IronThere is a form of unit pole which can be

transferred from one medium to another

without any uncertainty as to its behaviourand which has other advantages over thehistorical unit pole of a permanent magnet.If a rod of square section-it need not besquare, but this form lends itself to simplecalculation-be wound uniformly with fineinsulated wireand the rod bethen removed,we shall havea coil similar tothat shown inFig. I. Weassume that thelength is greatcompared withthe side of thesquare section,and that theside is small compared with the centimetre.The leads to and from the coil can beused as a bifilar suspension. Two suchcoils can be made and, the current ineach being the same, this current can beadjusted until the two like poles repel eachother when the coils are placed i cm. apartin the same straight line with a force of onedyne. So long then as we keep this current

1

Fig. I.

April, 1933 18o THE WIRELESS ENGINEER &

unchanged our coils have unit pole strength.It may occur to the reader that this wouldbe one way of defining unit current. If 1 bethe length of the coil, a the side of the squareand T the number of turns, and we make thenumber of turns per cm. T/l = ila2, andthen carry out the above adjustment ofcurrent, we may define this current as ourunit of current, but we shall be introducingno new unit, for it will be the old electro-magnetic unit of 10 amperes. If the squarerod has z mm. side it must be wound withioo turns per cm. which sounds somewhatimpracticable, but we will call such a coila standard coil.

Now let one of our coils so constructed beplaced at right angles to a magnetic field asshown in plan in Fig. 2. All the vertical

Fig. 2.

wires will experience forces as shown bythe arrows and every force acts at an arm ofa/2 about the centre of the coil. If I be thecurrent in amperes, the force on each wirewill be BaI and the total torque will beBar lio x Ta = Ba2TI/io where B is themagnetic induction of the uniform field inwhich the coil is placed. If the coil is astandard coil with a2T =1, then the torque

BlI/io. Now let us come into line withthe more familiar fiction and assume thatthe torque is really due to somethingcalled a quantity of magnetism m situatedat each end of the coil and acted on by aforce mH in the direction of the field. Wecan then equate this fictitious torque to thereal torque and thus find m.

InH1= Ba2TI/io .*. m= -12H

ITa2

101

B Iand in a standard coil m= -H. - /.4 fH io IoWe have assumed m to be at the ends of thecoil ; in referring to permanent magnetsthe poles are often stated somewhat vaguely

to be near the ends. If we consider themagnetic flux produced by the coil itself wehave

cf, Bca2 = pa2H,=- 12.2 IT---io 1

which is seen to be 4irm.If then the current be maintained constant

the torque on such a coil will be proportionalto the magnetic induction B of the field inwhich it is placed, but if on transferringit to another medium of different per-meability, the value of the current be alwaysadjusted until on reversal it gives the samedeflection as before on a ballistic galvano-meter connected to a coil wound around themiddle of the coil, then 4 is constant, andtherefore also m, and the torque on the coilwill be proportional to the magnetising forceH of the field in which it is placed. Thisappears to be the assumption made for theunit pole of a permanent magnet, althoughin that case the change of medium is onlyexternal to the magnet.

The Nature of H and BWith regard to the vexed question of H and

B Prof. Fortescue's letter shows that he doesnot realise the point of view of those whoprefer to regard them as representing thingsof different natures, nor the reasons thatled the International Electrotechnical Com-mission to give different names to the unitsof H and B. We suspect that the member ofthe Commission to whom Prof. Fortescuerefers was seeking an excuse for avoidingfurther discussion when he gave the humour-ous reason for the action of the Commission.To many people the assertion that H and Bare of the same nature savours of a claimto know more of the secrets of nature thanhave been revealed to the rest of us.

By the statement that the relation betweenthe magnetising force H and the magneticinduction B can be regarded as in any wayanalogous to the relation between stress andstrain, it must not be thought that it issuggested that magnetising force is in anyway of the nature of stress, or magnetic in-duction of the nature of strain. We do notknow enough of the real nature of either towarrant any such improbable assumption.What is suggested is that the mental conceptsto which we give the names of magnetisingforce and magnetic induction are related toone another in a similar way to the mental

EXPERIMENTAL WIRELESS 18,

concepts of stress and strain-perhaps it willmake our meaning clearer if we add-or tothe concepts of force and acceleration. It isa relationship between two related butdifferent concepts which it is often con-venient to regard as cause and effect.

A mechanical force F acts upon a testpiece and causes throughout the material aphenomenon called strain to which a definitenumerical value and direction can be assignedat every point. Mechanical engineers intro-duce a concept to which they give the name" stress " and assume it to exist at every pointof any cross-section such as that showndotted in Fig. 3. The actual force F is theoriginal cause of the whole phenomenon, butso far as this dotted cross-section is con-cerned it is replaced by this concept of stressintegrated over the whole area of the cross-section. (For the sake of simplicity weneglect the cosines of angles which enter intothe integration.) The stress at each pointis related to, and is often regarded as thecause of, the strain at that point. The natureof this concept of stress is that of a forcedivided by an area.

If the mental procedure which we havejust outlined, and which is universallyadopted in the study of mechanics, isallowable, are we not at liberty to followthe same mental procedure in consideringthe electromagnetic problem ? We em-phasise again that it is not a physical analogywhich we are suggesting but the analogyof formation of concepts. If the followingwording seems familiar a glance at the fore-going paragraph will show why.

An electric current Iacts upon a ring of iron(or wood, or air, orvacuum) and causesthroughout the materialor space a phenomenoncalled magnetic induc-tion to which a definitenumerical value anddirection can be assigned

Fig. 3. at every point. Electricalengineers introduce a con-

cept to which they give the name " magne-tising force " and assume it to exist atevery point of any closed curve such asthat shown dotted in Fig. 4. The actualcurrent I is the original cause of thewhole phenomenon but so far as this

April, 1933

dotted line is concerned it is replaced bythis concept of magnetising force inte-grated over the whole length of the path.(For the sake of simplicity we neglect thecosines of angles which enter into the in-tegration.) Themagnetising forceat each point isrelated to, and isoften regarded asthe cause of, themagnetic induc-tion at that point.The nature of thisconcept of mag-netising force isthat of a currentdivided by a length.

Just as the loadF is replaced overthe cross-section by a distributed system offorces, so the coil carrying the current I isreplaced over the closed curve by a dis-tributed system of current -carrying coils.Just as the stress will not be distributeduniformly, but the force per unit area atany point will be related to the strain at thatpoint in a manner depending on a property ofthe medium known as Young's modulus, sothe magnetising force H will not be distri-buted uniformly, but the ampere -turns percentimetre at any point will be related to themagnetic induction at that point in a mannerdepending on a property of the mediumknown as permeability (the Committee of theGerman V.D.E. proposed the name " in-duction constant ").

It is open to anyone to refuse to use theconcept of stress in the mechanical case orthe concept of magnetising force in theelectromagnetic case and to maintain thatthey can get on quite well without it.People sometimes boast that they can per-form some feat with one hand tied behindthem, but cui bono ? It is also open to any-one to assert that he has been granted adeeper insight into the secrets of nature andthat it has been revealed to him that theapparent cause and effect-whether stressand strain or magnetising force and magneticinduction-which appear so different toother people are really of the same natureand only differ from each other as a footdiffers from a yard.

In his letter Prof. Fortescue says that

Fig. 4.

April, 1933

" stress and strain admit of independentmeasurement --the one by the load and theother by the extension-whereas in themagnetic case there is only one possiblemeasurement, i.e., the mechanical force onan elementary pole." We maintain that ineach case there are three quantities capableof measurement ; in one they are force orload, area, and strain or total extension, inthe other, current, length, and magneticinduction or total flux. Not only are thesecapable of measurement but they are ineach case the three magnitudes which areactually measured, or deduced from thereadings of spring balances, extensometers,ammeters, flux meters and foot -rules.

What we have outlined here is the justi-fication for what we are pleased to see Prof.Fortescue refers to as the more usual point ofview, and although he claims to have beendriven to a contrary point of view by " thelogic of the situation," we would suggestthat logic is like the railway ; it will takeyou anywhere depending on the train youhappen to get into.

With regard to his closing remark that thestress -strain analogy unlike the conceptionof the magnetic particle does not stand uponobjective experiment, we will leave ourreaders to form their own opinion whetherwhat we have outlined above-which,although parallel with, is entirely indepen-dent of any stress -strain analogy-or Prof.Fortescue's magnetic particle stands themore firmly upon objective experiment.

Mental Concepts and Physical AnalogiesWe have emphasised that we do not suggest

that H is of the nature of a stress nor B of thenature of a strain. If one were to attempt toderive from a comparison of electromagneticphenomena such as wave propagation andthe storage of energy in electric and magneticfields with the corresponding mechanicalphenomena an analogy between the two setsof concepts, one would certainly look for theanalogy to the static phenomena of stress

182 THE WIRELESS ENGINEER &

and strain in the static phenomena of electricforce and electric displacement. This wouldsuggest that the electromagnetic analogy toYoung's modulus is in the permittivity ofthe medium, or rather its reciprocal and notin the permeability.

If mechanical force is applied to matterwhich can move freely without resistance itcauses a uniform increase in the velocity,and the relation between the acceleration andthe applied force depends upon the dimen-sions of the body and a property of thematerial known as its density. Similarly, ifelectromotive force is applied to a perfectlyconducting circuit, i.e., to electricity whichis free to move without resistance, it causesa uniform increase in the current and in themagnetic induction at every point, and therelation between the rate of growth of thecurrent and the applied electromotive forcedepends upon the dimensions of the circuitand a property of the medium known as itspermeability. It is interesting to note thatthe rate of increase of B depends only on theapplied electromotive force and is indepen-dent of the medium, whereas the rate ofincrease of the current and therefore of H isinversely proportional to the permeability ofthe medium. This suggests that B wouldbe more correctly associated with the cause,and H with the effect which depends on thenature of the medium. This reversal of theordinary conception has been pointed outby Livens, but we do not think that anyuseful purpose would be served by followingup this analogy.

In conclusion we wish to say that, howeverpossible it may be to do so, we feel sure thatany attempt to merge the various pairs ofconcepts which we have considered --whetherit be stress and strain, force and acceleration,or magnetising force and magnetic induc-tion-into one, would be a retrograde step inthe evolution of the system of concepts bymeans of which we describe and calculatemechanical and electromagnetic phenomena.

G. W. 0. H.

EXPERIMENTAL WIRELESS 183 April, 1933

Iron -Content Cores for High -frequency CoilsBy Alfred Schneider

THE idea of the employment of magneticcore materials for high frequency coilshas recently come to the fore again

through the development by Hans Vogt ofthe so-called "Ferrocart " material, sothat in view of the interesting editorial inthe January issue of Wireless Engineer, somefurther contribution about the theoreticalconsiderations which led to this remarkablenew magnetic material may be of interest towireless engineers.

Theoretically speaking, the conclusiongiven in the editorial, that very thin ironwire, insulated and wound in the direc-tion of the magnetic flux, should givethe highest permeability at lowest losses, isdoubtless correct, but also the doubts of theeditor as to price and possibility of manu-facturing such very thin wire, are perfectlyjustified.

them and of making cores of such wire,proved to be so enormous, however, thatsuch a material could not be produced at amarketable price.

Vogt therefore tried to attain the sameeffect in another way, and, indeed, by hismethod a material was created, of a structure,which in its effect is almost identical with theideal one of insulated magnetic filament,although for technological reasons he startsfrom iron particles and not from filaments.

The structure of the Vogt iron (Ferrocart)will be clear from Fig. i. The insulated ironparticles, having a shape similar to an ellipsoidand being of a very small size, are arrangedin a very thin layer on an insulating paperstrip, small channels are then made so as tosub -divide the continuous layer into smallstrips. These two stages, continuous layerand sub -division into channels, are shown

ENLARGED VIEW SHOWINGIRON PARTICLES ALIGNEDAND DRIED IN POSITION AFTERPASSING THROUGH D.C.COIL

11111 111111111 11111

11111111 11111110 11111111

111111 111111111 111111

ENLARGED VIEW SHOWINGSEPARATION OF IRON

PARTICLES INTO ROWS.

TO D.C.

SUPPLY

COMB SEPARATING IRON PARTICLESINTO ROWS ON INSULATING PAPER STRIP.

DIRECTION Of TRAVEL

D.C.SOLENOIO COIL

PARTICLES ALIGNED AND DRIED IN ALONGITUDINAL POSITION AFTERPASSING THROUGH D.C. COIL

INSULATED IRON PARTICLES

INSULATING PAPER STRIP

ENLARGED VIEW SHOWING IRONPARTICLES ON PAPER STRIP

Fig. I.-Illustrating the processes in the manufacture of Ferrocart

As a matter of fact, the inventor of Ferro -cart, when starting his researches, thoroughlyexamined the problem of making the materialof insulated iron filaments ; the calculationsproved, however, that the diameter of suchwire would have to be less than 5µ, i.e., 0.005mm. The difficulties of drawing filamentsof this diameter out of the special alloysrequired for high frequency, of insulating

in magnification. Now, by means of aD.C. solenoid coil through which the paperstrip passes in its further course, the semi-liquid iron layer is solidified, the particlesbeing all aligned in the longitudinal directionof the paper strip and solidified in thisposition. Thus, many continuous chains ofparticles, being interrupted by the thininsulating skin of each particle only, are

April, 1933

obtained in the longitudinal direction. Ascan be seen from Fig. 2, the coil cores arebuilt up of many layers of such paper strips,which are jointed by a hot liquid binder. Thepaper strips are so arranged in the core, thattheir longitudinal axes corresponds to thedirection of the magnetic flux in the coil.

From theabove explana-tions it will beclear that inthis case it isactually pos-sible to base thecalculation onthe assumptionthat ellipsoidparticles are inquestion, havingtheir longer axesin the directionof the magneticfield. In factthrough the particular arrangement and mag-netic aligning the effect of insulated magneticwire is so to say" imitated," the more so as themagnetic alignment also causes the particlesto separate transversely due to the repellingforce developed by their identical magnetisa-tion. In addition, the magnetic alignmenthas a further favourable effect in so far as theobjectionable end surfaces of the particles,which are difficult to insulate and tend todestroy the insulating skin of the neighbour-ing particles, are all directed so as to be inthe longitudinal direction, where a metalliccontact not only is without danger withregard to eddy current (as the eddy currentelectromotive forces are induced only in thetransverse direction parallel to the inducingwindings), but even desirable to improvethe magnetic conductivity. Now, the ques-tion may be raised why, if each particle isinsulated individually and aligned, theadditional sub -division by paper and groovesis necessary. There are two reasons fordoing this.

- First, as a matter of fact, the insulatingskin around each particle is very thin andmust be very thin in order to have a suffic-ient proportion of iron in the compositionto obtain a reasonable permeability. In-creasing the insulating skin by only a fewmillionths of a millimetre considerably re-duces the effective permeability, because the

Pain 11/

II/

Fig. 2.-Showing how Ferrocartis built up.


proportion of iron is decreased, and the airgaps between the particles are augmented.Therefore the particles cannot be absolutelyinsulated from each other, although thecontact resistance between the particles israther high. Accordingly, a composition ofsuch particles, without additional sub-division, represents a sort of semi -conductor,having a conductivity intermediate betweenthat of good electric conductors and goodelectric insulators, as has been statedby one writer.* Accordingly, such com-positions still represent a certain path foreddy currents, and, due to the relatively greatelectromotive forces, induced by the highfrequency currents, especially with fre-quencies beyond 500 kilocycles, considerableeddy current losses are produced. To pre-vent these losses it has been proposed toremove the core with increasing frequency,thus effecting the so-called " permeabilitytuning."* This method, however, has notfound its way into practice for variousreasons. Vogt, on the contrary, as explainedabove, has found a definite solution of theproblem of creating a material with lowlosses, even at frequencies beyond 50o kilo-cycles, by preventing the eddy currents fromextending over the whole magnetic core byadditional sub -division in the form of inter-mediate paper layers and grooves in themagnetic layers.

The intermediate paper layers and thegrooves, however, have still another functionthan forming an interruption for .galvaniceddy current, and this is the most remarkablenew feature of the Vogt iron. As a matterof fact, many authors hitherto, when dealingwith the question of magnetic compositions,have expressed the opinion that the lossescannot all be ascribed to hystereses and eddycurrents, but that an unknown additionalsource of losses exists, the nature of whichcould not be determined. Also in theEditorial in the January issue, these addi-tional losses are referred to and the hypo-thesis is mentioned according to which thelosses might be due to mechanical movementof the particles in the insulating matrix.

From the investigations made in con-nection with the Vogt iron, it became obviousthat such additional losses actually existand represent a considerable share of the

* R. H. Langley, Proceedings of the Radio Club ofAmerica, April, 1932.

EXPERIMENTAL WIRELESS 185

total losses in the magnetic core. Mechanicalmovements, however, proved not to be thecause, and, indeed, it is quite improbablethat there should occur any mechanicalmovement of the particles in the matrix inconsideration of the fact that the particles,which of course could be caused to move onlyby the changing magnetic field, would haveto make about a million oscillations persecond in both directions in the case of highfrequency induction coils ; that means, theywould practically remain in their originalposition like the pointer of a D.C. measuringinstrument when connected to A.C. Indeed,the actual source of the additional losses isto be found from quite a different and newelectro-physical consideration. There areproduced in the core not only galvanic eddycurrents, flowing in each individual particleand extending over a group of particles beingin contact with one another, but also eddycurrents, due to capacitive coupling, whichwere termed " capacitive eddy currents."Each two magnetic particles, which areseparated from each other by the insulatingskin, form a little condenser, the insulatingskin representing the dielectric and the twoparticles representing the two condenserplates. Thus a capacitive coupling is formedand the eddy current, which cannot flow in theusual galvanic way due to the insulating skin,passes as a dielectric displacement currentfrom particle to particle. This is, of course,possible only in the case of high frequency,where the capacitive effects are very con-siderable.

Low Losses ExplainedThis is the fundamental new discovery on

which the Vogt iron is based, and, havingdiscovered this source of losses, it was pos-sible to take steps against them ; thus theadditional sub -division was so dimensioned asto form not only an interruption for galvaniceddy currents but also to form a high " capaci-tive resistance " between the various groupsof particles. In fact, measurements showedthat although a thin intermediate paperlayer was perfectly sufficient to insulate onelayer of magnetic composition against thenext one, the losses were further reduced bymaking the paper thicker (this is identical toincreasing the distance between two con-denser plates to increase the capacitiveimpedance). Also the dielectric conduc-

April, 1933

tivity of the insulating layer and of theinsulating skin of each particle has a con-siderable effect.

This phenomenon is the principal reasonfor the low losses of the Vogt iron, and thisalso will explain why other inventors,although using very finely divided and wellinsulated iron particles, could not attain adefinite success especially in the short waverange 200-600 m, but had to make the coremovable, withdrawing it from the coil atthese shorter waves.

Further ObservationsIn this connection, some other observations

made during the researche§ in connectionwith Ferrocart may he of interest. First,the conclusions in the January Editorial,regarding the maximum attainable resultingpermeability and regarding the influence ofthe specific permeability of the magneticparticles are perfectly verified by theseinvestigations. Considering the fact thatwith regard to the above explained capacitiveeddy currents the iron must be finely dis-tributed, the share of magnetic material mustbe rather low, and therefore the specific per-meability of this material is of quite aninferior importance due to the influenceof the many small air gaps between theparticles. In fact, the resulting permeabilitycannot be increased much beyond i8 withoutincreasing the losses to an extent far out-weighing the advantages of the higher per-meability.

Another question, often raised with regardto such magnetic compositions, is whether thepermeability is constant over a wide range ofdifferent loads. Indeed this question is ofparticular importance in case of tuning coils,for any change of the permeability wouldmean a change of the inductance, in otherwords, of the syntonisation. These objec-tions are doubtless justified and in fact in thecase of transmitter coils due care must betaken that the saturation current is notreached. In case of receivers, however, themagnetic fields are so extremely weak thatthe magnetic induction is strictly propor-tional to the magnetic field intensity, in otherwords, the permeability is constant.

I hope to have contributed in someway to elucidating the actual conditions,effects and properties of magnetic composi-tions under high frequency conditions.

A,

April, 1933 186 THE WIRELESS ENGINEER &

Broadcasting on 7.85 Metres*Experimental Work in Amsterdam

By P. J. H. A. Nordlohne(Physical Laboratory, N. V. Philips' Gloeilampenfabrie¢en, Eindhoven, Holland)

FROM November, 1930, till October, 1931,the Physical Laboratory of N. V.Philips' Gloeilampenfabrieken, Eind-

hoven (Holland), carried out experiments onbroadcasting on a wavelength of 7.85 m.,first in Eindhoven, and then, from April1931, in Amsterdam.

These experiments were intended as atest of the usefulness of such a wavelengthfor local broadcasting. That this broad-casting remains local, is due to the fact that :

1. Oscillations of such a high frequency(of the order of 40,000 kc/s.) are not reflectedback to earth from the Kennelly -Heavisidelayer ;

2. The direct rays cannot follow thecurvature of the earth -surface to any greatextent.

The absence of rays reflected from theupper atmosphere (the so-called indirectrays) has the advantage that fading and dis-tortion, often involved thereby, do notoccur. The reliability of the communica-tion is the same during the whole of the 24hours and is only dependent on local sur-roundings. Experience has taught that thecondition of the lower atmosphere has noinfluence on the transmission over a fewkilometres as required for local broadcastingin cities.

The experiments for testing the usefulnessof such short wavelengths for local broad-casting involved the study of the emissionas well as the transmission and the receptionwhich will all be treated separately here.

As far as we know the Philips experimentsare the first of that kind conducted in Hol-land, but we are aware that much researchhas been carried out abroad.

Dunmore and Engel (1) in 1923 reportedon experiments with a wavelength of 10metres and called attention to the nearlycomplete absence of atmospheric distur-bances.

* Lecture held before the Nederlandsch RadioGenootschap at Amsterdam on February 13th,1932.

Wells (2) in 1925 described experiments ofthe Marconi Co. in 1922 and following yearswith a wavelength of 4 to 6 metres nearInchkeith. Also the French experimentsfrom the Eiffel Tower are well known ; theywere described by Beauvais and Mesny (3)in 1927. Further, the publication of Jouaustand Beauvais (4) relating to experimentsbetween Fort de la Reve near Nice and theCol de Tegbirie on Corsica is well known ;a wavelength of 5 metres was used for thiscommunication. The last-named authorsclaimed a range over sea that was 3o percent. more than could be calculated from aconsideration of the curvature of the earth -surface. They observed a pronounced de-pendence of the field on the condition of thelower atmosphere, as a result of an elevationor sinking of the electrical horizon, due toscattering.

Very much experimental work has beendone by the Radio Corporation of Americasome years ago (5), using frequencies from30,000 to 6o,000 kc/s. These experimentsled to a radiotelephonic communicationamong islands of the Hawaian group, cover-ing distances of 200 miles with a wavelengthof 8 metres, so that this communicationcould be connected to the internationalradio -telephone network. The transmittingas well as the receiving aerials were builton high hills or mountains. Radiotelephonywas used because of the difficulty and costof laying cables in the very deep seas separa-ting the islands. Experiments with aero-planes were also carried out. A ferry -boatover the Hudson was equipped with aninstallation for wireless navigation in foggyweather with much success. In connectionwith these latter experiments attention wascalled to the possibility of local wirelessbroadcasting provided the transmittingaerial has an elevated position.

Interesting experiments also were thosecarried out in 1929 by Gerth and Schepp-mann (6), in aeroplanes and from theBrocken in the Harz Mountains (Germany).

EXPERIMENTAL WIRELESS

Experiments have been carried out byDr. Schaeringer, in the neighbourhood ofSalzburg, but not published so far as I know.

Very important are the researches ofProf. Dr. A. Esau (7), Physical Institute inJena (Germany). Prof. Esau perhaps is thefirst who systematically tested the usefulnessof wavelengths below io metres for broad-casting in cities, for which purpose alreadyin 1929 experiments were made in Chemnitz,repeated in 193o with the aid of a multi-stage transmitter built by the C. LorenzA.G. in Berlin. Later on this transmitterwas removed to Berlin, in which city theexperiments were continued in co-operationwith the " Reichspost." Esau has also usedultra -short waves for communicating be-tween the engine -driver and the guard of along goods -train.

Dr. F. Schroter (8) has described experi-ments carried out in 193o by the TelefunkenCo. Dr. Schroter explains the possibilitiesfor local broadcasting, and emphasises theparticular usefulness of ultra -short waves fortelevision, due both to the possibility ofusing a broad frequency -band, and to theabsence of fading, and, as a consequence,of selective fading.

The experiments in Amsterdam undertakenby the Philips laboratory have confirmedthat local broadcasting on an ultra -shortwavelength of 7 to 8 metres is possible withsimple receiving apparatus and with atransmitter of to I kw. aerial -power ; thatthis broadcasting is very reliable and thatit has an absolutely silent background dueto the total absence of atmospheric disturb-ances, even when conditions are such thatthese disturbances practically prevent thereception of the ordinary broadcast -stations.Also other local electrical disturbances haveless influence than on the longer waves.

Only motor -cars cause a disturbance dueto a broad spectrum of damped wavescaused by the discharge through the spark -plugs ; the wires leading to spark -plugs andinterruptor behaving as an aerial for shortwaves. These motor -car disturbances arecommonly discernible up to a distance ofabout 20 metres. In the inner -town theyare of course more frequent than in the outersuburbs, but this fact is counteracted by thecircumstance that in the inner -town areceiver need not have critical reaction, sothat in many cases these disturbances are

187 April, 1933

not heard at all. It was possible to havegood loud -speaker reception free from thesedisturbances in a room level with the streetin Damrak, notwithstanding the consider-able motor -car traffic in that part of thecity: Moreover, the different motor -cartypes behave differently in that respect.For instance, the new Ford motor -car doesnot disturb at all, due undoubtedly to theparticularly short strips from the spark -plugs to the interruptor. In connectionherewith I may make mention of the factthat disturbances of this kind may beavoided by simply inserting resistances ofro,000 ohms between the spark -plugs andspark -plug -cables and near the interruptor.In some cases one resistance near the in-terruptor suffices. We have been successfulin avoiding all disturbance from a Fordmotor -car by one resistance as describedabove, to such extent that the disturbancecould not be detected in a sensitive super-heterodyne receiving apparatus placed withits antenna at a distance of 3 metres fromthe car.

Generally speaking, it may be stated thatmotor -car disturbances have more effect inthe outer suburbs of the town where thereceiver has to be more sensitive, and where,of course, the ratio of disturbance to signalis bigger. In that case a passing motor -carcauses a disturbance of a low pitch for ashort time ; it is much less irritable than anatmospheric disturbance ; in back rooms thedisturbances as a rule are much less severe.

The Transmitting ApparatusThe first step was to design a transmitter

of high quality, so that receiving -apparatuscould be qualitatively tested. In testingreceiving -apparatus there had to be no doubtof the quality of the transmitting apparatus.As it could be predicted that for the biggerdistances critical reaction in the receiving -sets would be necessary, and on account ofthe intention to design adaptor -apparatusfor superheterodyne -reception in combina-tion with selective broadcast -receiving -sets,one of the primary objects was to avoidfrequency -modulation, so that from thebeginning the use of a piezo-quartz-crystalwas decided upon.

It was further desirable to keep the trans-mitter as small as possible and to avoidmachines in view of the small available room

April, 1933

to be expected. This led to the design of atransmitter fed with alternating current, thenecessary direct current being taken fromrectifiers. Water-cooled valves, too, hadto be avoided for the same cause, as water-cooled valves occupy a relatively large space.For the final stage we decided on the use oftwo big air-cooled screen -grid transmittingvalves.

As our experience with quartz -crystalswith a frequency in the order of 5,000 to10,000 kc/s. was not favourable, our startingpoint was a crystal of about 2,398 kc/s.that was in our possession, so that after 4times frequency -doubling a final frequencyof 38,368 kc/s. could be obtained, equiva-lent to a wavelength of 7.85 m. The use oftourmalin-crystals was not yet known to us,otherwise we should certainly have con-templated a io,000 to 20,000 kc/s. tourmalin-crystal.

Screen -grid transmitting valves were used,as this type of valve has the advantages thatneutrodyning becomes unnecessary and thatthe excitation -power needed is less than thatfor an ordinary triode. This is because ascreen -grid valve has a very small anode -reaction and a big grid space, so that a triodewith a given amplification -factor (a bigamplification -factor is advantageous becauseof the small anode -reaction) draws moregrid -current than a tetrode at equal excita-tion.

The final stage was equipped with twoPhilips-tetrodes QB 3/50o and the foregoingstage with two QB 2/75. These tetrodeshave thoriated tungsten filaments and candeliver respectively 500 and 75 watts high -frequency energy at longer wavelengths.The inter -electrode -capacities are very small.

Cap. in c.m. QB 3/500 QB 2/75

Cag 0.01 0.022Caf 8.9 6.3Cgf 16 9.1

When we designed the transmitter wehad no smaller transmitting tetrodes, sothat the first stages are not equipped withthe more modern QC 05/15 but with TCo4/10 and TC 03/5, respectively, smalltriodes with oxide -coated filament for ioand 5 watts h.f. energy respectively whenused on longer wavelengths. The trans -


mitter has 7 stages. The first stage containsa TC 03/5 and the quartz -crystal for2398 kc/s. The second stage doubles thefrequency and contains two TC 03/5 valves.The excitation of these two triodes is suchthat no grid -current flows, so that thecrystal -stage is not affected by reaction fromlater stages due to modulation. Hence thisstage doubles the frequency and at the sametime acts as a buffer ; as a consequence, itcan only deliver a very small amount ofpower to the 3rd stage.

In this 3rd stage the frequency is doubledagain. Here also two valves TC 03/5 aremounted. The 4th stage is an amplifierand contains 2 valves TC 03/5. The 5thstage, containing two valves TC 04/10,doubles the frequency for the third time. Inthis stage the plate -tension is modulatedwith the aid of a Philips 5o watt l.f. amplifierwith a special output -transformer. Such a5o watt -amplifier, fed with a.c., is excitedby a microphone -gramophone amplifier, tothe input -terminal of which are connecteda gramophone -pick-up and three or lessmicrophones. The filament -current and themicrophone -current are drawn from a 6 -voltstorage battery, the only d.c. source that isused in the transmitter ; the plate -tensionis delivered by commercial anode -rectifyingapparatus. The stages 2 to 5 are allneutralised, this being necessary for fre-quency doubling stages. Frequency -doub-ling is performed by strong distortion of theanode -current and filtering the 2nd har-monic.

The 6th stage, containing 2 tetrodes QB2/75 as remarked above, is a linear amplifierfor the modulated energy and excites thegrids of the valves in the final stage. This7th stage, the final stage, contains twotetrodes QB 3/500 ; in this stage the poweris raised to the final level and at the sametime the frequency is doubled for the lasttime.

Each stage is mounted in a separatecopper enclosure, the measuring -instruments,filters, etc., being enclosed in separate coppercases for each stage. A copper -plate sepa-rates the h.f. part of the transmitter fromthe l.f. part underneath, containing trans-formers, rectifiers, l.f. filters, etc.

In every stage the anode -direct current,the grid -direct current, anode -tension andfilament -tension are measured, and in the


last two stages also the screen -grid -tension.A thermo-ammeter measures the 1.f. currentin the plate oscillating circuit of the 1st stage.Another thermo-ammeter is coupled to thefeeders between the transmitter and aerial.

The whole transmitter is built in an ironframe, covered with copper, copper -gauzeand perforated iron. The back consists oftwo doors ; all high tension is automaticallycut off by opening these doors. To obtainaccess to the first five stages the copper

189 April, 1933

screening plates must be removed. Theupper part of the front is of copper : allmeasuring instruments, tuning knobs,switches, etc., are mounted here.

All filaments, including the first stage, arefed with a.c. and all anode -tensions andgrid -bias -tensions are taken from rectifiers,making use of small high -vacuum -diodeswith oxide -coated filaments, and in the lasttwo stages mercury -vapour diodes also withoxide -coated filaments.

Front view of the transmitter.

April, 1933

That there is no trace of a hum, notwith-standing this general use of a.c. throughoutthe transmitter, is due in the first place tothe use of a quartz -crystal ; omitting thisa hum is heard due to frequency -modulation.

In the second place, a careful filtering ofall anode- and grid bias -rectifiers is responsiblefor this absence of hum. These excellentfilters for the first five stages are not ex-pensive as only low tensions and smallcurrents are dealt with. The filtering ofthese stages has to be superior also for thereason that the modulation -depth of a humgrows with each frequency -doubling. Thecause of this is the relatively big grid -biasrequired for frequency -doubling.

For the same reason, however, the gridof each foregoing frequency -doubling stageneeds much less than ioo per cent. modula-tion for a required ioo per cent. modulationof the final stage. The grid -oscillatorytension of the 7th stage, a frequency -doublingstage, will need, say, x per cent. modulation -depth for ioo per cent. modulation of theanode -power. As a consequence, also thegrid of the linearly amplifying 6th stageneeds only to be modulated with x per cent.The 5th stage is a frequency -doubler asmentioned above, in which the anode -tensionis modulated. As a frequency -doubler, theanode -tension of this stage needs only, say,y per cent. modulation for ioo per cent.modulation of its anode -energy, and conse-quently, of the grid -tension of the 6th stage.As this latter has to be modulated x percent., as we have seen, the anode -tension ofthe 5th stage has only to be modulated xyper cent. for zoo per cent. modulation of theaerial -power. The product xy is in theorder of 25 to 4o per cent. ; this low modula-tion -depth of the anode -tension of the 5thstage is easily obtainable linearly.

As before the 5th stage the frequency istwice doubled, it will be clear from theforegoing reasoning that more care has to beattributed to the 1.f. filters as the first stageis approached. We have modulated thetransmitter ioo per cent. by modulating thesecond stage only a few per cent., but nogood quality could be maintained due to verysmall variations in the main -supply tensionnecessitating relatively big variations of therequired settings of all succeeding stages,which is no simple matter, especially whenusing rectifiers. To avoid complications,


we decided on modulating the 5th stagewhich could be done comparatively easilyand did not suffer from important adjust-ment -difficulties when ordinary variationsof the main -supply tension are prevalent.To modulate two TC ozi/io valves 25 to4o per cent. a 5o watt amplifier is notnecessary ; a 25 watt amplifier would workas well.

The oscillating -circuit of the 7th stagelooks somewhat strange, as it appears to bebuilt up by connecting in series a self-inductance strip and a comparatively largevariable condenser. The truth is that theoscillating circuit is built up by a self-inductance strip and the inter -electrodevalve -capacities ; the variable condenser isa tuning device in the self-inductance partof the circuit ; it diminishes the inductivereactance and as a consequence has to betaken as large as possible ; on the other hand,it must be small enough to serve as a smoothtuning -device. A much higher circuit -im-pedance is obtained from such a series -connected condenser than from a parallel -condenser with its long connecting leads andbig dimensions and unavoidably large zero -capacity. One of the chief difficulties inultra -short wave design consists in obtaininghigh impedance of the oscillating circuits.It should not be forgotten that coupling ofthe aerial already diminishes the anode -circuit -impedance considerably, so that nomeans should be neglected of increasing theimpedance of the anode -circuit -impedanceitself. The efficiency of the final stageremains, however, as low as 20 to 25 percent., notwithstanding all precautions taken.

The antenna is a copper rod of half awavelength. It is fastened by insulatorsto a wooden board which is attached to awooden mast kept in position by steel guyssubdivided by insulators. Close to thecurrent-antinode a thermo-ammeter iscoupled to the antenna. This thermo-ammeter is needed when searching for thebest coupling of antenna and antenna feeders.Once this most favourable coupling is foundit is of no further use, since then a secondthermo-ammeter, coupled to the feeders inthe transmitter room, may replace it.

The feeders are galvanically coupled withthe antenna, and magnetically with theanode -circuit of the final valve -stage. Sincethe mast is located on the highest point of


the Canton Hotel (with the exception of thetower), the mid -point of the vertical dipoleis about 45 m. over the street -pavement.The transmitter itself is located in a smallroom in the upper floor of the hotel, so thatthe antenna -feeders have a length of about20 metres.

At first the reports on the quality of thetransmissions were not unanimously favour-able, notwithstanding the fact that a control -

April, 1933

receiving apparatus in the immediate vicinityof the transmitter gave excellent results andneither frequency -modulation nor fading -distortion was to be feared. The cause ofthis was the poor construction of the re-ceiving - apparatus used, the design ofwhich demanded special care as will beshown later on. As these receivers weregradually improved it was recognised moreand more that the quality of the transmitter

Back view of the transmitter with screens removed from top sections.B 2

April, 1933

was in no respect inferior to that of the bestbroadcasting -transmitters.

The Receiving ApparatusWe shall now treat the receiver -apparatus,

reserving the important link, the trans-mission, for the last chapter.

In the first place, we have to considerwhich types have to be taken into con-sideration and whether certain types havespecial advantages for ultra -short-wave re-ception. Our attention has been nearlyexclusively limited to the design of adaptor -apparatus, which, combined with theordinary broadcast -receivers, should con-stitute complete ultra -short-wave receivers.

The most simple adaptor has a reactivedetector -valve, the grid -circuit being tunedto the ultra -high frequency, and the outputterminals of which have to be connected tothe gramophone -terminals of the broadcast -receiver, so that only the low -frequency -amplifying part of that apparatus is madeuse of. This simple and cheap adaptor isquite satisfactory up to a distance from thetransmitter of i to 2 km. when using a smallaerial of about 3 m. length in a room at thestreet -level.

A more sensitive device is a superhetero-dyne -adaptor which assures a good receptionup to a distance of 4 to 5 km., other circum-stances being equal. It is even possible toget very strong reception at much greaterdistances, but the Schrot-effect of theamplifying valves in that case becomes sopronounced that the ratio of signal toparasitic noise is too small.

With super -regenerative reception it isthe Schrot-effect again that limits the rangeto the same order as with superheterodyne -reception. A great advantage, however, isthe simplicity in tuning, the latter being notcritical, and the absence of so-called capacity -effect. This capacity -effect is one of theserious difficulties in all ultra -short-wavereception.

Loose antenna -coupling, screening, andlong insulated axes in tuning and reaction -devices are indispensable to freedom fromthis effect. The aerial -coupling has to be soloose that it has practically no effect on thenatural frequency and the damping of theoscillation -circuits in the apparatus. Thisis necessary as usually fairly critical reactionhas to be used and an effect of the nearby


human body on the natural frequency of theaerial cannot be avoided. With too tightlycoupled aerials the human body has a markedeffect on the reaction and the tuning, render-ing it impossible to get satisfactory reception,unless the electro-magnetic field is so strong(distance z to 2 km.) that critical reactionhas not to be applied.

The screening must be good in order thatthe hand, when tuning, has no effect on thenatural frequency and damping of theoscillating circuit in the apparatus ; thesame applies to the long tuning -rods. Noweffective screening is not so simple as isusually the case when dealing with lowerfrequencies, due to the absence of a well-defined earthing point of the screen itself.The earth -lead, indeed, has a length of somemetres, that is in the order of one or morequarters of a wavelength. A good earth isonly to be had when the screen is a potentialnode of the aerial formed also by the earth -leads. Even in that case the earth -point isnot defined as the different points on thescreen already have not the same tension.

Since the antenna -coupling must notexceed a certain relatively low value, and,since the reaction depends on this coupling,this, together with the desired range, mustbe taken into account when deciding on thepower of the transmitter.

In superheterodyne -reception capacity -effect is still more troublesome, since thenatural frequency of the generator -valve -circuit is directly influenced by the approachof the human body, so that the intermediatefrequency is also dependent on it. It is, ofcourse, possible, after roughly tuning theadaptor -apparatus, to adjust the broadcastreceiver ultimately when capacity -effect istroublesome, but this procedure is not easyunless the broadcast -receiver has one -knobtuning. Furthermore, care has to be takenthat no ultra -high frequency energy pene-trates in the broadcast -receiver, since in thatcase capacity -effect, when tuning the broad-cast -receiver, is still present to some degree.There is still the danger that, after adjustingthe broadcast -receiver, this appears acci-dentally to be tuned to a strong broadcast -transmitter in the ordinary broadcast -wave range, so that this transmitter interfereswith the ultra -short wave -transmitter unlessa very effective screening is provided of theadaptor -apparatus as well as of the leads


connecting this with the aerial and earth -terminals of the broadcast -receiver. Thenit is still possible to be troubled by an inter-ference -note from a broadcast -transmitterwith the transposed frequency. For thesereasons it is better to construct the adaptorin such a way that it may not be necessary toadjust the broadcast -receiver.

As soon as the superheterodyne -adaptoris provided with a separate loose -coupledgenerator -valve all difficulties disappearwhen the other precautions mentioned aboveare not neglected.

In that case the aerial may be tightlycoupled to the tuning circuit in the adaptor -apparatus, as this circuit, when very looselycoupled to the generator -valve, does notaffect the natural frequency of the generator -circuit in an appreciable degree. The de-tection, too, is improved, since it appearsdifficult to get fair detection from an oscilla-ting valve, especially when the oscillation -frequency is so high, as in this case theamplitude of the oscillations is not so easilycontrolled. It is, of course, necessary todecouple common leads to generator- anddetector -valve (filament- and anode -leads)effectively for the ultra -high frequency bymeans of small filter -coils and air condensers.

In conclusion, I must call attention to thedisagreeable property of the superhetero-dyne -adaptor that during the first quarterof an hour or so the frequency drifts, duechiefly to the gradual heating up of thegenerator valve, and, in much less degree, bythe heating up of the circuit -components.We proved this by placing the valve outsidethe apparatus and cooling this valve, or, asan alternative, the circuit, by liquid airwhich immediately affected the frequency.After a quarter of an hour or so the frequencybecomes constant. When using indirectly -heated cathodes the frequency -drift is muchbigger than when using directly -heatedfilaments, since the temperature of valveswith indirectly -heated cathodes becomesmuch higher: The above -mentioned valueof a quarter of an hour is valid for indirectly -heated cathodes. Within that time thefrequency may drift as much as 20o kc/s.,that is about -I- per cent. Still it remainsindispensable to use valves with indirectly -heated filaments in an adaptor fed from themains ; otherwise a powerful hum due tofrequency -modulation cannot be avoided,

April, '933

so that it seems impossible to avoid thisquarter of an hour.

For a superheterodyne -adaptor it is stillmore important than for a reaction -apparatusthat variable condensers have to meetspecial requirements ; the spindles must beabsolutely free from torsion and tolerance ;the condenser must be very finely variableand must be so designed that there is nofriction of metal on metal when adjustingthe moving plates, so that the applicationof a condenser with two fixed plates and onevariable plate moving in the vicinity of thefixed plates seems almost unavoidable.The other assembly -parts, too, must be ofrobust and rigid construction. With reac-tion -apparatus special attention must bepaid to designs having the least dielectricand ohmic losses.

From the foregoing it may be seen thatthe reception of ultra -short waves in theorder of 7 to 8 metres calls for special pre-cautions, which are the more necessary asthe field strength diminishes. This neednot, however, lead to expensive apparatus ;special attention must be paid to certaindifficulties which do not arise in the receptionof ordinary broadcast -waves, even down to15 metres or so, and which are due to non -stationary potential distribution over theapparatu§ and to the small reactances ofsmall stray -capacities.

The TransmissionAs was remarked above, ultra -short waves

such as are dealt with in this paper,are not reflected by the Kennelly -Heavisidelayer. Observations have shown, however,that this does not apply to waves of morethan 8 metres length. It appears, however,that waves of 5 to 6 metres are sometimesreflected as a freak -phenomenon, as may beseen from the description of the R.C.A.experiments. These reflections, however,rarely occurred and were always of veryshort duration (one minute or so). Reflec-tions were observed of higher harmonics fromRocky Point at a distance of 800 miles ;the receiver was tuned to 5-6 metres. Onshorter distances these reflections were neverobserved.

Practically speaking, however, one maysay that nothing comes back from theHeaviside layer when the wavelength is lowerthan 8 metres, and only the direct rays need

4

April, 1933

be taken into account, so that the range islimited and fading cannot occur. Thisrange remains relatively restricted due to :

(I) Considerable absorption ;(2) the fact that such rays cannot follow

the curvature of the earth -surface to anygreat extent, so that there is a horizonbeyond which they cannot be received, how-ever great the transmitting power.

This horizon, of course, depends on theheight of transmitting and receiving aerial.The French experiments between the Rivieraand Corsica have taught, moreover, thatthis horizon is subject to variations, depen-dent on conditions of the lower atmosphere.

Assuming the rays not to follow at all thecurvature of the earth -surface, the signal -range is (sufficient transmitting power andreceiver -sensitivity being assumed) :

D = 3.56 (Olt Vrir)wherein

D = signal -range in km.ht.-- height of transmitting aerial in m.14.= height of receiving aerial in m.

Assuming the height of the mid -point of thedipole of the Philips -transmitter on theCarlton Hotel to be 45 m. above street -level,and assuming a receiving -aerial at a heightof io metres, the signal -range would cometo 35 km. (absorption disregarded), so thatUtrecht and Leiden would lie at the limitof signal -range.

Only from Haarlem and Heemstede (at18 kin. distance) were reports received fromlisteners not living in Amsterdam.

As for the absorption in the city, this issoon perceived when driving throughoutthe city in an autocar (as we did) fitted witha receiving -apparatus. The diminishing offield strength is very evident on passingfrom a broad street into a narrow by -street.We have carried out several measurements,some results of which we shall now discuss.

Position.Volts Output

of the Receiver.

HeerengrachtKalfje, street -level .

id., ist floor ..id., ist floor at the end of a

long room ..id., outside the Kalfje, close

to the Amstel..

37.52.54

2

12 to 14


Particular attention is called to a measure-ment taken in the Philips showrooms(Heerengracht) at a distance of about qoometres from the transmitter (in the CarltonHotel) in a locality level with the street, andto a measurement in and near the restaurant" Het Kalfje " at the Amstel, at a distanceof 6 km. from the transmitter.

The considerable increase of the field -strength when moving the receiving -appa-ratus from the interior of the restaurant tothe side of the river Amstel, separated fromthe restaurant only by a road, is very re-markable.

Schroter made a rough estimate of theabsorption in Berlin by comparing what maybe expected when absorption is not takeninto account and what actually is measured.He arrives at an absorption of r neper foreach 500 metres that the waves travelthrough the houses, which corresponds to areduction of the field -strength to i/ro aftertravelling through I km. of buildings. Thisabsorption can be diminished by choosing avery high position for the transmittingaerial ; a high receiving aerial of course addsto the same improvement. The receiving -aerial must not be allowed to come too nearthe gable of the house and the lead-ininsulator should be of good quality and ofsmall capacity. The effective length of theaerial is very important, this being adjustedin the receiving -room or in the receiver -apparatus itself by an extra length of wireor by a small aerial -loading coil, so that afavourable current- and potential -distribu-tion is arrived at, suitable for a convenientcoupling with the receiver -circuit.

In connection with the transmission ofthese short waves, I may call attention tothe following. From different sides (9, io)it is proposed to avoid difficulties with super-heterodyne -receiving -apparatus by omittingthe generator -valve in each individual appa-ratus and replacing it by one commontransmitter, having a frequency -differencefrom that of the modulated transmitter of200 to 1,500 kc/s. As an alternative, it isproposed to modulate the broadcast -trans-mitters constantly with 200 to 1,50o kc/s.which modulation in its turn is modulatedby the broadcast -programme.

Both procedures seem very attractive,but are quantitatively defective, as may beshown by the following reasoning. Our


starting -point is the reception we observedrepeatedly in " Het Kalfje " with the trans-mitter in the Carlton Hotel. We made usethere of a two -valve superheterodyne -adaptorapparatus in combination with a Philips -broadcast -receiving apparatus, type 2511 (2stages high -frequency ampl., linear de-tection, one stage low -frequency ampl. witha pentode). With this receiving assemblythe Philips -transmitter was received and thegridswing of the 1st detector (in the adaptor)was about o.i my., the heterodyning gene-rator giving about i volt gridswing on thedetector. The value of 0.1 my. is calculatedfrom the accurately known properties ofthe receiver 2511. One volt induced aux-iliary grid -voltage -swing has been taken asthis is a suitable value for the type of valveused ; increase of this value does not im-prove the receiving properties, whereasdecrease causes the output voltage of thereceiver to decrease in the same proportion.

If it is desired to produce this one voltby an auxiliary transmitter placed besidethe main Philips -transmitter, the formerwould have to produce a field -strength 104times that produced by the Philips -trans-mitter itself. The power of this auxiliarytransmitter then would have to be 108 timesthat of the Philips -transmitter. The aux-iliary transmitter would have to have apower of 25.106 kw. to produce the samereceiving -conditions in " Het Kalfje " asobtained with a local oscillator in thereceiving -apparatus itself, as used by us.Instead of a broadcast -transmitter of I kw.and an auxiliary transmitter of 25.106 kw.,one might as well design two transmittersof each 2,50o kw., for, roughly speaking,the output of the broadcast -receiver is pro-portional to the product of the voltagesinduced on the grid of the 1st detector, dueto signal and to auxiliary generator. Theproduct of the powers of broadcast -trans-mitter and auxiliary transmitter thereforehas to be kept constant. If only one trans-mitter is used, modulated at a high fre-quency which in its turn is modulated bythe broadcast -programme, as mentionedabove, then one may reckon on 5o per cent.high -frequency modulation at the most, asthe high -frequency modulating voltage is tobe modulated Too per cent. by the trans-mitted signal. In this case it is easilycalculated that the transmitter needs a power

195 April, 1933

of about io,000 kw. In the case of a broad-cast -transmitter with an auxiliary trans-mitter of the same power, we have

a sin cot + a sin (w + itico)t,and in the case of one low- and high -fre-quency modulated transmitter, we haveb sin cot(r 0.5 cos ticut) = b sin cot

o.25 b sin ((t) zlco)t

so that a2 = 0.25 b2 for equal receiving -conditions.

a2 and 1,2 determine the power of thetransmitters considered.

One may raise the question how muchthese values change with the distance.From the measurements in our possessionit is easily calculated that for i and 2 km.distance of the receiver from the trans-mitter, the 2 equal transmitters have eachto be designed for II kw. and 115 kw.respectively. These calculations are basedon measurements on the spot ; they do notclaim to be rigorously exact, but they arecertainly of the right order. They showwhy the idea of a so-called superheterodynetransmitter was ruled out by us from thebeginning. The required power, necessaryfor this procedure, could not be realised,quite apart from the economic side of thematter. Moreover, with the calculatedtransmitting powers we get receiving -con-ditions equivalent to those we obtained in" Het Kalfje " using the most sensitiveapparatus we could attain to, and with alocal oscillator ; this reception was somewhattoo weak for general use, since sufficientloud -speaker strength was accompanied bya prohibitive Schrot-effect. All powers,therefore, would have still to be multipliedby a factor of about 5 for real good receiving -conditions.

ConclusionIn conclusion, we may state that with an

ultra -short-wave transmitter of about kw.aerial -telephony -power, transmitting on awavelength of 7 to 8 metres, reliable localbroadcasting of high quality is possible in abig city, provided a suitable receivingadaptor apparatus in combination with asensitive broadcast receiving apparatus anda good aerial is used. The reception is charac-terised by a particularly silent background,even at times when ordinary broadcast -reception is totally spoiled by atmosphericdisturbances. The frequency -range at our

April, 1933

disposal is very broad ; for equal frequency -difference of transmitting stations, as manystations as can be located within the wave -range of 200 to 2,000 m. may find a placewithin a wave -range of 7.5 to 7.8 m. ; by thismeans, all important European broadcast -stations could be located within thiswave -range. However, this wave -range isnot needed since the signal -range of ultra -short -wave transmitters is limited, so thateventually all bigger cities might use thesame wavelength without the possibility ofinterference. Principally these waves areparticularly adapted for television, thatmeans for development of the home -cinema,in which case synchronisation is easilyobtained by making use of the local elec-tricity -mains. Installations as describedhere are of much importance for communica-tion between islands or between islands andthe continent, where cables are too expensive,e.g., for the communication between Javaand Bali, two islands separated by a verydeep sea. Furthermore, for communicationover shorter distances, when this communica-tion cannot readily be established in anotherway, ordinary radio being impossible becauseof interference with other services (police,fireguard, army and navy, long trains, ferry-boats in foggy weather, etc.).

New


1.

REFERENCESDunmore and Engel : " Scientific Papers." Bur. Stand.,No. 469, t. 10, April, 1923.

2. N. Wells : " The Marconi Wireless -Beam Reflector of Inch-keith." Engineering. No. 119, 309, 1925 I.

3. G. A. Beauvais : " Resultats experimentaux de Telecom-munications avec les ondes tres courtes." Onde Elects.,1930, pag. 484.

R. Mesny : " Les ondes courtes." Onde Electr., 3 janv., 1924.R. Mesny : " Les ondes ilectriques courtes." Ed. 1927.

4. R. Jouaust : " Some details relating to the propagation ofvery short waves." Proc. I.R.E., 19, 3, 1931.

G. A. Beauvais : " Experiences de communication radio-telephonique sur ondes tres courtes entre la Corse et leContinent." Annales des P.T.T., 1930, pag. 293.

5. Beverage, Peterson and Hansel : " Application of frequenciesabove 30,000 kilocycles to communication problems."Proc. I.R.E., 19, 8, Aug., 1931.

F. Gerth and W. Scheppmann : " Untersuchungen iiber dieAusbreitungs vorgange ultrakurzer Wellen." jahrb. derdrahtl. T.u.T., 33, 1, Jan., 29.

6. F. Gerth : " Der derzeitige Stand der Entwickelung derultrakurzen Wellen unter Beriicloichtigung ihrer Ver-wendungsmaglichkeiten fiir Rundfunkzwecke." E.N.T.,8, 1, 1931.

7. A. Esau : " Die Entwicklung der Ultrakurzwellen fur denRundfunk." Funkbastler, 11 Sept., 1931.

8. F. Schroter: " Erwiderung auf die vorstehenden Ausfiihrun-gen." Funkbastkr, 11 Sept., 1931.

F. Schroter: " Ultrakurzwellen-Rundfunk." Telefunken-Zeitung, 57, April, 1931.

F. SchrOter : " Aussichten des Ultrakurz-Rundfunks."Die Norag., 18, Mai, 1931.

F. Schroter : " Ultrakurzwellen-Femsehen." Telefunken-Zeitung, 58, Juli, 1931.

9. Manfred von Ardenne : " Zur Technik des Sendens andEmpfangens von Ultrakurzwellen, die mit mehreren modu-lierten Hochfrequenzen moduliert sind." Jahrb. derdraht. 1 T.u.T., 37, 1, Januar, 1931.

10. F. Schroter: " Ultrakurzwellen-Rundfunk." Telefunken-Zeitung, 57, April, 1931.

Short-wave Radio ReceptionBy W. Oliver.A short book of elementary instruction and

advice to amateurs wishing to construct and operateshort-wave receivers, giving a brief descriptionof typical sets and hints on tuning, learning theMorse code, tracing and repairing faults, etc.,and including lists of the distinctive InternationalCall letters, Amateur Nationality Prefixes and theAbbreviations used in Morse Transmission. Pp. rotwith 14 diagrams. Published by W. Foulshamand Co., London. Price is.

Wireless Really ExplainedBy P. J. Risdon. (Revised by W. Oliver.)Explaining in simple terms the principles of

broadcast transmission and reception, includingthe elements of electricity and magnetism, wirelesswaves, valves, inter -valve coupling and the actionof typical receivers. Pp. 94 with 26 illustrationsand diagrams. Published by W. Foulsham andCo., Ltd., London. Price is.

Wireless Wrongs and How to Right ThemBy " A Man in the Street."A brief summary of faults prevalent in receivers,

Bookstheir more common causes, tests and possiblemeans of remedy. Arranged in a convenientlytabulated form and followed, in the concludingpages of the booklet, with a few hints and tipsregarding aerials, earths, batteries, valves andloud speakers. Pp. 25. Published by J. W.Arrowsmith, Ltd., Bristol. Price 6d.

Physics in MeteorologyBy G. C. Simpson, C.B., C.B.E., D.Sc., LL.D.,

F.Inst. P., F.R.S., Director of the MeteorologicalOffice, London.

A reprint of a lecture given before the Instituteof Physics in November, 1932, showing the applica-tion of sound, light, heat, magnetism and elec-tricity to the solution of problems of meteorology.Pp. 22 with 5 diagrams. Issued by the Instituteof Physics.

Rundfunk Jahrbuch, 1933Giving a general résumé of the progress of broad-

casting in Germany, and the principal events ofinterest during the past year. Pp. 167, with manyillustrations. Published by the Reichs-Rundfunk,Gesellschaft, Berlin. Price Rm. 1.50.


A Magnetron

By E.

197 April, 1933

Oscillator for Ultra -shortWavelengths*C. S. Megaw, B.Sc., D.I.C.

(Research Laboratories of The General Electric Co., Ltd., Wembley, England)

SUMMARY.-The paper describes the characteristics and performance of a magnetron oscil-lator of improved design. The oscillator is primarily intended for operation on wavelengthsbetween r metre and io metres, but it can be used on both shorter and longer wavelengths.At wavelengths down to about 3 metres an output of 40 --so watts is obtainable with a con-version efficiency of the order of 5o per cent. At shorter wavelengths the output falls off,reaching about io watts at r metre. In the range mentioned, the performance comparesfavourably with that of the normal triode oscillator, particularly at the shorter wavelengths.

IntroductionTHE limitations of the conventional

triode oscillator at very short wave-lengths have led to an investigation

of the practical possibilities of other circuits.Of these the magnetron circuit has beenfound particularly satisfactory and anaccount is given here of a medium poweroscillator of this type for wavelengths of afew metres.

There are several possible ways in whichmagnetrons, that is valves operating in amagnetic field, can be used to produce shortwave oscillations. An account of thesedifferent methods will be given in anotherpaper. t It is only necessaryto state here that the opera-tion of the oscillator to bedescribed on wavelengths ofabout i metre and longerdepends on a " dynatron "(negative resistance) effectarising from the motion, ofelectrons in combined electricand magnetic fields. Theexistence of such an effectwas first demonstrated theo-retically and experimentallyby Habannl and a practicaloscillator employing it hasbeen briefly described recentlyby McArthur and Spitzer.' Ithas also been shown3 thatunder certain conditions electronic oscilla-tions (analogous to the triode oscillations ofBarkhausen and Kurz) can be produced mmagnetron circuits and by this means wave-lengths down to a few centimetres have beenreached. Such oscillations can be producedby the circuit to be described, but the output

* Received by the Editor, August, 1932.f To appear shortly in Journ. I.E.E.

is small compared with the power obtainablefrom the longer wavelength dynatron oscilla-tions. Only the latter are dealt with here.

The Valve and its CharacteristicsThe valve used consists of a cylindrical

diode system with the anode divided intotwo equal segments and operates with amagnetic field in the direction of the elec-trode axis. The oscillatory circuit is con-nected between the anode segments. Thisarrangement has been called a " split anode "magnetron. Fig. i shows the valve used.The anode and filament leads are broughtout at opposite ends of the bulb and the

Fig. I. -5o watt split -anode magnetron.

electrode system is arranged so as to intro-duce the least possible capacity, inductanceand resistance into the oscillatory circuit.The bulb is made of low -loss hard glass.

The pure tungsten filament takes a currentof 2.5 A. at 4.o v. for about 8o mA. emissioncurrent.

Without the magnetic field the anodevoltage -current characteristics are those of

April, 1933

an ordinary diode. The valve ordinarilyoperates under conditions which give emissionsaturation when the magnetic fieldis removed.

If the relation between total anode currentand magnetic field strength is plotted withconstant (common) anode voltage and con -

soMIIIIM1111111111111111111NEEM1111MIN1111NIIMIE11111111111111=11111111MINIIM 1111111111M11111

, =MENNEN111111EMIll11111M1111111111MEMENIENMINIMMIIMEINEMI111111111111111111111111M1M1111111111MMIIMMIIIIMIIIIMMIIIIIIIMM111111111111111

01111MENIMMIIIIIIIIMIEnE11!02 04 06

FIELD CURRENT IN AMPERES

Fig. 2.-Cut-off characteristic for Ea = 1,000 v.,= 5o mA., and da = 1.o cm.

stant filament emission a curve of the typeshown in Fig. 2 is obtained. This is the" cut-off " characteristic of Hull.4 Thecritical field strength is independent ofemission and is given approximately by

He /(i80 Ea)Ida (C.G.S. units)where Ea = anode -cathode p.d. (volts)

da = anode diameter (cm.)To produce dynatron oscillations the

field strength must be greater than thecritical value He. Under these conditionsnone, or, in practice, only a small proportionof the electrons leaving the cathode reachthe anode so long as the two anode segmentsare maintained at the same potential Ea.The emitted electrons form a rotatingcylinder of space charge whose diameteris smaller than that of the anode. If ane.m.f. is applied between the anode seg-ments so that the potential of one segmentis raised to Ea + AEa and that of the othersegment depressed to Ea - AEa the rotarymotion is disturbed and electrons startreaching the anode. For a certain range ofvalues of AEa more electrons reach the lowpotential segment than the high potentialone. If the electron current reaching thehigh potential anode segment is I, and thatreaching the low potential segment is I,there will he a resultant power output to


the circuit connecting the anode segmentsof

6,Ea . - AE, .= AEa(.12- .11)= e .

where e is the inter -segment voltage (2 AEa)and i is therefore equal to (I, -That is to say, the net output to the circuitis the same as that which would be producedby a current (I, -- ID flowing in opposi-tion to the inter -segment voltage e. If thecircuit connecting the anode segments hassuitable constants oscillations will be main-tained in it in a manner somewhat similarto the operation of a triode push-pulloscillator.

The relation between e and i is the operat-ing characteristic of the valve. This rela-tion can be obtained from static measure-ments of AEa, I1 and /2 and is of the formshown in Fig. 3 (full line). With increasing

.e (positive or negative) i increases in thenegative direction up to a point and thendecreases, passes through zero and for largevalues of a (of the order of twice Ea) becomespositive. In Fig. 3 negative i is to be inter-preted as current flowing in opposition to theapplied e.m.f.

I-

0.0.I-

0

INTER -SEGMENT VOLTAGE

Fig. 3.-" Dynatron " operating characteristicsshowing variation of effective output currenti =(I$ -I and input current I = (4+12)

with inter -segment voltage e.

The second curve (dotted line) in Fig. 3shows the form of the relation between e and/a = (I, + I,), the current taken from theH.T. supply. From the two curves theoutput from the valve to the circuit and theinput to the valve from the H.T. supply canbe calculated, respectively, for various


operating conditions. Such calculations haveshown very close agreement with measure-ments made at frequencies sufficiently low

-r

-H.T.+

FIELD MAGNET

TOLOAD

Fig. 4.-Magnetron oscillator circuit.

for circuit impedances and radio frequencyvoltages to be determined with accuracy.

When the operating frequency is so highas to approach the rotation frequency of theelectrons in the magnetron the static char-acteristic is no longer followed and thedynatron effect begins to disappear. How-ever, this limit occurs at a considerablyhigher frequency than the correspondingeffect in the triode reaction oscillator. Theeffect of the valve constants on the oscilla-tory circuit impedance is also less in themagnetron than inthe triode case. Theeffective inter -anodecapacity of the mag-netron described is onlyabout 2.0 p,p,F whenmounted between themagnet poles.

The CircuitFig. 4 shows the

circuit diagram of theoscillator. The wave-length is determinedby the constants ofthe circuit between thetwo anode segments.The H.T. is suppliedvia the electrical centreof this circuit. NoH.T. choke is required.For wavelengths above2 or 3 metres it isquite practicable touse " lumped " in-ductance and capacity

April, 1933

in the tuned circuit owing to the lowcapacity of the valve itself. For shorterwavelengths a Lecher wire tuning system ispreferable.

Apart from wavelength adjustment thebehaviour of the circuit can be entirelycontrolled by means of the filament and fieldcurrent rheostats. In the sense that itcontrols the oscillation strength and theefficiency of the oscillator the field control isanalogous to the reaction control in anordinary triode oscillator. The fact thatthis control is outside the radio frequencypart of the circuit is an appreciable advantageat very short wavelengths.

The maximum output depends on thefilament emission. As the space charge satu-ration current is much greater than the maxi-mum permissible anode current the outputincreases continuously with filament emissionin the working range.

The magnetic field is supplied by a fieldmagnet described in the next section. Thevalve is mounted between the poles of thismagnet with the electrode axis in the direc-tion of the magnetic lines of force.

Fig. 5 shows a photograph of a completeoscillator in which the various parts of thecircuit will be easily identified.

Fig. 5. --- Magnetron oscillator.

April, 1933

The Field MagnetThe magnetron described has been de-

signed to operate between the poles of aniron cored field magnet instead of inside asolenoid as was the case in previous shortwave magnetron oscillators. Two importantadvantages result from this. The powerrequired to produce the necessary fieldstrength is greatly reduced and the air -blastcooling required by a comparable solenoidsystem can be dis-pensed with. Lessimportant advantagesare a reduced capacitybetween the anodesegments and the fieldmagnet and the factthat the electrodes areeasily visible duringoperation.

A typical field mag-net is shown in Fig 6The dimensions of theiron circuit, which isconstructed of mildsteel, are shown in thefigure. The coils areeach wound with 2,000turns of No. 22 s.w.g.,d.c.c. wire. With a current of z amperethis magnet gives a field strength at thecentre of the gap approaching r,000 c.g.s.units. This involves a power of 35 to 4owatts. Saturation occurs at about 1,70oc.g.s. units. With a 4o mm. airgap the fieldis sensibly uniform in the space occupied bythe electrodes.

OperationThe following figures give typical opera-

ting data for a magnetron of the typedescribed.Normal anode voltageMaximum anode voltageMaximum anode currentMaximum anode lossMaximum magnetic field

strength required

= 500 to 1,000= 1,200.= 8o mA.= 5o watts.

= about i,000c.g.s. units.

Maximum power output :5o watts at 10 m. wavelength.4o watts at 3 m.lo watts at I m.

Optimum load impedance at I,000 anode


volts (estimated from experimental resultsat 3 m. wavelength) = 3,000-5,000 ohms.

Example of operating conditions :-Ea = 1,000 H = 90o c.g.s. units./a = 8o mA. A = 3.o m.Output = 4o w. Conversion efficiency

= 5o per cent.Total efficiency, including field and fila-

ment power = 3o per cent.

Fig. 6.-Field magnet.

The power output figures quoted abovehave been obtained by means of a load lampconnected to the oscillatory circuit. Byadjusting an identical lamp, connected tosupply mains, to the same brilliancy as theload lamp a fairly accurate estimate of thepower output can be made.

The relation between oscillation wave-length and the length of a Lecher wireoscillatory circuit (wire diameter = 0.5 cm.,spacing = 2.5 cm.) is shown in Fig. 7 forwavelengths up to 4 metres. The circuitlength is measured from the anode centreto the Lecher wire bridge.

The filament emission and anode voltagehave no critical optimum values and thepossible output increases with both thesequantities provided the field strength isincreased with the anode voltage. Thepractical limits are set by filament evapora-tion and anode temperature.

Below about i metre the energy of thedynatron oscillation falls off rapidly. Pro-vided the operating wavelength is not toonear the limit the greatest output is ob-tained with a field strength of the order of


twice the cut-off value. The optimum fieldstrength decreases as the wavelength limitis approached.

If the operating wavelength is near theshort wave limit, the efficiency decreasesappreciably with decreasing anode voltage(and magnetic field strength) since the limitdepends on the rotation frequency of theelectrons in the valve. Experimental re-sults indicate that the efficiency falls rapidlyat a frequency which is about 4 times theelectron rotation frequency for the anodevoltage used and the corresponding cut-offfield strength. The electron frequency isgiven approximately by

3.3.1071 / E,fo = da

The relation between output and anodevoltage with fixed magnetic field strength isimportant since the output is most easilymodulated by anode voltage variation. Theform of this relation is shown in Fig. 8,which also shows the relation between meananode current and anode voltage (dottedcurve). The output reaches a maximum atan anode voltage of about 0.7 of the cut-offvoltage. On the high voltage side of themaximum the oscillation amplitude decreasescontinuously reaching zero at approximatelythe cut-off value of anode voltage. On the

60

40

20

01 2 3 4

WAVELENGTH IN METRES

Fig. 7.-Relation between oscillation wave-length and length of Lecher wire oscillatory

circuit.

low voltage side, however, the oscillationscease suddenly at a point which depends onthe operating conditions and in particularon the circuit loading. As the anode currentincreases continuously with anode voltageup to the cut-off value this means that for

linear modulation there is a choice betweenlarge percentage modulation with low effi-ciency, and restricted modulation range withmuch higher efficiency, according to whether

I -zLilCcCC

0

0z0

0z

t+ EcANODE VOLTAGE (EA)

Fig. 8.-Variation of (i) oscillatory current and(2) mean anode current with anode voltage

(magnetic field strength constant).

the mean anode voltage is above or belowthe value for maximum output. By loadingthe circuit heavily it is possible to avoid thediscontinuity but only at the cost of reducedmaximum output and efficiency.

ConclusionDevelopments in radio communication

during the past year or two have indicatedan increasing need for a reliable mediumpower generator of continuous waves of afew metres length. Short distance direc-tional communication, television, and radiobeacons come under this heading. Inseveral departments of physical and en-gineering research, such as the study ofdielectrics and of wave propagation, similarrequirements have arisen. It is hopedthat the magnetron oscillator describedhere will suggest a means of meetingsome of these requirements more satisfac-torily than is possible with the conventionaltriode oscillator. The chief advantages ofthe former are (I) greater output and effi-ciency at wavelengths under about 5 metres,(2) lower effective inter -electrode capacity,

0

zcccc

-cc

-J0in0 tEc

April, 1933

and (3) greater simplicity of adjustment andgreater reliability of operation at high load-ing. The chief disadvantage of the mag-netron oscillator is the necessity of anauxiliary supply for the field magnet. Thisdisadvantage has been reduced to such anextent in the present design that a totalefficiency of about 3o per cent. including thefield and filament power is possible at 3metres wavelength (which compares favour-ably with the anode conversion efficiencyof most triode oscillators at this wavelength).It is also quite practicable to operate amobile transmitter, with an output of theorder of 20 watts at a few metres wavelength,from a 12 volt car battery.

The problem of high percentage Modula-tion with good efficiency in the magnetronoscillator is at present under investigationand is not likely to prove insoluble. Mag-netron designs for considerably larger outputare also being investigated.


While the chief field of usefulness for themagnetron is undoubtedly for very shortwavelengths, it is likely to find applicationsat longer wavelengths in circuits where avery low internal capacity is of importance.Also as a symmetrical negative resistance,available at any frequency at which measure-ments are possible, it may prove a valuableaddition to the radio laboratory.

In conclusion, I have much pleasure inthanking Mr. B. S. Gossling for valuableadvice and suggestions during the develop-ment of the oscillator described in thispaper.

REFERENCES1 Habann, Jahrb. d. drahtl. Tel. 24, 115 and

135 (1924).2 McArthur and Spitzer, Proc. I.R.E., 19, 1971

(1931).3 Zacek, Jahrb. d. drahtl. Tel. 32, 172 (1929),

(summary of paper published in 1924). Okabe,Proc. I.R.E., 17, 652 (1929).

Hull, Phys. Rev., 18, 13 (1921).

CorrespondenceLetters of technical interest are always welcome. In publishing such communications theEditors do not necessarily endorse any technical or general statements which they may contain.

The Principles of Electro-MagnetismTo the Editor, The Wireless Engineer.

SIR,-Your February editorial, " The Principlesof Electro-Magnetism," calls attention to the con-fusion of thought which often arises between Band H (not to mention Bo). As you point out" both views are correct," so long as one does notmix them up, but I would suggest that the " Heavi-side view," as you term it, is made more confusingthan it need be because we are in the habit ofdefining H in one way, from " unit pole " con-siderations, and then treating it as if we haddefined it electrically [4/r/T/rol].

The difficulty can only be illustrated by startingat the beginning ; if my argument seems to lackrigidity, this must be put down to the need of cuttingit as short as possible.

We begin by defining that " purely mathematicalabstraction," a " unit pole," from the equationmimoir2 = f, the force in dynes between poles ; wedo not consider the values which might arise in amagnetic medium, and do not therefore need tointroduce ts. Having thus provided a " unit pole,"we attach it to a spring -balance (of brass or non-magnetic material), which is graduated in dynes,and send an observer to walk about with thisinstrument in the neighbourhood of " permanentmagnets " and wires carrying " electric currents "the observer finds that he gets a force on his " unitpole," measured by his spring -balance, in these

cases, and we will call this force F or " Field -strength " [it is, of course, usually called H, butas will be seen later, it is at present desirable tocall it by some other name, avoiding the use ofH, and F for " field " seems suitable].

Now let us build a large anchor -ring, big enoughfor our observer to get inside ; to fix ideas, let ussuppose a fireclay tube, 2 metres in diameter and akilometre long, the ends being bent round to meet.Our observer in the inside of the ring measures thecross-sectional area, which he calls A cm2, and thedistance he has to walk round to return to hisstarting -point (we have suggested a kilometre),which he calls / cm. While he is doing this, we windround the ring a winding of T turns, which weconnect in series with a reversing switch, carbonrheostat, and a " blank " moving -coil ammeter,to a battery. (The ammeter is of necessity blank,since we have not yet defined a unit of current.)On closing the switch and screwing up the rheostat,the observer in the inside of the ring, who has withhim his " unit pole," reports from time to timethe force on it, and we note that equal incrementsof force correspond to equal increments of angularmovement of the ammeter needle. When theobserver finds the force F on his unit pole is equalto 42rrito/ we mark the scale of our ammeter " i,"when F has twice this value, we mark the scale" 2," etc., and we have thereby defined the Ampere,ten of which form the absolute CGS ElectromagneticUnit of Current ; and we use for current the


symbol I. Having done this, we can calculatethe force F the observer will find is acting on hisunit pole, by the equation

F = 4n/T/1o/ (a)

from the reading of the ammeter which we havenow calibrated. To save circumlocution, we mustnow mention something not yet defined, namelythe Volt, using the symbol E for " number ofvolts," and we want an instrument which will readthe time -integral of volts, f Edt. Being practicalelectricians, we can imagine something that woulddo this, such as, for example, a commercial ampere -hour meter in series with a large constant resistance(only it would have to be vastly more sensitive),or we have, in fact, the very instrument readymade and called a " fluxmeter " (only pleaseremember we have not yet heard of " flux," stillless defined it).

We now wind a " secondary coil " of T2 turnsround our anchor -ring, and join the ends of it toour meter measuring fEdt, or as we may call it,our " Volt -second meter." We then find that anychange in the force F on our observer's unit poleis accompanied by a proportional change in thedeflection of our volt -second meter.

When the change of F, say 4F, amounts to1o8/AT we mark the scale of our new meter" r," a further change of the same amount andwe mark the scale " 2," etc. We have thusdefined the " volt -second," and thereby, since wepossess a clock, we define the volt. Note that upto now we have not had any magnetic materialpresent, no need for any a.

If our observer is not available, we can stillcalculate the force F on his unit pole from thereading of our new meter, by the equation

z1F - fEdt X 108(P)AT,

Suppose, having rescued our observer from theinterior, we fill the anchor -ring with molten castiron, and let it solidify ; after it has cooled, ourobserver propounds the conundrum, " what is nowthe force F on my unit pole, which I forgot torescue and which is now somewhere inside the castiron ? " We cannot measure it directly, and weare not inclined to agree .with the suggestion thatwe should bore a " small " hole in the iron, intro-duce another unit pole, and see ; for we shouldnot then be measuring F actually inside the iron,but only in air in a hole in the iron, which wesuspect may be quite different. The argumentthat as a " unit pole " is a " purely mathematicalabstraction " we can suppose it small enough topass between the constituent atoms of the castiron does not, to the writer at any rate, make itany easier to imagine. I even venture to suggestthat the value of F is quite unknowable. " Oh no,calculate it from your ammeter reading, usingequation [a]," says someone ; another suggestsconfirming the result by the reading of the Volt -second meter, but behold this instrument hasbuzzed right round the scale an incredible numberof times, and gives a value for F (calculated byequation [/3], a thousand or more times (actually1.4 times) as big as the result calculated by ourammeter reading and equation [a]. Which is

" right " ? I suggest that we cannot possibly tell ;quite possibly neither 1 But the two numericalvalues calculated from [a] and [if] must have names,and we christen them H and B respectively.

H is now (and for the first time) defined by theequation :

H = 47TITI1el (Y)

and similarly B is defined by :fEdt x 108

B AT,Note the sequence of the definitions, the " unit

pole " enables us to define F, the " field -strength "or " Magnetic Force " From F we define current Iand volts or volt -seconds fEdt. From I we defineH, and from fEdt we define B. µ is then a ratioof two physically different quantities (as E =stress/strain).

Now the confusion arises in the usual practice,for we next proceed to confuse F (magnetic force)with H (magnetising force) and proceed to coverup our confusion by using the symbol H to standnot only for that defined by [y], but for what Ihave hitherto called F. They are certainly differentconceptions, for F is used via I, in order to define H.

It is true that we have so chosen or " cooked "our units of measurement, that in air, F, H, and Bare all represented numerically by the samenumber, but they are really distinct and differentconcepts.

In Iron (or magnetic material) F is inconceivable,as even with the assistance of mathematical abstrac-tions we cannot conceive of a way of measuring it(this is unimportant, for F and the " unit pole "have served their turn, and as soon as we candefine I and E by their means, they can be dis-carded). It would be no conceivable use to know F,the force on a non-existent " unit pole," in anypractical case.

If anyone be found to maintain that in iron Fmust of necessity be equal to H, I will asseveratethat F must be equal to B, or to 'GBH, or to someother function* of B and H, and every argumentused in favour of one equality will be equally validfor the other.

The truth is that we are enquiring, when we askthe value of F in iron, " What happens to a purelymathematical abstraction when it is transferredto another medium," as you put it in youreditorial. I submit that the answer was given toAlice by Humpty-dumpty, " It means just what Ichoose it to mean, neither more nor less."

C. R. COSENS.

To the Editor, The Wireless Engineer.SIR, -In his letter to you of February loth,

Professor Fortescue makes two bold statements,upon each of which, with your permission, I shouldlike to ask him a question.

The first statement is that by no possible experi-ment can H and B be distinguished. Now I havealways believed that, in the case of a long air -coredsolenoid, the value of H was 4w/io times the

* Any function of B and If that reduces to B (or H) whenB H will do!

April, 1933

ampere -turns per cm ; measurable, therefore, bycounting the turns and reading an ammeter. Thevalue of B, however, cannot be derived from thesemeasurements, but it can be obtained by measuringthe area of the core and reading the throw of aballistic galvanometer attached to a search coil whenthe current is interrupted. These measurementsin turn give no idea of H. Does Professor Fortescuemean to deny the existence or reliability of suchtests ?

The second statement is that it is absurd toadopt different dimensions for H and B when eachis referred to a crevasse measurement. Now, inhis daily work, Professor Fortesque must surelyoften use the Electrostatic Units. Does he realisethat, in this system, which is just as logical as theE.M. system, the dimensions of H and B aredifferent ? Would he say that, on the E.S. system,H and B are analogous to stress and strain, but noton the E.M. system ; or would he say that thecrevasse theory is applicable to the E.M system,but that in the E.S. system it is " absurd " ?

The University, WILLIAM CRAMP.Birmingham.

Vibrations of a Coil -driven Paper ConeTo the Editor, The Wireless Engineer.

SIR,-To solve Mr. F. R. W. Strafford's problem(see p. 141, March, 1933, issue) we have to prove theexistence of a force on the coil the frequency of whichis half that applied to the power valve. Assumingthe beat oscillator to be above suspicion, and thatno negative resistance effects involving relaxationoscillations are present in the valve circuits (W.E.,p. 134, March, 1932), the precedure is given below.Neglecting radiation and losses the force equationis Accelerational + Axial Constraint = DrivingForce

Or I22 Dt2 se = F cos wt .. (I)

where m = effective mass, s = axial constraint= axial displacement from equilibrium posi-

tion. From the given data one concludes that thespeaker magnet is small. The axial distribution,assumed symmetrical, will probably resemble aninverted parabola. The flux density at any pointdistant 6 from the central plane of the gap isB. = B B. - 02 where B = density at centreand is a constant depending upon the flux dis-tribution. The force on the coil per unit current

leis C = fBbd where 1 is total length of wire uni-formly spaced on coil, and b is the axial width ofthe gap. The limits of integration when the coilcentre is distant e from the central plane are- (b/2 - e) and (b/2 ). Using the precedingvalue of B9, performing the integration and insertinglimits, we obtain C = Cl - OP, where C1 = Bl- tklb2112. The driving force F = CI, so (r)becomes

m Utz + se = (C, - 01e2)I cos cot .. (2)


Putting wt = 2Z we get bt = 22fw

which on substitution in (z) yields2E

(a62 - b) cos 2Z + ce = o

and bt2co'

(3)

b=421C1-1 and c= =4 - wo(7:)2w2mwm am

corresponding to the fundamental frequency of thediaphragm on the axial constraint. The solutionto (3) is on somewhat similar lines to that givenby the writer in W.E. October, 1932. It containsan infinite number of terms (alien frequencies) thefrequency of the one being half that of the funda-mental. The physical explanation is that owingto variation of the field within the magnet, the coilin moving to one side of its central position tendsto move further, i.e., the rectification effect. Itsprogress in this direction, however, is foiled by theaxial constraint, so the coil dallies at each end of itsexcursion. The net result is that a motion ofhalf frequency, together with a myriad of others, issuperposed on the fundamental. Owing to theradial mode resonances its presence is revealed,otherwise it would be of minor account at frequenciesof L000 , especially with a large diaphragmor a gap of greater width giving a more uniformfield distribution axially.

This probably diagnoses Mr. Strafford's trouble,but I should be glad if he would confirm the diag-nosis by using a long coil protruding 4 to 6 mm.from each side of the gap. The flux interlinkageswill then be substantially constant throughout thetravel of the coil, even though the driving currentdistorts the field. The constancy of interlinkageis, of course, the criterion for absence of alienfrequencies. At the same time, he might test theinfluence of varying a and b by halving the currentand of reducing c by removing fairly large piecesfrom the annular surround and centering device.

London. N. W. McLAcHrtix.

The DynatronTo the Editor, The Wireless Engineer.

SIR, -I was interested to see the paper by F. M.Colebrook on the use of the " Dynatron " arrange-ment for amplification, as I had tried the schemequalitatively a year or two ago. From the academicpoint of view one is always pleased to see theoreticalanalysis confirmed by experiment ; but the use ofaudio -frequencies for measurements obscures thepractical difficulties which would be experienced inapplying such a circuit to selective radio -frequencyamplification.

In the first place the use of a low anode voltagetends to increase cross -modulation ; admittedlyvariable -mu valves were not available at the timeof my experiments, but in the absence of definiteinformation I am inclined to doubt whether thesevalves would remain free from this trouble whenused in a dynatron circuit. The absence of adefinite reaction coupling associated with the tuned


circuit has its disadvantages ; for example, if aH.F. choke is employed at any point in the anodecircuit, oscillation may occur at its resonantfrequency before the desired degree of selectivityof the tuned circuit has been attained. Thisprecludes the use of the tuned grid type of inter -valve coupling. Then again the chief advantagesuggested is the absence of special componentsto provide reaction, but reaction is most commonlyrequired in " straight " sets where the tuned circuitswill have to cover a wide range of frequencies.

April, 1933

Whereas a reaction circuit can be designedto give fairly constant results, using a reactioncondenser ganged to the tuning if necessary (cf.Wireless World " Autotone " receiver), the dynatroncircuit requires readjustment of one of the valvepotentials at each frequency to compensate for thechange in decrement of the tuned circuit. Finally,it would be impossible to use the valve in questionfor volume control by variation of grid bias.

D. A. BELL.Magdalen College, Oxford.

The Researches of the late Dr. D. W. Dyeon the Vibrations

THE meeting of the I.E.E. Wireless Sectionon Wednesday, ist February, was devotedto a lecture on the above subject by Dr. E. H.

Rayner of the National Physical Laboratory.Referring first to the work of Langevin in the

submarine application of the quartz crystal as asupersonic source (for depth sounding), the lecturerdiscussed the work of Prof. Cady in the applicationto radio -frequency measurements and the associa-tion of the late Dr. D. W. Dye in this connection.He then showed various models of quartz crystals,including a large -size model illustrative of thehexagonal shape. From this he explained thevarious axes, optical and electrical, stating thatthe resistance across the electrical axes was of theorder of 10-5 of that across the optical. Variousatomic models were also exhibited, showing thedisposition of the atoms within the quartz structureand their effect upon the optical and electricalaxes.

The lecturer then turned to Dr. Dye's own work,referring particularly to the need for knowledgeof the mechanical processes involved in the vibra-tion of quartz. The equivalent circuit of the quartzcrystal was illustrated as shown in Fig. I, and the

(b)

Fig. I. -Quartz resonator and its equivalentelectrical circuit.

lecturer showed the application of the crystal as aresonator, illustrating the crevasses and the differentfrequencies of response. He then discussed theeffect of air gaps between the crystal and electrodes,the increase of frequency shown (in a slide) beingdue to increase of the air -gap having the effect of

of Quartzdecreasing the equivalent series condenser. Re-ferring briefly to other well-known piezo-electriccrystal materials, the lecturer quoted rochelle salt,which was objectionable because of the tendencyto liquefy, but suggested that tourmaline mighthave application to the higher frequencies.

Dr. Rayner then proceeded to describe in detailDye's classical work on the modes of vibrationof the quartz crystal.First he illustrated theprinciple of stationarypatterns by interferencefringes, on which Dye'smethod was based, anddescribed Dye's arrange-ments for examining thecrystal under conditionsof vibration, including hismethod of examining thecrystal under short lightflashes. Thereafter he Fig. 2.-Dye's ring oscil-showed an extensive series lator and electrode system.of Dye's pictorial results,illustrating flexural stresses and compressionalvibrations. Dye's methods of examining bothsides of the crystal while under vibration werealso discussed, and the various designs illustratedwere explained. From this the lecturer describedDye's reconstruction of a three-dimensional vibrat-ing crystal system from the two-dimensionalpictures, a model again being illustrated.

Lastly, the speaker described Dye's type of ringoscillator, developed just before his death and itsperformance being confirmed posthumously. Thering and the disposition of the electrodes is shownin Fig. 2, while the lecturer showed the method ofcutting the oscillator from the piece of quartzand of predetermining the frequency. The per-formance of this oscillator as a frequency -standardwas then shown, the zo kc. oscillator being steppeddown to r kc. for control of a phonic motor, andcompared both with a known r,000 ,-- fork anda standard clock. The performance of such anoscillator brought it definitely into the class of astandard -frequency source.


Abstracts and ReferencesCompiled by the Radio Research Board and reproduced by arrangement with the Department

of Scientific and Industrial Research.

PAGE PAGE

Propagation of Waves .. .. 206 Directional Wireless 216Atmospherics and Atmospheric Acoustics and Audio -Frequencies .. 216

Electricity 208 Phototelegraphy and Television .. 220Properties of Circuits 210 Measurements and Standards .. 222Transmission 211 Subsidiary Apparatus and Materials 224Reception .. 211 Stations, Design and Operation 228Aerials and Aerial Systems 214 General Physical Articles .. 230Valves and Thermionics 215 Miscellaneous 230

PROPAGATION OF WAVESDIE BEEINFLUSSUNG ELEKTROMAGNETISCHER

WELLEN DURCH HOCHSPANNUNGSFREI-LEITUNGEN (The Effect of High -TensionOverhead Lines on the Propagation ofElectromagnetic Waves). - H. Zuhrt.(E.N.T., Jan., 1933, Vol. to, pp. 25-38.)

When electromagnetic waves strike such a line,they produce in the pylon high -frequency currentswhich flow through the earth wire to earth on bothsides, with the result that a current distributiontypical of a T -aerial takes place. If the waveshave a frequency in the neighbourhood of thesharply marked natural frequency of such aT -aerial, reflection and screening effects occur.The magnitudes of these effects are theoreticallydetermined in this paper. For the sake of simpli-fication an infinitely long series of pylons isconsidered, equally spaced at such a distance thatthey do not mutually affect their natural fre-quencies ; the transmitter is so far away that thewave front is considered as plane, and the wavesare in resonance with the fundamental wave ofthe system.

Section I deals with the calculation of thefundamental wave (equations 5 and 5a) ; section IIwith the secondary radiation field (made up fromthe pylon radiation and the earth -wire radiation) ;section III with the radiation resistance and thecurrent induced in the pylon system ; and section IVwith the resultant field. Section V gives anumerical example, for a line with 20 -metre pylonsspaced 200 m (such as is often used for a to-kvsupply) ; it is found that at a distance of 200 metrespast the line a decrease in amplitude of aboutto% occurs.

ELECTROMAGNETIC SHIELDING AT RADIO FRE-QUENCIES.-L. V. King. (Phil. Mag., Feb.,1933, Series 7, Vol. 15, No. 97, pp. 201-223.)

This theoretical paper starts from the Maxwell -Faraday laws and finds formulae for the alternatingfield inside (a) a closed conducting spherical shellof uniform thickness, and (b) a cylindrical shell ofuniform thickness, produced by plane electro-magnetic waves of length large compared with thelinear dimensions of the shell in question. Formulaesuitable for numerical calculation are found in thecase of thin shells and a method of calculating the

shielding ratio for thin shells is developed, whichalso takes into account the effect of magneticpermeability. " Numerical calculations indicatemarked advantages in the use of shielding con-tainers of iron." An approximate formulae isgiven for the shielding ratio at radio frequencies.

MESSUNGEN DER REFLEXION AN DER HEAVISIDE-SCHICHT (Measurements of the Reflection atthe Heaviside Layer).-G. J. Elias andC. G. A. von Lindem. (E.N.T., Jan., 1933,Vol. to, pp. 1-8.)

Apart from a few slight changes in the numericaldeductions, this paper corresponds to the Dutchpaper dealt with in 1932 Abstracts, p. 396.

COMPARATIVE FIELD STRENGTHS GIVEN BY LONGAND SHORT BROADCAST WAVES [550-450 kc/s].-Byrne. (See abstract under

" Stations, Design and Operation.")

SOME CHARACTERISTICS OF SHORT-WAVE PRO-PAGATION.-J. Hollingworth. (WirelessEngineer, Feb., 1933, Vol. to, No. 113,pp. 89-90.)

Abstract of the I.E.E. paper, a short summary ofwhich was dealt with in February Abstracts, p. 91.Some of the subsequent discussion is included.

A NOTE ON AN AUTOMATIC FIELD STRENGTH ANDSTATIC RECORDER.-Mutch. (See under" Measurements and Standards.")

PORTABLE LONG WAVE TESTING APPARATUS[Field -Strength Measuring Set].-deCoutouly. (Ibid.)

NORTH ATLANTIC SHIP -SHORE RADIOTELEPHONETRANSMISSION DURING 1930 AND 1931[Analysis of Short -Wave Data : FieldStrength Contour Diagrams, etc.].-C. N.Anderson. (Proc. Inst. Rad. Eng., Jan.,1933, Vol. 21, pp. 81-101.)

SHORT-WAVE TRANSMISSION TO SOUTH AMERICA[Survey of Transmission Conditions : Absenceof Seasonal Effect : Characteristics ofAtmospheric Noise, etc.].-C. R. Burrowsand E. J. Howard. (Proc. Inst. Rad. Eng.,Jan., 1933, Vol. 21, pp. 102-113.)

For a summarised comparison with transatlantic


transmission, and reasons for the differences, seepp. 109 and 112-113.

TRANSOCEANIC RECEPTION OF HIGH -FREQUENCYTELEPHONE SIGNALS [and the Correlationof Magnetic Activity and Signal Strength].-Morris and Brown. (See under " Stations,Design and Operation.")

TIAORIE PHOTOMATRIQTJE DES ECLIPSES DE LUNE(Photometric Theory of Lunar Eclipses[Leading to the Deduction of an AbsorbingLayer at about 150 Kilometres]).-F. Link.(Comptes Rendus, 23rd Jan., 1933, Vol. 196,No. 4, pp. 251-253.)

THE IONISING EFFECT OF METEORS IN RELATIONTO RADIO PROPAGATION.-A. M. Skellett.(Proc. Inst. Rad. Eng., Dec., 1932, Vol. 20,PP. 1933-1940)

See February Abstracts, p. 92: also below.Author's summary :-" From a study. of availablemeteor data it is concluded : (1) that meteorsexpend the larger part of their energy in the Kennelly -Heaviside regions, that is, in the regions of theupper atmosphere which control the propagationof all long-distance radio waves ; (2) that themajor portion of a meteor's energy goes intoionisation of the gases around its path ; (3) thatthis ionisation extends to a considerable distancefrom the actual path-in some cases severalkilometres or more-and lasts for some minutesafter the meteor has passed ; (4) meteor trains areproduced only in the lower Kennelly -Heavisidelayer.

" A table of the various sources of ionisationof the upper atmosphere is given with values foreach in ergs cm -2 sec -2. These include sunlight,moonlight, starlight, cosmic rays, and meteors.During meteoric showers the ionising effect doesnot appear to be negligible compared with thatdue to other ionising agencies occurring at night.A meteor of one -gram mass or greater will produce,on the above assumptions, sufficient ionisationto affect propagation. One explanation of thegeneral turbulent condition of the ionised layersmay be provided by the continuous bombardmentof meteors."

OBSERVATIONS OF KENNELLY-HEAVISIDE LAYERHEIGHTS DURING THE LEONID METEORSHOWER OF NOVEMBER, 1931.-J. P.Schafer and W. M. Goodall. (Ibid., pp. 1941-1945.)

See preceding abstract and back reference." While the results are not conclusive, due to thefact that a moderate magnetic disturbance occurredduring this same period, there is some reason tobelieve that the presence of meteors in un-usual numbers caused increased ,ionisation of anintermittent nature in the region of the lowerlayer."

THE THEORETICAL FREQUENCY DISTRIBUTION OFPHOTOGRAPHIC METEORS.-P. M. Millman.(Proc. Nat. Acad. Sci., Jan., 1933, Vol. 19,PP. 34-39.)

THE INTERFERENCE PATTERN OF THREE CARRIERWAVES [Investigation of Three Transmitterseach radiating the Same Speech Message witha Frequency Separation of about 7 Kilo-cycles/Second].-A. W. Ladner and H. J.H. Wassell. (Marconi Review, Nov. -Dec.,1932, No. 39, pp. 19-29.)

SUPERPOSITION OF Two MODULATED RADIO FRE-QUENCIES [and the Distortions in theResulting Envelope : Treatment applicableto All Cases where the Signal consists of aDirect and a Reflected Beam].-Roder.(See under " Properties of Circuits.")

RADIO COMMUNICATION BY MEANS OF VERYSHORT ELECTRIC WAVES.-G. Marconi.(Marconi Review, Nov. -Dec., 1932, No. 39,pp. i-6.)

First part of the R.I. lecture dealt with in Februaryand March Abstracts, pp. 91 and 153.

FIELD STRENGTH MEASUREMENTS OF ULTRA -SHORT WAVES FROM EMPIRE STATE BUILD-ING. (See abstract under " Miscellaneous "-" Looking Ahead as 1933 Opens.")

THE ULTRA -SHORT-WAVE SERVICE BETWEENCORSICA AND THE CONTINENT.-de Clejoulx.(See under " Stations, Design and Operation.")

DER FARADAYEFFEKT IN IONISIERTEN GASENBEI WELLEN VON 4 CM LANGE (The FaradayEffect in Ionised Gases for Waves of Length4 cm.).-P. Keck. (Ann. der Physik, 1932,Series 5, Vol. 15, No. 8, pp. 903-925.)

This paper gives a full description of workalready referred to in February Abstracts, p. 92.

DISPERSIONSUNTERSUCHUNGEN MIT UNGEDAMPFTENULTRAKURZWELLEN (Investigations of Dis-persion [in Liquids] with Undamped Ultra -Short Waves).-M. Seeberger. (Ann. derPhysik, Jan., 1933, Series 5, Vol. 16, No. 1,PP. 77-99)

This paper describes measurements of the re-fractive index and absorption of various liquids(water, propyl alcohol, glycerine, solutions ofvarious salts) at wavelengths of 12.6, 13.6, 14,19 and 24 cm. Three methods were investigated(i) Lecher wire system including the liquidcontainer, (2) reaction on the emitter of the wavesreflected from the end of a trough filled with liquidand containing the emitter, (3) Lecher system withthe liquid container at the end. The last methodwas found to be the best. The numerical resultsof the measurements are given. The emitterused was that described by Collenbusch (1932Abstracts, pp. 403-404)

CHER DIE VERWENDBARKEIT DER RESONANZ METHODEZUR MESSUNG VON DIELEKTRIZITATSKON-STANTEN WASSERIGER ELEKTROLYTL6SUNGEN(The Applicability of the Resonance Methodto the Measurement of Dielectric Constantsof Aqueous Solutions).-M. Jeiewski.(Physik. Zeitschr., 15th Jan., 1933, Vol. 34,No. 2, pp. 88-94.)

C 2

April, 1933

VARIATIONS DE PHASE PAR REFLEXION SUR COUCHESMETALLIQUES TRES MINCES (Phase Changes atReflection by Very Thin Metallic Layers). -P. Rouard. (Comptes Rendus, 3oth Jan.,1933, Vol. 196, No. 5, PP. 339-341.)

Some workers have found that such reflectionproduces an advancement of phase, others aretardation, while Wernicke found sometimesone and sometimes the other, according to thestate of the metal. The writer describes his ownexperiments, with silver films of gradually varyingthickness, which show that the phase variationfor reflection at a silver -on -glass surface is first alag which, for any wavelength, passes through amaximum, diminishes, and vanishes as the thicknessincreases ; it then changes to a lead which growsvery rapidly, passes through a maximum, diminishes,and becomes constant around o.30 A. The thicknessof metal which annuls the phase difference increaseswith the wavelength. Silver -on -mica surfacesgive similar results.

THE REFLECTION OF LIGHT AT A SURFACE COVEREDBY A MONOMOLECULAR FILM [TheoreticalInvestigation]. -C. Strachan. (Proc. Camb.Phil. Soc., Jan., 1933, Vol. 29, No. 1, pp.116-130) .

This paper investigates the effect of a two-dimensional assembly of scattering centres at aplane boundary between two optically different,homogeneous media. Allowance is made for thepossibility of molecular orientation.

LICHTSTRAHLEN UND WELLENFLACHEN IN ALLGE-MEINANISOTROPEN KORPERN (Light Rays andWave Surfaces in General AnisotropicBodies). -P. Frank. (Zeitschr. f. Physik,1933, Vol. 8o, No. 1/2, pp. 4-18.)

SULLA PRESSIONE DI RADIAZIONE PER RIFRAZIONE(On the Pressure of Radiation at Refraction).-C. Bellia. (Nuovo Cim., Dec., 1932, Vol. 9,No. I0, pp. 328-334.)

THE CHARACTERISTICS OF A DEEP Focus EARTH-QUAKE : A STUDY OF THE DISTURBANCE OFFEBRUARY 20, 931.-F. J. Scrase. (Phil.Trans. Roy. Soc. London, 13th January,1933, Vol. 231, No. A 699, pp. 207-234.)

ATMOSPHERICS AND ATMOSPHERICELECTRICITY

SONNENROTATION UND LUFTSTORUNGEN (SolarRotation and Atmospherics). -F. Schindel -homer. (Naturwiss., 13th Jan., 1933, Vol.21, No. 2, p. 25.)

This letter is a second preliminary communication(for the first see 1932 Abstracts, p. 633) on thecorrelation between the 27 -day period of solarrotation and the frequency of atmospherics. Fromthe primary impulse described in the first com-munication, r, 2 and 3 solar rotation periods arecounted before and afterwards, and pulses areobtained which show a gradual doubling of themaximum given by the primary impulse. Thewriter deduces from this the existence of twoelectrically active layers on the sun of which onerotates in 27 days and the other slightly more


slowly. The existence of this second layer hasalready been considered by other geophysicalwriters.

DIRECTIONAL STUDIES OF ATMOSPHERICS AT HIGHFREQUENCIES. -K. G. Jansky. (Proc. Inst.Rad. Eng., Dec., 1932, Vol. 20, pp. 1920-1932.)

Author's summary :-A system for recording thedirection of arrival and intensity of static on shortwaves is described. The system consists of a rotat-ing directional antenna array, a double detectionreceiver and an energy -operated automatic recorder.The operation of the system is such that the outputof the receiver is kept constant regardless of theintensity of the static. Data obtained with thissystem show the presence of three separate groupsof static : Group 1, static from local thunderstorms ;Group 2, static from distant thunderstorms, andGroup 3, a steady hiss -type static of unknownorigin.

Curves are given showing the direction of arrivaland intensity of static of the first group plottedagainst time of day and for several different thunder-storms. Static of the second group was found tocorrespond to that on long waves in the direction ofarrival and is heard only when the long -wave staticis very strong. The static of this group comes mostof the time from directions lying between south-east and south-west, as does the long -wave static.

Curves are given showing the direction of arrivalof static of group 3 plotted against time ofday. The direction varies gradually throughoutthe day going almost completely around the compassin 24 hours. The evidence indicates that the sourceof this static is somehow associated with the sun.

A NOTE ON AN AUTOMATIC FIELD STRENGTH ANDSTATIC RECORDER.-Mutch. (See under" Measurements and Standards.")

RELATION ENTRE LES VARIATIONS DIURNES DUCOURANT TELLURIQUE ET DU CHAMP MAG.NETIQUE TERRESTRE (Relation between theDiurnal Variations of the Earth Currentand of the Terrestrial Magnetic . Field). -D. Stenquist . (Comptes Rendus, 16thJan., 1933, Vol. 196, No. 3, pp. 205-207.)

ON THE RELATIONS OF STELLAR ELECTRICITY ANDMAGNETISM TO THE PHENOMENA OF THESUN'S ATMOSPHERE. -R. Gunn. (Science,23rd Dec., 1933, Vol. 76, No. 1982, Pp. 577-583.)

ON THE PROPAGATION OF A LIGHTNING DISCHARGETHROUGH THE ATMOSPHERE. -E. C. Halliday.(Phil. Mag., Feb., 1933, Supp. No., Series 7,Vol. 15, No. 98, pp. 409-420.)

From the author's summary :-Experimentshave been described in which a camera with rotatinglenses has been successfully used to obtain someindication of the manner of propagation of threelightning flashes. It has been shown that it ispossible for a lightning discharge to be propagatedfrom the ground towards the cloud, and that aflash starting at the ground may join up with aflash starting in the cloud, as was earlier shown


by Boys. Two instances have been describedin which discharges started at a point above theground and were propagated both upwards anddownwards. The photographs have demonstratedthat two flashes may, for part of the distancebetween cloud and ground, traverse the same path.A flash has been photographed showing forkingtowards the cloud at the top and towards theground at the bottom. In this case at least, thedirection of the forking cannot be a criterion of thedirection of propagation of the flash.

SPARK -GAP BREAKDOWN : CHARACTERISTICPROPERTIES OF THE GLOW DISCHARGE.-R. Hellman : Rogowski. (Electrician, 13thJan., x933, Vol. Ito, No. 285o, pp. 31-32.)See also Rogowski, 1932 Abstracts, p. 635.

DIE KORPIJSKULARE NATUR DER ULTRASTRAHLUNGUND IHR ERDMAGNETISCHER EFFEKT (TheCorpuscular Nature of Cosmic Radiationand the Effect of Terrestrial Magnetism).-J. Clay. (Naturwiss., loth Jan., 1933, Vol.21, No. 3, pp. 43-44.)

A short account of work already published inProc. Roy. Acad. Amsterdam, 17th Dec., 5932.The writer has confirmed and extended resultsalready obtained (Jan. Abstracts, p. 34) on thevariation of cosmic ray intensity with magneticlatitude (for which he gives a graph) which indicatedthat cosmic radiation was corpuscular by nature.He determines the minimal value of the energywhich the rays must have to penetrate the earth'satmosphere and gives a curve for the distributionof corpuscles as regards energy and the variationof the primary rays with the earth's magneticfield. He reaches the conclusion that the primarycosmic radiation observed at the earth's surfaceis the part possessing most energy of a corpuscularradiation of almost exponential distribution andmean energy 3 x io10 electron volts. This radiationis cut off at 3.6 x Io9 electron volts, and thisexplains the observed constancy of the radiationbetween 5o° magnetic latitude and the magneticpoles. He also finds that the radiation mustbecome harder as the equator is approached.

BEMERKUNG ZUM INTENSITATSVERLAUF DERULTRASTRAHLUNG IN GROSSEN HollEN(Remark on the Variations of Cosmic RayIntensity at Great Altitudes).-H. Kulen-kampff. (Naturwiss., 13th Jan., 1933,Vol. 21, No. 2, pp. 25-26.)

Recent measurements of Regener ( Jan. Abstracts,p. 33, and below), appear to contradict the presenceof a maximum in the altitude/intensity curve ofcosmic rays, which was hitherto expected to beproduced by the gradual saturation of primaryby secondary radiation. The writer of this lettershows that consideration of the effects of the earth'smagnetic field in producing curvature of the pathsof the secondary electrons gives no theoreticalmaximum, but a gradual tendency to a limitingvalue, as found experimentally by Regener.

ENERGY OF COSMIC RAYS.-E. Regener. (Nature,28th Jan., 1933, Vol. 131, p. 13o.)

This preliminary letter corrects the value of the

intensity /co of cosmic radiation at its entranceto the atmosphere, given recently by the writer(Abstracts, 1932, p. 634 ; also Jan., p. 33), to a valuecorresponding to the production of 333 pairs ofions per cc per sec. in air at N.T.P. 1.02 x 108pairs of ions are found to be produced altogetherby total absorption of cosmic rays by a column ofair of x sq.cm section. " The flux S of energycoming to the earth from the cosmic rays is foundto be 5.2 x to-e erg per sq. cm per sec." Thetemperature of space corresponding to the flux ofcosmic rays is also considered.

COSMIC RAYS MAY BE BORN OF NEUTRAL RAYS FROMSTARS.-R. Gunn. (Sci. News Letter, 21stJan., 1933, Vol, 23, No. 615, p. 41.)

A note on a new theory suggested by Gunn." Out in the stars that are still young there areborn in great ' star -spots,' like sun spots on our.sun, negatively charged ions or electrons which aregiven great energies. These pass through the outerlayers of the star and each .picks up a positive ion,and the two together form an energetic neutralpair of some to to too thousand million electronvolts. When such particles hit the earth theybecome separated in the upper atmosphere byelectromagnetic forces or collisions and becomecosmic rays."

MILLIKAN AND COMPTON DEBATE COSMIC RAYFACTS AND THEORIES.-Millikan : Compton.(Sci. News Letter, 7th Jan., 1933, Vol. 23,No. 613, pp. 6-7.)

ON COMPTON'S LATITUDE EFFECT OF COSMICRADIATION.-G. Lemaitre and M. S. Val-larta. (Phys. Review, 15th Jan., 1933,Series 2, Vol. 43, No. 2, pp. 87-91.)

The full paper, an abstract of which was dealtwith in March Abstracts, p. 152.

EINSTEIN BACKS LEMAITRE IDEA THAT COSMICRAYS ARE BIRTH CRIES.-Lemaitre. (Sci.News Letter, 21st Jan., 1933, Vol. 23, No.615, p. 37.) See Feb. Abstracts, p. 95.

How FAR Do COSMIC RAYS TRAVEL ?-F. Zwicky.(Phys. Review, 15th Jan., 1933, Series 2,Vol. 43, No. z, pp. 147-148.)

This preliminary letter examines the two proposedsuggestions as to the nature of cosmic rays, (t) thatthey are of local origin, and (2) that they have beenproduced throughout interstellar or intergalacticspaces, and establishes " a relation between themand the red shift of extragalactic nebulae."

'USER DIE DRUCK- UND TEMPERATURABHANGIGKEITDER IONISATION BEI ULTRASTRAHLUNG (TheDependence of Cosmic Ray Ionisation onPressure and Temperature).-B. Gross.(Zeitschr. f. Physik, 1933, Vol. 8o, No. 1/2,pp. 125-133.)

X-RAYS MADE AS COSMIC RAYS PLOUGH THROUGHGAS [and cause Part of the Charge in anIonisation Chamber used for Ray Measure-ments].-G. L. Locher. (Sci. News. Letter,21st Jan., 1933, Vol. 23, No. 615, p. 37.)

April, '933 210 THE WIRELESS ENGINEER &

SOME COMMENTS ON THE RELATION BETWEENIONISATION AND IONISATION CURRENT INGASES AT HIGH PRESSURES.-W. R. Harper.(Proc. Carob. Phil. Soc., Jan., 1933, Vol. 29,No. I, pp 149-155.)

" The phenomena which limit saturation in ahigh-pressure ionisation chamber are examined withparticular reference to the role played by preferen-tial recombination."

STUDIE USER DIE METHODIK DER IONENZAHLUNG(Study of the Methods of Ion Counting [inAir]).-H. T. Graziadei. (Physik. Zeitschr.,15th Jan., 1933, Vol. 34, No. 2, pp. 82-88.)

PROPERTIES OF CIRCUITS

TRANSMISSION CURVES OF HIGH -FREQUENCY NET-WORKS.-S. J. Model. (Proc. Inst. Rad. Eng.,Jan., 1933, Vol. 21, pp. "4-143.)

From the Radio Transmitter Research Institute,Leningrad. Author's summary :-The aim of thisarticle is to deduce the design data on frequencytransmission, which are necessary in designing thenetworks for high -frequency modulated waves. Inthe first section the general laws of current variationsin circuits tuned to carrier frequency with andwithout tube .generator are derived ; the inter-dependence of current curves of various circuitsforming a given network is also clarified.

The second section gives the derivation of thetransmission curve equation for a two -circuit system,analyses the curve shape as dependent upon theparameters of the circuits and the coupling factorbetween them, and also explains the part the tubegenerator plays in restricting the field of applicationof formulas derived. Moreover, the equation of thecurrent curve for inaccurately tuned circuits isgiven, which shows that inaccurate tuning resultsin asymmetrical curves and for that reason is notto be recommended. The third section contains thederivation of the transmission curve equation fora three -circuit system and the analysis of the curveshape as dependent on parameters of circuits andcoupling factors between them.

EQUIVALENT CIRCUITS OF AN ACTIVE NETWORK.-J. G. Brainerd. (Proc. Inst. Rad. Eng., Jan.,1933, Vol. 21, pp. 144-153.)

Author's summary :-The problem considered inthis paper is that of setting up an exactly equivalentcircuit for a four -terminal network containing onee.m.f. and any number of impedances connected inany manner. Several relatively simple circuitswhich can be used to represent any such active net-work, and some others which can be used in specialcases, are completely specified.

VEER DIE THEORIE DES FREISCHWINGENDENKREISES (On the Theory of the Freely Oscil-lating Circuit).-M. Rousseau : Osnos.(Hochf:tech. u. Elek:akus., Jan., 1933, Vol.41, pp. 27-28.)

After referring to Osnos' papers (1932 Abstracts,pp. 457 and 635-636) the writer gives some moretransformations of the usual formulae and anumber of deductions to be made from them. Thus

" half the ohmic resistance must equal the geo-metric mean of the inductive resistance and theabsolute value of the whole (necessarily negative)inductive -capacitive resistance." Again, the con-dition for oscillatory discharge is shown to be thatthe time constant of the L, R system must begreater than a quarter of the relaxation time of theC, R system. The writer uses one section to showthat the use of the ideas of " impedance " and" non -ohmic resistance " is only logical in the theoryof forced oscillations.

SOME NOTES ON DEMODULATION.-Roder. (Seeunder " Reception.")

SUPERPOSITION OF Two MODULATED RADIO FRE-QUENCIES [and the Distortions in the Re-sulting Envelope].-H. Roder. (Proc. Inst.Rad. Eng., Dec., 1932, Vol. 2o, pp. 1962-197o)

Author's summary :-If two modulated radiofrequencies are added, the envelope of the resultingsignal can be found by means of a vector diagram.A number of envelopes corresponding to variousconditions are shown. The distortions in the re-sulting envelopes are mainly due to phase shifts ofaudio frequency. Low percentage modulation helpsto keep distortions small. The treatment appliesto common -frequency broadcasting and to all casesin which the received signal consists of a direct anda reflected beam.

FREQUENCY DOUBLING IN A TRIODE VACUUM TUBECIRCUIT.-C. E. Smith. (Proc. Inst. Rad.Eng., Jan., 1933, Vol. 21, pp. 37-50.)

Author's summary :-" This paper gives aquantitative analysis of operating performance ofa triode vacuum tube as a frequency doubler. Withslight changes this analysis can be applied totripling, quadrupling, etc. Three methods of attackhave been outlined : theoretical analysis, graphicalsolution, and experimental results. The theoreticalanalysis furnishes a general solution to the problem,from which general conclusions as to the bestoperating conditions can be drawn. The graphicalsolution furnishes a simple method by which rapidcalculations can be made for a particular case. Theexperimental results are valuable for comparisonbecause they give the actual results of operation.

" The primary object of this work has been toinvestigate the conditions that will give maximumplate efficiency and consequently the most de-sirable operating conditions, also keeping in mindthat power output and power amplification areimportant factors to consider in the practicalapplication of the frequency doubler." Thewriter refers to the work of Marique, Mesny, andBunimowitch (Abstracts, 1929, p. 325 ; 1930, p.332 ; and 1931, p. 209).

VOLTAGE AMPLIFICATION WITH HIGH SELECTIVITYBY MEANS. OF THE DYNATRON CIRCUIT.-Colebrook. (See under " Reception.")

GRID CURRENT COMPENSATION IN POWER AMPLI-FIERS.-Baggally. (See under " Acousticsand Audio -frequencies.")


TRANSMISSIONBERECHNUNG DES ANFACHMECHANISMUS VON

SCHWINGUNGEN BEI EINER ELEKTRONEN-. RoHRE IN BARKHAUSEN-KURZ-SCHALTUNG

(Calculation of the Mechanism of OscillationOnset with a Valve in the Barkhausen-KurzCircuit).-H. Edler. (Archiv f. Elektrot.,9th Dec., 1932, Vol. 26, No. 12, pp. 841-849.)

The writer adopts energy considerations as hisstarting -point and finds a differential equationgoverning the motion of electrons in the valve,under the assumption that an external alternatinge.m.f. from the connected Lecher system actsbetween anode and cathode of the valve ; he neglectsinternal friction and interaction between theelectrons themselves. Solving the equation, hefinds an expression for the amplitude of oscillationof the electrons which itself varies with time. Thefluctuations of the amplitude give the boundariesof the oscillation regions ; the amplitudes of theoscillation voltages can also be calculated and theinfluence exerted by the dimensions of the valve andthe voltages on its electrons can be estimated.The maximum developable power is determined,with the conditions under which it is attained.Good agreement is obtained with the experimentalresults of Wundt (1931 Abstracts, pp. 94-95)SHORT- AND ULTRA -SHORT-WAVE TRANSMITTING

CIRCUIT WITH SEVERAL LECHER PAIRS OFMICA -INSULATED COPPER STRIP.-H. Rotheand H. 0. Roosenstein. (Summary ofGerman Patent in Hochf:tech. u. Elek:akus.,Nov., 1932, Vol. 4o, No. 5, p. 184.) SeeFig. 17.

POINTS OF CONNECTION OF FEEDERS TO LECHERSYSTEM IN ULTRA -SHORT-WAVE TRANS-MITTER.-B.D.H. Tellegen : Philips' Com-pany. (Ibid., p. 184.)

To prevent the leakage of r.f. energy to thecathode, the feeders are not connected to thepotential nodes but are symmetrically displacedwith respect to these, the points 14 and 15 (Fig.18) being so chosen that no a.c. potential differenceexists between them and the cathode. The choke 13can then be dispensed with.

DISCUSSION ON " A NEW CIRCUIT FOR THE PRO-DUCTION OF ULTRA -SHORT-WAVE OSCIL-LATIONS." -N. Carrara : Kozanowski. (Proc.Inst. Rad. Eng., Dec., 1932, Vol. 20, p.1975)

Pointing out the similarity between the Kozan-owksi circuit and Carrara's own circuit-cf. Febru-ary Abstracts, p. 114, Carrara.

AN OSCILLATOR HAVING A LINEAR OPERATINGCHARACTERISTIC [and High Frequency Sta-bility]. --L. B. Arguimbau. (Proc. Inst. Rad.Eng., Jan., 1933, Vol. 21, pp. 14-28.)

Author's summary :-" A review of conventionallinear equilibrium conditions is given. It is shownthat these conditions are not usually at all applic-able to practical oscillators because the operatingregion is not to be so simply described. The relationof non-linear effects to frequency modulation ispointed out. A modified type of oscillator whichconforms to the elementary linear conditions is

211 April, 1933

described and its properties are discussed. Theresults are applied to a conventional grid -leak andcondenser oscillator to explain its modulatingcharacteristics. The simplicity of the new circuitpermits a discussion of amplitude inertia effects."

In conclusion, the author sums up :-" It hasbeen seen that an oscillator, where the change in.plate resistance is brought about by an auto-matically shifted bias, conforms closely to lineartheory. Computations show that such an oscillatorhas high frequency stability and linear modulatingcharacteristics. It has been shown that dynamiceffects can be minimised by a proper choice of theconstant k [equation ro]. In the practical designof an oscillator, a compromise is usually necessarybetween freedom from frequency modulation andfreedom from dynamic hysteresis. A desirablesolution to both of these troubles can usually beobtained by choosing a sufficiently high proportion-ality factor between the rectifier output and theamplitude of swing. This sharper control of equi-librium conditions is the essential advantage of thelinear oscillator which has been described."

CARRIER NOISE IN SHORT-WAVE TRANSMITTERS.-A. J. A. Gracie and E. J. C. Dixon. (P.O.Elec. Eng. Journ., Jan., 1933, Vol. 25, Part4, Pp. 300-303.)

The writers sum up as follows :-Carrier hum isincreased by doubler circuits adjusted to operate atgood efficiency, and may limit the commercialeffectiveness of the transmitter. In telephonetransmitters, where the carrier hum must be at avery low level, it is best to eliminate the ripple atthe source by adequate smoothing arrangementsand the use of d.c. supplies wherever possible. Intelegraph transmitters .a.c. hum may be saturatedout at the doubler stages, and the required modu-lation applied at their output. The saturation pointin bright -emitting cathodes of high efficiency isfairly critical, and the effect of using valves of highergain with dull -emitting filaments is being investi-gated.

LES EMETTEURS A ONDES ENTRETENUES MODULfESDES STATIONS COTIERES FRAN cAISES (TheModulated C.W. Transmitters of the FrenchCoast Stations).-M. Bruniaux. (Ann. desP.T.T., Jan., 1933, Vol. 22, No. I, pp.59-78.)

NEW B.B.C. HIGH -POWER TRANSMITTER TO RE-PLACE DAVENTRY LONG -WAVE SET [using" Series Modulation "].-(Marconi Review,Nov. -Dec., 1932, No. 39, p. 3o.)

Paragraph mentioning that the new trans-mitter will employ " a new system of modulation,known as ' series modulation.' " up to 9o%. Thefrequency characteristic will be straight between3o and ro 000 c/s.

RECEPTIONFILTER FUR ZWISCHENFREQUENZVERSTARKER (A

Filter for the Intermediate -Frequency Ampli-fier [of a Superheterodyne Receiver]).-H.Piesch. (Hochf:tech. u. Elek:akus., Jan.,1933, Vol. 41, pp. 23-26.)

Description and analysis of " a new type of filter

April, 1933 21

circuit " specially suitable for the purpose named.A single stage is found to be considerably superiorin sharpness of tuning and in shape of characteristicto a single -stage i.f. amplifier with two resonancetransformers. Within the band-pass zone the no-load output voltage is higher than the input voltage.This property and the fact that its point of maximumeffect can be adjusted within the pass band, enablesthe filter to be adjusted if desired to correct anyfrequency distortion in the r.f. or a.f. part of thesuperheterodyne receiver.

Each series element of the filter consists of aresonance circuit with a capacity in series ; eachparallel element consists of a similar combination.The calculated damping characteristic for a loss -free filter for a band of Io kc/s is shown in Fig. 3,left-hand curve : the measured characteristic isgiven by the right-hand curve. Fig. 4 refers to asingle -stage filter with a middle frequency of 77kc/s and a band -breadth of only 5 kc/s. The left-hand curve shows the frequency characteristic ofthe output voltage, the right-hand curve that of theratio output voltage/input voltage. The latter curve,with its dip near the middle frequency, represents afilter which could well be used with a superhetero-dyne receiver having a sharply tuned r.f. part ;the dip would compensate the r.f. resonance curveand give the combination the desired rectangularcharacteristic. On the other hand if compensationis required in the a.f. part of the receiver, thiscorrection can be accomplished by detuning theparallel element of the filter.

For the broadcast and short-wave receivers atpresent on the market, a single -stage filter is selec-tive enough, but there is no difficulty in extendingthe principle to two stages (Fig. 6).

FIDELITY COMPENSATION BY REGENERATION [byFeeding -back Part of Output in Phase toScreen -Grid Circuit of Tetrode].-A. C.Matthews. (Electronics, Jan., 1933, p. 18.)

The compensation applied to the s.g. circuitdoes not affect the otherwise normal frequencycharacteristic. The impedance of the s.g. circuitdecreases with frequency, so that the voltage fedback through the tetrode is greater for the lowerfrequencies. If compensation at the higher fre-quencies is also required, a double -humped curve isobtained by a series resonant circuit shunted acrossone of the s.g. circuit resistors.

EIN NEUARTIGER NIEDERFREQUENZVERSTARKER (ANew Type of L.F. Amplifier [Glow -DischargeTube Coupling : for Frequencies from o toso 000 c/s]).-O. Schafer : H. Peek. (Radio,B., F. f. Alle, Feb., 1933, pp. 49-55.)

An almost rhapsodic appreciation of Peek's newamplifier for direct and alternating currents, dealtwith in 1932 Abstracts, p. 535. Apart from its usefor photoelectric cell researches (for which it is" outstandingly suitable ") it is admirably adaptedto the broadcast reception of the local station, incombination with a very good loud speaker : theperfection with which the lowest notes are repro-duced " spoils " the listener for the ordinaryreceiver. On long-distance reception its advan-tages are wasted : even the local station trans-missions are not worthy of it if too long a cable is


involved. The new amplifier can be simply andcheaply constructed from ordinary commercialcomponents.

The writer discusses at length the developmentof the new circuit from the ordinary direct -coupledamplifier circuit with a compensating battery inthe grid lead, and from the Loftin -White direct -coupled circuit which, unlike the latter, is suitablefor mains operation. The advantages of the newcircuit over these, and over the resistance -coupledamplifier, are described : as regards the latteramplifier, the fundamental advantage presentedby the new system is that the glow -discharge tubehas, for a.c., a resistance practically independentof frequency and ranging from about so to looohms according to the design of the tube. Finally,he gives the design data and diagram of a two -valveamplifier for use with an a.c. mains unit, giving atabout I c/s an amplification factor of half thenormal value (the latter, with an output valve offactor Jo, being about 270), while between 5o and8 000 c/s the amplification is constant, " which can besaid of hardly any other circuit ! " A worn-outtriode is used, in conjunction with a bell trans-former in parallel with the heating circuits, as arectifier to provide the necessary grid voltage. Amilliammeter in the anode circuit of the outputvalve is necessary for the adjustment of theamplifier.

VOLTAGE AMPLIFICATION WITH HIGH SELECTIVITYBY MEANS OF THE DYNATRON CIRCUIT.-F. M. Colebrook. (Wireless Engineer, Feb.,1933, Vol. so, No. 113, pp. 69-73.)

Author's summary :-In an ordinary tuned -circuit triode or tetrode amplifying stage withoutretroaction, the selectivity is less than that of thetuned circuit alone. If, however, a tetrode valve isused in a dynatron or negative resistance condition,the selectivity of the amplifying stage is con-siderably greater than that of the tuned circuitalone, and the voltage amplification comparablewith that given by the valve used in the normalmanner. With a given tuned circuit at a givenfrequency the normal and the dynatron tetrodeamplifying circuits can only be made comparablein selectivity by using a high degree of retroactionwith the former. In this condition the voltageamplification given by the normal circuit will exceedthat of the dynatron circuit, but the dynatron cir-cuit has the advantage of simplicity, since noadditional winding or other auxiliary apparatus isrequired.

The above distinction between the two types ofcircuit is deduced analytically and confirmed byexperiments with a tuned audio -frequency ampli-fier.

SOME NOTES ON DEMODULATION.-H. Roder.(Proc. Inst. Rad. Eng., Dec., 1932, Vol. 20,PP. 1946-1961.)

Author's summary :-" The audio output of adetector depends upon two magnitudes which areindependent each from the other. These magni-tudes are the radio -frequency envelope of the inputand the rectification characteristic of the detector.In problems dealing with the distortions caused byselective circuits or by the superposition of several


modulated signals it becomes necessary to treatboth magnitudes separately.

" A new method to find graphically the radio -frequency envelope for the 'general modulatedsignal,' is given. An analysis is made showing howto find the detector output if the input radio -frequency envelope and the rectification char-acteristic of the detector are known."

The author feels justified in dealing with thesubject in some detail because " numerous pub-lications on these problems do not consider thatdistortions in the audio output are due to twoindependent sources, one of which is arbitrarilyadjustable to a much higher extent than the other."DISCUSSION ON " SOME NOTES ON GRID CIRCUIT

AND DIODE RECTIFICATION."-F. E. Ter -man : Nelson. (Proc. Inst. Rad. Eng., Dec.,1932, Vol. 20, pp. 1971-1974.)

Terman, referring to Nelson's paper (1932Abstracts, p. 583), points out that the expressionthere given for the input resistance, R/2cos28,yields a value always higher than that actuallyobtained ; it is derived by assuming that thelosses in the input resistance equal the energyconsumed by the grid leak, so that the energy lostin the rectifying device at the input electrode isneglected. He also considers Nelson's relationfor distortionless grid rectification, to beless strictly correct than the Terman-Morganrelation XIR.InIVI - m1, though unless thedegree of modulation is high the difference is toosmall to be very important.

Nelson replies : Fig. 3 of his paper shows thatR/2cos20 gives results which are too high, while theTerman-Morgan R/2cos8 gives (as the authorsshowed) results which are too low. His expressionX/R > m picks the point where the rate of change ofthe envelope is greatest, i.e. when either a sine ora cosine wave is passing through zero. " This holdsfor all percentages of modulation and this theoryexplains why it is possible to obtain good qualitywith i00% modulation and X/R > 1." He addsthat " some might object to the theory developedby saying that the distortion is caused by a lowratio of a.c. to d.c. impedance. There is sometruth in this, particularly when the X/R ratio isvery low. The writer hopes to cover this subjectin the near future, as a few preliminary checks showthat the effect of too low a ratio of a.c. to d.c.impedance may be separated from the effect dis-cussed here."INTER -CARRIER NOISE SUPPRESSION : A NEW

SYSTEM [using the Wunderlich B Valve andAutomatic Volume Control].-N. E. Wunder-lich. (Electronics, Jan., 1933, p. 13.)

" A simplified system which requires no additionaltubes and hence is much less expensive than morecomplicated systems, although performing thesame functions. These results are attained bytaking advantage of circuit arrangements madepossible by the strategic electrode construction ofthe Wunderlich B tube, which has a longer cathodeand a small additional anode placed at the top ofthe structure and shielded from the other elements.By this extra anode, the AVC potentials areamplified before being applied.

" Performance curves show the threshold '

level of the receiver may be set to any desiredvalue while retaining the AVC action. The sus-taining action of the AVC may be caused to releaseat any desired lower signal level. The point ofrelease may be adjusted by means of the sensitivitycontrol." This action is very sharply definedbecause of the above -mentioned amplification ofthe AVC potentials.

AUTOMATIC VOLUME CONTROL [with Double -DiodeTriode, giving Quiet, Amplified and De-layed AVC].-C. N. Smyth. (WirelessWorld, 17th Feb., 1933, Vol. 32, psi. 134-136.)

CIRCUITS FOR AMPLIFIED AUTOMATIC VOLUME CON-TROL [using the Duplex Diode Triode].-L. E. Barton : C. Travis. (Electronics,Jan., 1933, pp. 16-17.)

For references to these valves see Abstracts,Jan. and Feb., pp. 41 and toe, 1-h cols. ; 1932, p.587, r -h col. In the circuit here described thesteady d.c. voltage generated by the diode detectoris amplified in the triode unit of the valve togetherwith the detected a.f. signal, and is used as anAVC bias after this amplification. " There resultsa very sensitive and level AVC system." Theadvantages of the circuit are analysed.

ADDING AUTOMATIC VOLUME CONTROL TO ANEXISTING RECEIVER WITH A VARIABLE -MUH.F. STAGE.-( World Radio, loth Feb., 1933,Vol. 16, No. 394, p. 200.)

NEW AUTOMATIC BIAS SCHEME.-E. G. Bowen.(Wireless World, 24th February, 1933, Vol.32, P. 157.)

There is a definite tendency among con-structors of battery sets to follow mains practice andprovide automatic bias for both r.f. and a.f. valves.A new bias scheme is described in this article whichovercomes the difficulties hitherto associated withbattery variable -mu valves.

NEW APPARATUS AND APPLIANCES [Special Trans-former and Output Choke for " QuiescentPush-Pull."].-(Electrician, loth Feb., 1933,Vol. r io, No. 2854, pp. 200-201.)

Q.P.P. WITH TRIODES . -W. I. G. Page. (WireleisWorld, 24th February, 1933, Vol. 32, pp.152-153.)

Up to the present the principles of quiescentpush-pull have been applied to pentode valves.Where, however, the major consideration is valvecost rather than power output, triodes in quiescentpush-pull will probably find wide application,especially as matching is not essential and moving -iron speakers can be used satisfactorily.

AN OBSCURE CAUSE OF POOR RECEPTION [FaultyCombination of Parallel -Feed Transformerand Autorhatic Grid Bias Resistor].-M. G.Scroggie. (World Radio, loth Feb., 1933,Vol. 16, No. 394, p. 202.)

THE ALL -WAVE MONODIAL SUPER.-W. T. Cocking.(Wireless World, 27th January and iothFebruary, 1933, Vol. 32, pp. 6o-63 and109-112.)

A seven -valve battery -operated receiver for the

April, 1933

amateur constructor. The principal feature is thatthe receiver covers a waveband of 12 to roo metresin addition to the customary medium and longbroadcast wavebands.

AUTO -RADIO AN EXPANDING MARKET IN 1933POLICE RADIO INCREASES.-(Electronics, Jan.1933, PP. 8-9.)

DIE FUNKINDUSTRIE IN DER SAISON 1932/1933(The Radio Industry in the Season 1932/1933[Analysis of Types, Prices and Sales]).-H.Menzl. (E.T.Z., 12th Jan., 5933, Vol. 54,No. 2, pp. 25-26.)

COUNTERACTING FADING BY PERIODIC DETUNINGOF RECEIVER.-Deutsche Tel:werke u. Kabel-ind. A.G. (Summary of German Patent inHochf:tech. u. Elek:akus., Nov., 1932, Vol.40, No. 5, p. 184.)

The receiver is detuned by 500 to r 000 c/s ata frequency of from 5 to 5o per second, by meansof a rotating condenser or a valve device.

AIRPLANE RADIO FREED FROM IGNITION INTER-FERENCE.-Jeuldas. (See under " Aerialsand Aerial Systems.")

AERIALS AND AERIAL SYSTEMSANPASSUNG VON SPEISELEITUNGEN AN KURZ-

WELLEN-SENDEANTENNEN (Matching Feederswith Short- [and Ultra -Short-] Wave Trans-mitting Aerials).-S. Issakowitsch-Kosta.(E.N.T., Jan., 1933, Vol. 10, pp. 9-19.)

Of the four components into which the over-allefficiency of a transmitting system can be divided,namely the partial efficiencies of generator, feeder -coupling circuit, feeder, and dipole aerial, thewriter deals with the middle two. The matchingconditions are derived for the single -wire feederwith return through the dipole capacity to earthand the earth itself, and for the parallel -wire feeder.With the former, a transmission of energy is onlypossible when the matching is perfect : with thelatter and with the mechanically difficult con-centric feeder (which is not dealt with in thispaper, except that a reference is given to Moser'swork-Abstracts, 1929, p. 105), the conditions arenot so critical. The efficiencies and over-all damp-ing are found in terms of Roosenstein's " waveratio " (the ratio of the potentials at node andantinode-1931, p. 36: see also p. 441), and thechoice of the point or points of connection on theaerial is investigated. Finally, the theoreticalresults are confirmed by tests on a 9-m wave.

RECHNERISCHE ERMITTLUNG DER IMPEDANZ VONANTENNEN (The Calculation of the Impe-dance of Aerials [with Mid -Point Feed : forVarious Values of 1/A)].-J. Labus. (Hochf:tech. u. Elek:akus., Jan., 1933, Vol. 41, pp.17-23.)

The usual practical formula for the reactance of anaerial is X = -1/ITC . cot (colVLC), which isderived under the assumption that the aerial can betreated as a homogeneous line with uniformlydistributed inductance and capacity. This formula


is strictly applicable only when the aerial length isnegligible compared with the wavelength. Otherworkers have calculated the reactance of centre -fedaerials (e.g., Carter, 1932 Abstracts, p. 585) butonly for the special case where the aerial length is a.whole number of half wavelengths (dipole aerial).

The writer therefore starts from the Maxwellequations and calculates, with the help of the lineintegral (r), the required impedance, which is madeup of tvzo components, one real (the radiation re-sistance) and the other imaginary (the reactance).To do this he first determines the scalar and vectorpotentials at the surface of the conductor. Hearrives at equation (18) for the reactance at thefeeding point : if the aerial length is small com-pared with the wavelength this simplifies down toequation (19), in which the only determiningvalues are the length and radius of the wire. Theimportant difference between this latter formula andthe usual one quoted at the beginning of thisabstract is that by the former the reactance isindependent of the wire spacing d of the feeder (Fig.lb), whereas by the latter the reactance involvesthis factor. The discrepancy between the twoformulae is shown in Fig. 6, where the reactance isplotted against different values of //A.

THE RADIATION CHARACTERISTICS OF A VERTICALHALF -WAVE ANTENNA.-J. A. Stratton andH. A. Chinn. (Proc. Inst. Rad. Eng., Dec.,1932, Vol. 20, pp. 1892-1913.)

Authors' summary :-Extensive measurementshave been made of the field distribution about avertical antenna operating at 29 megacycles oversea water. With the help of a small, non -rigidairship observations were made at various altitudeswhich permitted the plotting of the space char-acteristic and the determination of the attenuationas dependent on altitude. The experimentalresults are compared with field intensities computedfor the given physical conditions from the theo-retical expressions of Sommerfeld and Strutt. Ofinterest is the manner in which the field is attenu-ated in the immediate vicinity of the ground plane,and the marked effect on the general intensitydistribution of the effective height of the antennaabove ground.

Details are given of the design of the apparatusrequired for carrying out the measurements.

MUTUAL IMPEDANCE OF Two SKEW ANTENNAWIRES.-F. H. Murray. (Proc. Inst. Rad.Eng., Jan., 1933, Vol. 21, pp. 154-158.)

Author's summary :-In a recent communicationP. S. Carter derived a formula for the mutualimpedance of two antenna wires intersecting at an.angle 8, as a sum of two definite integrals [1932Abstracts, p. 585]. It has been shown that relateddouble integrals can be reduced to sums of logar-ithmic integrals and their limits ; in the presentpaper this reduction will be generalised. Themutual impedance of two antenna wires in spacewhich do not intersect may be calculated in terms ofthe integrals discussed, but these integrals can beevaluated by known tables only if the wires are inthe same plane. The integrals of Carter result froma suitable limiting operation ; a direct evaluationis also possible and is given.


ON THE NUMERICAL CALCULATION OF THE CURRENTIN AN ANTENNA [and the Reduction ofRelated Double Integrals]. -F. H. Murray.(American Journ. of Mathematics, 1931, Vol.53, pp. 889-89o.) See also precedingabstract.

THE NEW HIGH -POWER BROADCASTING STATIONAT VIENNA, WITH REFLECTOR TOWER FORDIRECTIVE WORKING. -Telefuitken Com-pany. (E.T.Z., 9th Feb., 1933, Vol. 54,No. 6, pp. 137-138.)

Rapid progress is being made with Bisambergstation, on a mountain ridge to the north of Vienna.Since it lies near the north-eastern boundary ofAustria, radiation is required to be chiefly in thewest and south-west directions, and therefore areflector tower (13 -metre iron lattice), exactlysimilar to the aerial tower, is being erected IIometres to the east of the latter (cf. Wilmotte, MarchAbstracts, p. 158). Both towers oscillate at a quarterwavelength. A capacity earth (a network of 12kilometres of copper wire) is supported on 36 smallmasts. The modulated aerial circuit power is15o kw.

AIRPLANE RADIO FREED FROM IGNITION INTER-FERENCE. -C. F. Jenkins. (Sci. NewsLetter, 21st Jan., 1933, Vol. 23, No. 615,P. 40)

Brief summary of recent U.S. Patent. Theinventor found that radiation from the engineignition system spread spherically, but a cone -shaped zone free of radiation existed behind theplane, having its apex at a point directly behind theengine. The effect is probably due to absorption bythe metal parts of the fuselage. The antenna isflown from the tail structure axially in this zoneand thus avoids ignition interference.

DIE ELEKTRISCHE RAUHREIF-ABHEIZUNG VONFREILEITUNGEN (The Electrical Heatingof Overhead Lines against Hoar Frost). -K. Halbach. .(E.T.Z., 12th Jan., 1933, Vol.54, No. 2, PP 33-36.)

SCHWINGHEBELDAMPFER FOR FREILEITUNGEN (AVibration -Damping Lever for Attachmentto Overhead Lines near Point of Sus-pension). -H. Schmitt. (Zeitschr. V.D.I.,14th Jan., 1933, Vol. 77, No. 2, pp. 52-53.)

VALVES AND THERMIONICSOUTPUT STAGES : THE CHOICE OF VALVES FOR USE

IN LOW -FREQUENCY AMPLIFIERS [RapidTest Method]. -C. C. Whitehead : Brain.(Wireless Engineer, Feb., 1933, Vol. to, No.113, pp. 78-82.)

After giving a résumé of Brain's paper on deter-mining the operating conditions for a given valve formaximum undistorted output (1929 Abstracts,p. 388), the writer describes an extension of themethod to a rapid test on commercial power valvesto determine their suitability or otherwise for use inthe output stage of any given amplifier, the anodevoltage of which is limited to a definite figure." Incidentally, it also led to the solution of thefollowing problems :-(I) the minimum anode vol-

tage required to obtain a given output and efficiency,and (2) whether a reasonable efficiency [taken as25 °A] is obtainable without exceeding the maximumanode voltage recommended by the makers. . . .

It is surprising how many alleged power' valves. . . fail to pass this test."

TUBES WITH COLD CATHODES [Glow Discharge asElectron Source]. -A. Hund. (Electronics,Jan., 1933, pp. 6 and 22.)

Summary of a recent I.R.E. paper on the workreferred to in Abstracts, 1932, pp. 38 and 349 : cf.also p. 168, Guntherschulze and Keller. The onlyillustration shows one of Seibt's gas -dischargevalves (1932, p. 286), but Hund implies that hisown valves eliminate the " Seibt region " wherethe increase of current with voltage is very slowand uniform, and also the second region where thecurve is very steep, and work in the third region,giving a plate characteristic very similar to thatof an ordinary valve. Demonstrations were given.

A " Low -Hum " VACUUM TUBE [No. 262A High -Gain A.F. Amplifier for A.C. Mains]. -J. 0. McNally : Western Electric Company.(Bell Lab. Record, Feb., 1933, Vol. II, No.6, pp. 158-162.)

For public address systems, sound picture pro-jection, speech input equipment for broadcasting,etc. " Disturbance currents in the 262A tube dueto the alternating current supply of the heater aresomewhat greater than the shot effect and slightlyless than the thermal noise." See also 1932 Ab-stracts, pp. 639-640.

INTER -CARRIER NOISE SUPPRESSION : A NEWSYSTEM [using the Wunderlich B Valve withAdditional Anode for Amplifying the AVCPotentials].-Wunderlich. (See under " Re-ception.")

VACUUM TUBES [Survey for i932]. -L. J. Davies.(Electrician, 27th Jan., 1933, Vol. IIo, No.2852, p. fo3.)

THE PLACE OF NICKEL IN RADIO TUBE MAN U -FACTURE . -A. J. Marino. (Electronics, Jan.,1933, PP. 4-5.)

I.E.E. WIRELESS SECTION : CHAIRMAN'S ADDRESS.PART I : THERMIONIC VALVES. PART II :SPONTANEOUS FLUCTUATIONS IN VALVEAMPLIFIERS . -L. B. Turner. (Journ.Jan., 1933, Vol. 72, No. 433, pp. 10-30.)

In the section on improvements in receivingtriodes the writer compares the 1918 Army " R "valve with a modern directly heated type : " Ihave tabulated p.2/p because this, I think, measures,as well as any single quantity can, the practicalgain obtainable from an amplifying valve in termsof power or square of voltage. Despite a filamentpower divided by 13, in its operation the new valvemay be said to be so decibels better than the old."A further advance is given by the development ofindirectly heated cathodes (in England, oiat of everythree receiving valves sold only one is of the in-directly heated type: but on the other hand" while the sale of receiving valves is still increasing,


the sale of the directly heated type appears to havereached its maximum ").

The 500 kw demountable valve referred to byAngwin a year before (Abstracts, 1932, p. 112)has undergone alterations, and the new design isdescribed and illustrated, as is also a 4o kw valveof the same general type.

The penultimate section of Part 2 quotes someestimates and experimental determinations of thelimits of smallness of amplifiable signal voltage, andgives some results obtained with a sensitive wide -band amplifier (P.O. Research Department) forcross -talk and other telephonic measurements.The special low -noise valves of Metcalfe andDickinson (1932, p. 587) are referred to.

NON-LINEAR VALVE CHARACTERISTICS : A BRIEFDISCUSSION ON THEIR USE.-C. S. Bull.(Wireless Engineer, Feb., 1933, Vol. 1o, No.113, pp. 83-88.)

Author's summary :-In this article a method isgiven for showing how the frequencies in an inputsignal are added and subtracted by a curved valvecharacteristic. The resultant frequencies are not allable to give an output voltage across the anodeload, since the load is generally reactive or resistiveto a limited range of frequencies only. Simple rulesare given for determining the effective combinationof frequencies, and modulation rise, cross modula-tion, detection, modulation, and high -frequencymixing are considered in detail.

DIE BESTIMMUNG DER AUSTRITTSARBEIT AN OXYD-KATHODEN (Determination of the WorkFunction of [Barium] Oxide Cathodes).-W. Heinze. (Ann. der Physik, Jan., 1933,Series 5, Vol. 16, No. I, pp. 41-76.)

Author's summary :-Reasons are given forconsidering it more correct to determine the workfunction by means of the cooling caused by emis-sion rather than by using Richardson's equation.The errors of previous measurements by the coolingmethod are indicated. The experimental arrange-ment is improved and designed in such a way thaterrors, in particular those due to the internal re-sistance of the valve, are avoided. The heatingeffect of the emission current is either kept as smallas possible by suitable leads, or else calculated,using its relation with the heating current ; theeffect of re -radiation of heat from the anode iseliminated.

An empirical relation is found between the emissioncurrent in the saturation region and the anodevoltage ; in the case of oxide cathodes this takesthe place of the effect of the electrostatic imagerule and enables the work function found experi-mentally to be converted to that corresponding tozero external field. The work functions for [bariumoxide] cathodes made by the metallic vapour andpaste methods have mean values of 2.05 and 1.87volts respectively ; in this it appears probable thatthe work function increases with the temperature.

ETHER DIE RAUMLICHE VERTEILUNG DER EMISSIONVON GLOHKATHODEN [The Spatial Dis-tribution of the Emission from IncandescentCathodes).-F. Hamacher. (Archiv f. Elek-trot., 7th Jan., 1933, Vol. 27, No. 1, pp. 47-56.)

The distribution of emission from linear and non-

linear incandescent cathodes is investigated byobserving the fluorescent images they produce on ascreen which also serves as the anode. The effectof a Wehnelt cylinder is also considered, andmeasurements are made of the distribution behind aslit cylindrical anode, to explain the leakage roundstops raised to a high potential.

DIRECTIONAL WIRELESSLA RADIOGONIOMETRIE ET LES RADIOCOMPAS

STROBOSCOPIQUES (Direction Finding andStroboscopic Direction Finders).-R. Hardy.(Rev. Gin. de l'Elec., 21st Jan., 1933, Vol. 33,No. 3, pp. 85-94.)

Theory and practical development of the direct -reading direction finder dealt with in 1932 Ab-stracts, p. 642.

ACOUSTICS AND AUDIO -FREQUENCIESSUR LA PROPAGATION DES ONDES SONORES

D'AMPLITUDE FINIE (The Propagation ofSound Waves of Finite Amplitude [Mathe-matical Analysis with Results applicableto High -Power Horn Loud Speakers]).-Y. Rocard. (Comptes Rendus, 16th Jan.,1933, Vol. 196, No. 3, pp. 161-164.)

Equation 3 applies to a large exponential horn.If y = a sin wt represents the motion imposed on theair at the horn entrance, certain harmonics,super-posed on this, are found at the mouth : the ampli-tude of the pm harmonic is proportional to ay.If R is the ratio of the air velocities for the secondharmonic and the fundamental, at the horn mouth,the value of R is given by equation 4. If it isconsidered necessary, for the sake of quality, tolimit the value of R for a given frequency (sayR = I at 3 000 c/s), equation 4 gives a limit for a,which corresponds to a limit for the acoustic powerW supplied (per square centimetre) at the entranceto the horn. This limiting value is given by

W = R2pcsm2/8,2(1, + 1)2,12 ;

this, for a horn with a low cut-off at 7o c/s andfor R = I at 3 000 c/s, works out at W = o.82watt/cm2, corresponding to pressures of 25 000baryes at the entrance. " Such pressures areeasily obtained in powerful modern loud speakers,and this example (where the energy in free space ofthe second harmonic is equal to that of the funda-mental) shows that the consideration of waves offinite amplitude, and the equations governing them,may become very necessary in many cases."

ACCESSION TO INERTIA OF, AND POWER RADIATEDBY, A SPHERE VIBRATING IN VARIOUS WAYS ;WITH APPLICATIONS TO HORNLESS LOUDSPEAKERS [Theoretical Investigation].-N.W. McLachlan. (Phil. Mag., Feb., 1933,Supp. No., Series 7, Vol. 15, No. 98, pp.443-472)

The writer finds " that, under certain conditions,a radially vibrating sphere gives perfect repro-duction of acoustic pressures." He calculates theincrease of low -frequency output of discs and spheresdue to a large baffle. " When two diaphragmsvibrate in opposition along a common axis, it isindicated that a baffle is unessential provided theuseful radiation is confined to the external surfaces


of the vibrators." In the case of a sphere havingnodal circles, the low -frequency radiation is foundto decay rapidly with increase in the number ofnodal circles. " This arises from interference dueto oppositely vibrating segments. It follows thatthe 1.f. interference concomitant with free -edge discsor conical diaphragms executing diametral modesis extremely large, provided the system is sub-stantially symmetrical. By aid of formulae givenin the analysis, the mechanical impedance of varioustypes of spherical vibrator can be found."

SOUND DISTRIBUTION FROM A HORN [TheoreticalDetermination of Conditions for Validity of" Rigid Disc " Substitution for Mouth ofHorn].-N. W. McLachlan. (Wireless En-gineer, Jan., 1933, Vol. io, No. 112, pp.26-27.)

MESURE DE LA PHASE ET DE L'AMPLITUDE DE HAUT-PARLEURS ELECTRODYNAMIQUES (Measure-ment of the Phase and Amplitude of Moving -Coil Loud Speakers).-W. Binder. (Ann.des P.T.T., Jan., 1933, VOL 22, No. 1, pp.86-9E)

French version of the German paper referred toin 5932 Abstracts, p. 229.

LAUTSPRECHER-ANPASSUNG (Matching the LoudSpeaker with the Output Stage).-H.Keller. (Radio, B., F. f. Alle, Feb., 1933,PP. 67-69.)

The writer discusses the effects of under- andover -matching : the former (internal resistance ofvalve too large) causes a loss of the lower notes ; thelatter (internal resistance too high) a loss of thehigh notes. He concludes that in practice the bestcompromise between fidelity and amount of outputis obtained when " the internal resistance is about2 000 times the loud speaker inductance."

L' EVOLUTION ET L' ETAT ACTUEL DE LA TECHNIQUEMICROPHONIQUE (The Evolution and PresentCondition of Microphone Technique).-P.Chavasse. (Ann. des P.T.T., Jan., 1933,VOL 22, No. 1, pp. 9-58.)

VELOCITY [RIBBON] MICROPHONES: PRACTICALCONSTRUCTION OF Two TYPES THAT AREEASY TO MAKE.-C. W. Melotte : G. A.Elliot. (QST, Feb., 1933, Vol. 17, pp.23-25.)

EIN KOMPENSATIONSMIKROPHON SEHR KLEINERDIMENSIONIERUNG (A Compensation Micro-phone of Very Small Dimensions [for SoundField Measurements]) .-W. Geffcken.. (E.N.T.,Jan., 1933, Vol. bo; pp. 39-40.)

The " compensation " method of measuring soundintensities without distorting the sound field de-pends on the readings being taken when themeasuring microphone has its diaphragm at rest.But though this device prevents distortion by asecondary field, it does not get over the distortiondue to the obstacle presented to the sound wavesby the microphone ; it is therefore very desirablethat the dimensions of the latter should be verysmall. The compensation microphone here described

was made from a Siemens-Reiniger-Veifa " Ohr-sprecher " for the deaf, using the magnetic systemof this as the compensation device and adding asecond diaphragm to form a capacity microphone.The latter is connected in Riegger's " half resonancecurve " circuit, the output of which leads to anamplifier and telephones silence in the telephonesindicates that the diaphragm is at rest. The com-pensating voltage is led to the magnet systemthrough a Briicke " compensator " which allows thephase to be adjusted. The arrangement is very freefrom all kinds of external disturbance and is con-siderably more sensitive than a capacity microphoneof the usual size.

RETARDING RELAY FOR SYNCHRONISING DIRECTRECEPTION AND LOUD SPEAKER RECEPTIONIN PUBLIC ADDRESS SYSTEMS [using MagneticRepetition from Revolving Steel Ring].-L. Gaumont. (Summary of French Patentin Rev. Gen. de l'Elec., 28th Jan., 1933,Vol. 33, No. 4, p. 32 D.)

THE MODERN A.C. QUALITY AMPLIFIER [Self-contained Gramophone Equipment].-W. I. G. Page. (Wireless World, 17thFebruary, 1933, Vol. 32, pp. 128-131.)

A two -stage a.c. mains -driven amplifier for theamateur constructor. The speech output is sixwatts.

CONTROL OF SOUND QUALITY IN PICTURE PRO-DUCTION.-C. Dreher. (Electronics, Jan.,.1933, pp. 10-12 and 22.)

DAS LARMPROBLEM VOM STANDPUNKT DES INGEN-IEURS (The Noise Problem from the Stand-point of the Engineer.)-K. W. Wagner.(Zeitschr. V .D.I ., 7th Jan., 1933, Vol. 77,No. I, pp. 1-9.)

GESELLSCHAFT FUR ELEKTRISCHE MUSIK (TheSociety for Electrical Music).-(E.T.Z.,9th Feb., 1933, Vol. 54, No. 6, p. 139.)

Note on the recent formation of this Societyby Schiinemann, Director of the State High Schoolof Music, Berlin, and Leithauser, of the Heinrich -Hertz Institute.

KEYBOARD AND LOUD SPEAKER.-R. Raven -Hart.(Wireless World, 27th January, 1933, Vol. 32,p. 67.)

A short technical description of the " Neo-Bechstein " electrical piano.

SYNTHETIC SOUND.-Herbert Rosen. (WirelessWorld, 3rd February, 1933, Vol. 32, p. lox.)

The author describes a method developed by -Rudolph Pfenniger of producing " sound films "by making pencilled outlines-using the variablearea system-on paper strips.

GRID CURRENT COMPENSATION IN POWER AMPLI-FIERS.-W. Baggally. (Wireless Engineer,Feb., 1933, Vol. pp, No. 113, pp. 65-68.)

Taking a practical example of the design of a.power amplifier free from grid current distortion,the writer shows that for the conditions given the


input circuit resistance must not exceed 1000ohms and must deliver a total voltage excursion of35o volts without appreciable distortion. Withregard to these " very severe requirements " hesays : " It might be thought that we could reducethe impedance of the drive circuit by using astep-down transformer, but although this is donein the American ' class B ' amplifiers, using specialhigh -magnification output valves, it is not usuallyeasy to get sufficient undistorted voltage from thesecondary of the transformer to swing our Britishlow -impedance triodes. The arrangements now to bedescribed have the advantage of 'virtually reducingthe impedance of the penultimate stage to a lowvalue without in any way interfering with the voltage -handling capacity, though when used in this waythe stage does not amplify the applied voltage,but in fact introduces a small amount of attenuation.This is not usually of any consequence, as it merelyinvolves increasing the gain of the previous amplify-ing stages somewhat."

HARMONIC CONTENT IN AMPLIFIERS [IndiscriminateUse of Power Basis and Amplitude Basisin Specification : the Need for Uniformity].-H. L. Kirke. (Wireless Engineer, Jan.,1933, Vol. io, No. 112, p. 26.)

A NEW TYPE OF L.F. AMPLIFIER [Glow -DischargeTube Coupling : for Frequencies from o toTo 000 c/s].-Schaler : Peek. (See under" Reception.")

OUTPUT STAGES : THE CHOICE OF VALVES FORUSE IN LOW FREQUENCY AMPLIFIERS [RapidTest Method].-Whitehead : Brain. (Seeunder " Valves and Thermionics.")

ENDVERSTARKER FOR FERNSPRECHTEILNEHMER (Re-ceived Speech Amplifier for Telephone Sub-scribers [Single -Valve, Mains Operated :Five -fold Amplification]) .-Siemens & Halske.(E.T.Z., 2nd Feb., 1933, Vol. 54, No. 5, p.I I .)

THE AUDIO TRANSFORMER AS A SELECTIVE AMPLI-FIER.-M. Fawley. ( Journ. Franklin Inst.,Feb., 1933, Vol. 215, No. 2, pp. 133-147.)

Author's summary :-A selective amplifier isdeveloped which is simple, and more flexible thanother types. It utilises an audio transformer witha variable series resistance in the primary circuit,and a variable secondary load capacitance, bothof which may be adjusted to give broad or narrowpeaked amplification at any frequency within awide range, or flat amplification throughout thisrange. Theoretical and experimental amplificationcurves are shown, and a transient solution givesthe primary and secondary currents resulting whena voltage is suddenly applied to the amplifier.This analysis shows that the time constants for thecircuit are comparable with those for other peakedamplifiers. Oscillograms are shown illustrating thetransient behaviour of the amplifier.

LA QUALITE EN RADIOTELEPHONIE (Quality in[Broadcast] Radiotelephony).-P. David.(Rev. Gen. de l'Elec., 4th Feb., 1933, Vol.33, No. 5, PP. 139-145.)

The writer begins by discussing the far greater

complexity of the problem compared with that ofcommercial telephony (as dealt with by Chavassein the paper referred to in 1932 Abstracts, p. 44)thus in broadcasting the instantaneous power in-volved may vary in the ratio of I to io-7 for a meresymphony, while as regards " sensational " noisesthe American measurements are quoted (lion's roar87 db above audition threshold, aeroplane engineat 6 metres' distance io-I2o db, etc.). This (andthe consequent " compression " of light and shadein transmission) is only one of the complicatingfactors : others are the increase in frequency rangefrom si to 9 or 10 octaves ; the much more exactingdemands of the listener when bent on amusementthan when engaged in business ; fading, etc., etc.The practical worthlessness of the usual " intel-ligibility " tests in this sphere is pointed out : thefact that children, unaccustomed to listening, aremore exigent than adults, who become hardenedto certain faults, is quoted.

Thus to obtain quantitative results which canbe worked out and made to lead to further progress,the consideration of over-all quality must beabandoned in favour of separate analyses of thevarious causes of distortion, and in Section V thewriter deals with " non -uniform " distortion-i.e.,the unequal reproduction of different frequencies.Some results of Snow and of Fletcher (Abstracts,1932, pp. 98-99 : 1931, pp. 329 and 62o) are given,and embodied in a diagram showing the roles ofthe various portions of the spectrum (Fig. 2). Thefrequency characteristics of a m.c. loud speaker, acondenser microphone, and a ribbon microphone(taken from papers by Grutzmacher and Just andby Hartmann) are shown in Fig. 3, and the improvedreproduction given by combining several loudspeakers in Fig. 4. Section VI deals briefly withnon-linear distortion, and Fig. 7 gives some raspingfactor curves. Section VII deals with parasiticnoises.

FOURIER ANALYSIS AND VOWEL CURVES.-E. W.Scripture. (Nature, 24th Dec., 1932, Vol. 130,pp. 965-966.)

A critical comparison of the application of knownmethods of Fourier analysis to the resolution ofvowel curves. All are found unsatisfactory and therequirements of a new method, now under develop-ment, are stated.A NEW HARMONIC ANALYSER [Synchronous Disc

and M.C. Voltmeter].-B. G. Gates. (Journ.Scient. Instr., Dec., 1932, Vol. 9, pp. 38o-386.)

ON THE ACCURACY OF THE AURAL METHOD OFMEASURING NOISES [Unexpectedly HighAccuracy Found].-J. Obata and S. Morita.(Journ. A COUSt. Soc. Am., Oct., 1932, Vol. 4,No. 2, pp. 129-137.)

AKUSTISCHE MESSGERATE FOR PRAKTISCHENGEBRAUCH (Acoustic Measuring Instru-ments for Practical Use).-W. Jaekel :Siemens & Halske. (Zeitschr. V.D.I., 28thJan., 1933, Vol. 77, No. 4, pp. 98-99.)

AMPLITUDES OF SOUNDS FROM MUSICAL INSTRU-MENTS.-K. W. Wagner. (Electronics, Jan.,1933, p. 24.)

Summary only. A number of sound pressure


amplitudes are quoted (large choir, piano and organconcerts, soprano solo, etc.), and also the maximumpowers developed by various instruments.

THE VIBRATIONS OF A CHLADNI PLATE.-R. CColwell. (Phil. Mag., Feb., 1933, Series 7,Vol. 15, No. 97, pp. 317-324.) See alsoAbstracts 1931, p. 563 : January, 1933,p. 43 (two) : and back references.

DIAGONAL SYMMETRY IN CHLADNI PLATES.-R. C. Colwell. (Journ. Franklin Inst., Feb.,1933, Vol. 215, No. 2, pp. 169-177.)

A VACUUM TUBE OSCILLATOR FOR CHLADNI PLATES.-R. C. Colwell. (Science, 9th Dec., 1932,Vol. 76, No. 1980, pp. 547-548.)

A METHOD FOR THE DETERMINATION OF THEVELOCITY OF SOUND IN SOLIDS [based onBoyle and Rawlinson's Law and using aThermocouple to measure Intensity].-W. T. Richards. (Science, 8th July, 1932,Vol. 76, No. 1958, pp. 36-37.)

THE THEORY OF ACOUSTIC FILTRATION IN SOLIDRons.-R. B. Lindsay and F. E. White.(Journ. Acoust. Soc. Am., Oct., 1932, Vol. 4,No. 2, pp. 155-168.)

GRUNDLEGENDE UNTERSUCHUNGEN UBER SCHALL-ABSORPTION (Fundamental Investigationson the Absorption of Sound).-E. Winter-gerst and H. Klupp. (Zeitschr. V.D.I.,28th Jan., 1933, Vol. 7, No. 4, pp. 91-95.)

The sound absorption of wasinvestigated by placing a sample at the end of aio-cm wide tube, at whose other end a loud speakerprovided pure sinusoidal notes. The maximum andminimum pressure -amplitudes of the stationarywaves formed by reflection were measured bymeans of a microphone -probe (a long fine rigidtube attached to a microphone and adjustablealong the axis of the main tube) with amplifierand valve voltmeter. From these values theabsorption coefficient was calculated. The materialswere either freely hanging, stretched, or in theform of plates. The improvement of a material isdiscussed, as regards the serious defect of a co-efficient varying with frequency, by combining itwith another material, by partially perforatingit and thus using the resonance effect of the holes,and by other methods. Finally the " prices perunit of absorption " of the various materials arecompared.

TRANSMISSION OF SOUND THROUGH PARTITIONS.-A. H. Davis. (Phil. Mag., Feb., 1933,Series 7, Vol. 15, No. 97, pp. 309-316.)

COMBINED REVERBERATION TIME FOR ELECTRICALLYCOUPLED ROOMS.-A. P. Hill. (Journ.Acoust. Soc. Am., July, 1932, Vol. 4, No.Part 1, pp. 63-68.)

A MODIFIED FORMULA FOR REVERBERATION.-G. Millington. (Ibid., pp. 69-82.)

April, 1933

RESONANCE IN SMALL ROOMS.-V. 0. Knudsen.(Journ. Acoust. Soc. Am., July, 1932, Vol. 4,No. I, Part I, pp. 20-37.)

WEIGHT AS A DETERMINING FACTOR IN SOUNDTRANSMISSION.-P. E. Sabine. (Ibid.,PP. 38-43.)

ACOUSTICS OF BROADCASTING AND RECORDINGSninios.-G. T. Stanton and F. C. Schmid.(Ibid., pp. 44-55.)

ACOUSTICS OF LARGE AUDITORIUMS.-S. Lifshitz.(Journ. Acoust. Soc. Am., Oct., 1932,Vol. 4, No. 2, pp. 112-121.)

ACOUSTIC PICK-UP FOR PHILADELPHIA ORCHESTRABROADCASTS.-J. P. Maxfield. (Ibid., pp.122-128.)

RICHTWIRKUNG UND STRAHLUNGSLEISTUNG VONAKUSTISCHEN STRAHLERN UND STRAHLER-GRUPPEN IN DER NAME EINER REFLEK-TIERENDEN EBENEN FLACHE (Directive Effectand Radiated Power of Acoustic Radiatorsand Radiator Groups in the Neighbourhoodof a Reflecting Plane Surface).-F. A.Fischer. (E.N.T., Jan., 1933, Vol. io,pp. 19-24.)

A theoretical investigation. The writer dealsfirst with the case where the reflecting surface isinfinitely impervious to sound in comparison withthe propagating medium, so that there is no phase -change on reflection : a practical example of this isthe reflection of a sound wave in air at the surfaceof the earth or sea. He then considers the oppositecase, where the reflecting surface is, in comparison,completely permeable to sound, as when submarinesignals are reflected at the surface of the sea : inthis case a reversal of phase occurs.

In both cases he takes first the simplest exampleof a single radiator, small in comparison with thewavelength, at a distance h from the reflectingsurface. By combining the radiator and its imageinto a two -radiator group of spacing 2h and commonphase (for the first case) or opposed phase (for thesecond) and by using the appropriate formula fora two -radiator group (formula I or 7) he obtainsthe corresponding polar diagrams (e.g., Figs. 2and 7d, both for h = 3A). In the case of reflectionwithout phase -change, he points out that whateverthe relation between h and A there is always aradiation maximum coinciding with the surfaceitself, while with reflection with phase -reversalthere is always a radiation minimum along thesurface.

In the latter case there are no further minimaso long as 2h/A LS 1/2. Fig 7 gives the diagrams forvarious values of 2h/A. In curve a the ratio ismuch less than Or and the characteristic is thewell-known figure -of -eight curve of the wirelessframe aerial, of which, here, only one half has aphysical significance. As h/A increases, the minimumalong the surface becomes more and more sharp,till for h = A/2 a second minimum appears (curve c).

The calculation of the radiated power, in bothcases, can be accomplished by integration or byRayleigh's formula (equation 2 or 9). The curve

April, 1933

of the expression in the brackets is given in Fig. 3for unchanged -phase reflection and in Fig. 8 forreversed -phase reflection. The directive charac-teristic of a group of radiators is given by theproduct of the characteristic of the group in aninfinitely extended medium and the characteristicof a two -radiator group spaced by a distance 2h,of common phase in one case and of opposed phasein the other. Thus Fig. 9 gives the diagram of aquadruple group at a depth in the sea of 3k : it isthe product of 5a, the characteristic of such agroup in an infinitely extended medium, and 7d,the diagram for a simple radiator at a depth of 3k.Similarly, for reflection without phase -change,Fig. 5b shows the combination of the same group -characteristic 5a and the simple radiator diagramFig. 2. The radiated power from such groups ofradiators can be found from the appropriate Rayleighformula (equation 6 or II, according to the natureof the reflection).

The paper ends by a discussion of the morecomplex case where a second reflecting layer hasto be considered, as in submarine signalling wherethe sea bottom has its effect.

POSITION FINDING BY UNDER -WATER SQUNDSIGNALS.-B. R. Hubbard. (Journ. Acoust.Soc. Am., Oct., 1932, Vol. 4, No. 2, pp. 138-'54.)

PROCtDE OPTIQUE POUR LA MESURE DE L'ABSORP-TION DES ONDES ULTRA-SONORES PAR LESLIQUIDES (Optical Method of Measuring theAbsorption of Supersonic Waves by Liquids[based on the Diffraction of Light due tothe Supersonic Waves]).-P. Biquard.(Comptes Rendus, 23rd Jan., 1933, Vol. 196,No. 4, pp. 257-259.)

See March Abstracts, p. 15o (Lucas), for the workleading up to this method,

QUASI -STANDING WAVES IN A DISPERSIVE GAS.-D. G. Bourgin. (Ibid., pp. 108-I l 1.)

PREPARATION OF COLLOIDS BY SUPERSONIC DIS-PERSION.-N. Marinesco. (Comptes Rendus,3oth Jan., 1933, Vol. 196, No. 5, pp. 346-348.)

PHOTOTELEGRAPHY AND TELEVISIONAN INVESTIGATION OF VARIOUS ELECTRODE STRUC-

TURES OF CATHODE-RAY TUBES SUITABLEFOR TELEVISION RECEPTION [Size of Spotindependent of Intensity Modulation, etc.].-A. B. Du Mont. (Proc. Inst. Rad. Eng.,Dec., 1932, Vol. 20, pp. 1863-1877.)

Leading to the development of a tube with dualaccelerating electrodes (Fig. 13), the first a discwith a hole in it, and the second a hollow cylinder.To increase the spot intensity without loss of sensi-tivity, an additional anode of silver is deposited onthe funnel -shaped part of the bulb. The vacuumis of the order of one millionth of an atmosphere.A signal of several volts will modulate the spotfrom minimum to maximum brightness. Verysimilar results were obtained with only one accele-rating electrode (with a very fine " gun "-Fig. 12).


INTENSITATSSTEUERUNG GASKONZENTRIERTER ELEK-TRONENSTRAHL MITTELS WEHNELT-ZYLINDER(Intensity Control of the Gas -ConcentratedCathode Ray by Wehnelt Cylinder).-W.Kleen. (Hochf:tech. u. Elek:akus., Jan.,1933, Vol. 41, pp. 28-3o.)

A short paper on the work dealt with in FebruaryAbstracts, pp. 105-106.

CATHODE-RAY OSCILLOGRAPHY [Physical SocietyDiscourse].-Watson Watt. (See under" Subsidiary Apparatus and Materials.")

MODULATION FREQUENCIES IN VISUAL TRANS-PassION.-E. L. White. (Proc. Inst. Rad.Eng., Jan., 1933, Vol. 21, pp. 51-55.)

Author's summary :-A method of computingthe maximum frequencies produced in televisiontransmission is shown. This method is based onthe consideration of the degree of edge definitionproduced. It is shown that these frequencies areindependent of the amount of detail in transmittedpictures for equal edge definition.

For a picture scanned by the usual method, Npictures per second, with L lines, having a ratio rof width to height and having the ratio K betweenthe width of scanning line and width of edge shadow,it is shown that f = KNrLI/2. Two systemsvarying from the normal [decreased number oflines away from centre of picture] are discussedand compared with the normal system in regardto the magnitude of the frequencies produced.

MARCONI TELEVISION TRANSMITTER TYPE TT5.-R. J. Kemp. (Marconi Review, Nov.-Dec.,1932, No. 39, pp. 7-18.)

Using 50 -scan lines per picture and a picturerepetition frequency of 15/sec. This number ofscan lines was chosen because it gives detail suffi-ciently good to enable recognisable head andshoulders, and even full-length images, to bereceived. Any increase in the number would leadto an appreciable increase in the bulk of the appa-ratus and the necessity for a radio transmitter ofspecial design, whereas the apparatus is intendedfor use with the existing type of high -qualitybroadcasting transmitter.

TELEVISION TESTS ON ULTRA -SHORT WAVES FROMEMPIRE STATE BUILDING.-(See abstractunder " Miscellaneous "-" Looking Aheadas 1933 Opens.")

TELEVISION [Survey for I932].-W. G. W. Mitchell.(Electrician, 27th January, 1933, Vol. I Da,No. 2852, p. 115.)

DBER DIE ABHANGIGKEIT DER ELEKTROMOTORISCHENKRAFT VON KUPFEROXYDULPHOTOZELLEN(SPERRSCHICHTPHOTOZELLEN) VON DERBELEUCHTUNGSSTARKE UND WELLENLANGE(The Dependence of the E.M.F. of CopperOxide [Barrier Layer] Photocells on theIntensity of Illumination and Wavelength).-A. Goldmann and M. Lukasiewitsch.(Physik. Zeitschr., 15th Jan., 1933, Vol. 34,No. 2, pp. 66-73.)

The writers find that their measurements canall be comprised by the formula

V = M .LI(L + MN), (V = e.m.f., L = in-


tensity of illumination, M and N constants).Various bases for this formula are discussed andit is found that it can be deduced if the resistanceof the barrier layer is a linear function of the cellvoltage. Measurements were made with mono-chromatic light in various spectral regions and theboundary potential (for L-). co) appears to beindependent of the wavelength to a first approxima-tion.

'OBER DIE ABHANGIGKEIT DER ELEKTROMOTORIS-CHEN KRAFT DER SELENSPERRSCHICHTPHOTO -ZELLEN VON DER BELEUCHTUNGSINTENSITAT(The Dependence of the E.M.F. of SeleniumBarrier -Layer Photocells on the Intensityof Illumination).-A. Goldmann. (Physik.Zeitschr., 15th Jan., 1933, Vol. 34, No. 2,PP. 74-75.)

The writer finds that the formula for the e.m.f.of a barrier -layer photocell given in the pre-ceding abstract also holds for selenium cells.

ZUR FRAGE DER EXISTENZ EINES SPERRSCHICHT-PHOTOEFFEKTS AM BLEISULFID (The Questionof the Existence of a Barrier -Layer Photo-electric Effect in Lead Sulphide).-E. Tiedeand G. Briickmann. (Zeitschr. f. Physik,1933, Vol. 8o, No. 5/6, pp. 302-304.)

In this preliminary communication the writersstate that they find no " back surface " [Hinter -wand] photo -electric effect at a nickel/lead sul-phide boundary and that the " front surface "[Vorderwand] photoelectric effect is really athermoelectric effect.

REVERSAL OF THE CURRENT FROM A CUPROUSOXIDE PHOTO CELL IN REP. LIGHT.-H. H.Poole and W. R. G. Atkins. (Nature,28th Jan., 1933, Vol. 131, p. 133.)

Details are noted of a small reverse currentobtained with a cell of the " Serpidox " makeexposed to red light.

SELF -GENERATING PHOTOCELL [Dry -Disc Cell giving5-7 Milliamperes in Direct Sunlight].-T. Rhamstine. (Electronics, Jan., 1933,p. 25.)

SELENIUM CELL EMPLOYING NEW PRINCIPLE[Layer of Selenium fused on Iron Disc,with Transparent Film of Sputtered PreciousMetal].-Siiddeutsche Apparatefabrik. (Ibid.,p. 26.)

For a so cm2 surface the current ranges from 5.2to 445.5 microamperes, for a 25 -watt lamp and a50o -watt projector lamp respectively, at a distanceof one metre.

MODERN METHODS OF MANUFACTURING PHOTO-ELECTRIC CELLS. PART II.-G. Dejardin.(Rev. Get, de l' Elec., 14th Jan., 1933, Vol. 33,No. 2, pp. 36-53.)

Second and final part of the survey dealt within March Abstracts, p. 169. Throughout thepaper, footnotes give the appropriate journal andpatent references.

VBER DAS AUFTRETEN UND DIE DEUTUNG DERSELEKTIVEN LICHTELEKTRISCHEN ELEK-TRONENEMISSION AN ZUSAMMENGESETZTENALKALIKATHODEN (The Occurrence andExplanation of the Selective Emission ofPhotoelectrons at Composite AlkaliCathodes).-W. Kluge. (Physik. Zeitschr.,ist Feb., 1933, Vol. 34, No. 3, pp. 115-126.)

For composite layers of the type ag-m2o-m,where M is an alkali metal, the writer finds twoemission maxima, one at the high -frequency andone at the low -frequency end of the spectrum.The former are tentatively explained on linesalready indicated by Ives (Ives and Briggs, 1932Abstracts, p. 102) and the latter are satisfactorilyexplained by the theory of Gudden and Pohlpublished in 1925.

SUR LE CALCUL D'UN AMPLIFICATEUR DE BASSEFREQUENCE POUR CELLULE PHOTO-ELECTRIQUE (The Calculation of a L.F.Amplifier for a Photoelectric Cell).-G. A.Boutry. (Comptes Rendus, 27th Dec., 1932,Vol. 195, No. 26, pp. 1384-1387.)

For a densitometer working on constant deflec-tions. Since the cell impedance R, under theconditions of use, is very considerable (of the orderof 6 000 ohms), a transformer of impedance suitedto such a circuit is not practical : the classicalcoupling by resistance and condenser was thereforeinvestigated. If the cell delivers its output to aresistance r connected to the grid through a largecapacity C, and the total ohmic resistance of thegrid circuit (leakage, internal resistance, etc.) isa, while p and R, represent the internal resistanceof the anode circuit and the impedance of thecoupling to the next stage, respectively, then it isfound that the employment of screen -grid valvesis necessary and that the optimum energy amplifica-tion is obtained when p = R2 and when ar = R2.In practice it is impossible to exceed r = a =loo ohms without compromising the stability of thestage.

The above theoretical results have been confirmedexperimentally. The current amplifications of thestage, given approximately by the equation

= k .11(p+ R2).ral(o r), where k is the amplifica-tion coefficient of the valve, agrees within io %with the experimental value. It is easy to obtainvalues of p. between io 000 and 20 000: toexceed this, " a very well insulated and protectedmaterial must be available. If the stage is followedby a triode detecting by grid characteristic curvature,the intensity of a photoelectric current of io-2.acan be multiplied by 90 000."

PHOTOELECTRIC CURRENTS IN GASES BETWEENPARALLEL PLATE ELECTRODES [TheoreticalInvestigation].-L. A. Young and N. E.Bradbury. (Phys. Review, ist Jan., 1933,Series 2, Vol. 43, No. I, pp. 34-37.)

The theory developed in this paper is based onthe fact that the main factor governing the currentobserved is the Ramsauer scattering by gas atoms.An equation is found for the ratio of current to itssaturation value as a function of the field strengthand other constants of the gas. The ratio shouldvary as the square root of the field strength for small

April, 1933

field strengths. This agrees with experimentalresults already obtained by one of the writers(Bradbury, 1932 Abstracts, p. 532).

THE PASSAGE OF PHOTOELECTRONS THROUGHMICA.-N. E. Bradbury and L. A. Young.(Phys. Review, 1st Jan., 1933, Series 2,Vol. 43, No. 1, pp. 84-85.)

The theory referred to in the foregoing abstractmay be extended to apply to solid dielectrics.This letter describes experiments with mica whichgive agreement with the theory.

THE PHOTOELECTRIC EFFECT OF ELECTRIC SPARKRADIATION.-J. Meiklejohn. (Phil. Mag.,Jan., 1933, Series 7, Vol. is, No. g6, pp.146-163.)

From the author's summary :-The resultsobtained indicate that the intensity of the photo-electric radiation and the quantity of electricitydischarged from single sparks were both quadraticfunctions of the minimum sparking potential.. . .

The effect of the electrode substance tends to confirmthe photoelectric theory of sparking potentials . . .

the intensity of the photoelectric radiation wasapproximately a linear function of the quantityof electricity discharged and also of the energy ofthe spark. . . . Experiments on the effect of oxideon the sparking potential showed that surfaceconditions play a part in determining the sparkingpotential.

MEASUREMENTS AND STANDARDSDAS PARALLELDRAHTSYSTEM ALS MESSINSTRUMENT

IN DER KURZWELLENTECHNIK (The ParallelWire [Lecher Wire] System as a MeasuringInstrument in Short -Wave Technique).-O. Schmidt. (Hochf:tech. u. Elekokus.,Jan., 1933, Vol. 41, pp. 2-16.)

The writer's primary object is to point out themany applications of the " line without internallosses " as a measuring instrument in short-wavetechnique, where other methods fail: for themeasurement of resistances, characteristic im-pedances, currents and voltages (magnitude andphase), power and damping. In the course of hispaper he investigates the errors liable to arise frompractical deviations from the ideal conditions, suchas the presence of internal damping ; effects dueto the spacing not being infinitely small comparedwith the wavelength ; lack of homogeneity in thewires (e.g., the presence of insulator caps, bends,divergence of the wires, etc.) ; change of dielectricconstant by fog. Thus in a long Lecher pair (lengthfrom measuring point to end 93.5 m) an averagedisplacement of nodal point amounting to about30 cm was found as a result of autumnal fog. Rainhad no effect.

In the first (mathematical) part of his paper thewriter shows that a parallel wire system can beregarded as a non-quasistationary oscillatingcircuit, and brings out the analogies between sucha circuit and a quasistationary circuit with con-centrated capacity and inductance. He shows alsothat the geometrical locus of the impedance of aline without internal loss, for a constant coupleddamping, and the corresponding locus of the im-pedance for a constant distance between potential


nodes, form orthogonal circles to each other (Fig. 6).For the impedance in the case of a finite internaldamping he obtains an approximate formula (26),in which the internal damping of the line is expressedas a concentrated load and the line is consideredto be otherwise free from losses.

The final part of the paper deals with the variousapplications mentioned at the beginning of thisabstract.

MEASUREMENT OF THE FREQUENCY OF ULTRA -RADIO[Ultra -Short] WAVES [using Special LecherWire System].-J. B. Hoag. (Proc. Inst.Rad. Eng., Jan., 1933, Vol. 21, pp. 29-36.)

Author's summary :-" An analysis of a par-ticular Lecher wire system, whose characteristicimpedance matches its input impedance and whoseoutput end is short circuited, has led to an equationpermitting the measurement of ultra -radio fre-quencies to three significant figures. The methodis independent of end effects and has been appliedto a determination of the velocity of propagationalong iron wires."

In the course of the paper the writer points outthat the frequency and strength of the source mustremain constant during the time needed to obtaina complete set of data ; this time, however, maybe shortened without undue sacrifice of accuracyby limiting the data to two peaks at oppositeends of the Lecher wires. With a complete setof data the accuracy is fully ten times that of theusual peak -to -peak method for these frequencies,the calculated wavelengths being reliable to threesignificant figures.

MESSUNGEN VON WELLENWIDERSTAND UNDSCHEINWIDERSTAND VON SYMMETRISCHENUND UNSYMMETRISCHEN FREILEITUNGEN BEIHOHEN FREQUENZEN (Measurements of theCharacteristic Impedance and ApparentImpedance of Symmetrical and Unsym-metrical Overhead Lines at High Fre-quencies [in Connection with the Propagationof Interference with Broadcast Reception].-A. Clausing. (E.T.Z., 19th Jan., 1933,Vol. 54, No. 3, pp. 54-56.)

Undertaken by Siemens and Halske at the requestof the VDE. Theory demands that with increasingfrequency a definite limiting value of characteristicimpedance is reached asymptotically. The measure-ments have produced no evidence to the contrary,but could not be carried out with sufficient accuracy(for reasons which are discussed) to prove the theorycompletely.

A NOTE ON AN AUTOMATIC FIELD STRENGTH ANDSTATIC RECORDER.-W. W. Mutch. (Proc.Inst. Rad. Eng., Dec., 1932, Vol. 20, pp. 1914-1919.)

Sample records of short-wave signal intensitiesmade by this equipment were shown in the paperby Potter and Friis dealt with in 1932 Abstracts,p. 344. The primary aim of the equipment was torecord the energy received from a fading signalduring periods of the order of ten seconds. Themethod adopted is that the gain of the measuringset is changed at short intervals in such a way asto hold the average output constant, a commercialrecorder controller being used to operate the gain


control and to record the changes in gain withtime. This controller is an instrument containinga sensitive moving coil galvanometer actuated (inthis case) from the output circuit of the measuringset : a set of jaws closes every two seconds acrossthe arc of swing of the needle, and the position ofthe needle when thus clamped controls the motionof the attenuator potentiometer.

PORTABLE LONG WAVE TESTING APPARATUS [Field -Strength Measuring Set]. -G. C. de Coutouly.(Bell Lab. Record, Feb., 1933, Vol. II, No. 6,pp. 178-183.)

A NEW WAY OF MAKING USE OF THERMOELECTRICPHENOMENA.-Egal. (See under " Sub-sidiary Apparatus and Materials.")

THE BROWNIAN MOTION OF THE RESONANCERADIOMETER. -G. A. Van Lear, Jr. (ReviewScient. Instr., Jan., 1933, Vol. 4, pp. 21-27.)

ON A SOLUTION OF THE MOTION OF A GALVANO-METER EXCITED BY ANY PERIODIC CURRENT.-R. Yoneda. (Journ. I.E.E. Japan, Sept.,1932, Vol. 52 [No. 9], No. 53o : Englishsummary pp. 99-10o.)

THE LIGHT RAY AS POINTER FOR TECHNICALMEASURING INSTRUMENTS. -K. H. Sieker.(Zeitschr. f. Fernmeldetech., 15th Dec., 1932,Vol. 13, pp. 186-19o.)

A SENSITIVITY CONTROL FOR THE LINDEMANNELECTROMETER. -L. G. Grimmett. (Proc.Physical Soc., ist Jan., 1933, Vol. 45, Part1, No. 246, pp. 117-119.)

I.E.E. METER AND INSTRUMENT SECTION : CHAIR-MAN'S ADDRESS. -R. S. J. Spilsbury. (Journ.I.E.E., Jan., 1933, Vol. 72, No. 433, pp.3o-36.)

ALTERNATING -CURRENT BRIDGES FOR MEASURE-MENTS OF ELECTRICAL INSULATING MA-TERIALS [at Frequencies from 60 c/s to1.5 Mc/s).-W. A. Ford and S. I. Reynolds.(Gen. Elec. Review, Feb., 1933, Vol. 36, pp.99-105.)

MESSUNG KLEINER ELEKTRO-DYNAMISCHER KRAFTEMIT DEM KONDENSATORMIKROPHON (TheMeasurement of Small ElectrodynamicForces with the Condenser Microphone). -A. Agricola. (Hochf:tech. u. Elek:akus.,Jan., 1933, Vol. 41, pp. 30-33.)

Author's summary :-The calibration of a con-denser microphone for measuring small pressuresis described. The force is applied by the measur-able homogeneous magnetic field in a plate con-ductor, in which there is an opening into which themicrophone is introduced. The measured forceslie between the limits of 0.5 and 120 mg on the24 cm2 surface.

A TRIPLE INTERFEROMETER FOR DISTINGUISHINGFLEXURAL AND LONGITUDINAL VIBRATIONSIN QUARTZ. -H. Osterberg. (Journ. Opt.Soc. Am., Jan., 1933, Vol. 23, pp. 30-34.)

April, 1933

SILVERING TO LOWER CRYSTAL FREQUENCY [Con-trolled Deposition of Silver Electrodes tocure Quartz Crystals ground to Too High aFrequency]. -W. P. Hunter. (QST, Feb.,1933, Vol. 17, pp. 48-49.)

LANGS- UND BIEGUNGSSCHWINGUNGEN VON TUR-MALINPLATTEN (Longitudinal and FlexuralVibrations of Tourmalin Plates). -V.Petriilka. (Ann. der Phys., 1932, Series5, Vol. 15, No. 8, pp. 881-902.)

-OBER NORMALIEN FUR WELLENMESSUNG DERULTRAKURZEN WELLEN (On Standards forMeasurement of Ultra -Short Waves). -G.Leithauser and V. Petriilka. (Funktech-nische Monatshefte 9/1932.)

The writers recommend the use of tourmalininstead of quartz for resonators in the ultra -shortwave -band, as the optical and electrical axes oftourmalin coincide and circular plates can be used.A resonator using neon gas containing a littlehydrogen is described, but the direct use of this isnot advised, as the resonance point is not exactlydefined ; the parallel incandescent lamp method(Feb. Abstracts, p. 109), however, is accurate.Even harmonics of the tourmalin plate can also beused.

AN OSCILLATOR HAVING A LINEAR OPERATINGCHARACTERISTIC [and High FrequencyStability].-Arguimbau. (See under " Trans-mission.")

THE WIRELESS WORLD DIRECT READING STATIONFINDER. -H. F. Smith. (Wireless World,13th January, 5933, Vol. 32, pp. 20-22.)

An absorption wavemeter calibrated directlyin " station names," for the amateur constructor.

TIME -KEEPING --OLD AND NEW. -F. Hope -Jones.(Journ. Scient. Instr., Feb., 1933, Vol. io,PP 43-49.)

THE CLINKER ELECTRIC CLOCK FOR MAINS WHICHARE NOT TIME -CONTROLLED. -British Thom-son -Houston Company. (Ibid., p. 54.)

An LC combination has parallel resonance at theworking frequency when the pendulum, carrying amagnetic member, is remote from the magneticcircuit of the iron -cored inductance : the pendulumis thus always attracted to the central or verticalposition by a force greater than that which retardsit after this position is passed.

A NEW TYPE OF FREE -PENDULUM CLOCK [FreePendulum Self -Maintained in vacuo bymeans of Electrostatic Impulses and havinga Closely Governed Arc]. -G. A. Tomlinson.(Proc. Physical Soc., I $t Jan., 1933, Vol. 45,Part I, No. 246, pp 41-48.)

AUTOMATIC EQUIPMENT FOR THE TRANSMISSION OFTHE EXACT TIME OVER THE TELEPHONESYSTEM. -(Alta Frequenza, Sept., 1932, Vol.1, No. 3, PP. 474-477.)

SYNCHRONISATION OF CLOCKS OVER TELEPHONENETWORK.-Brillie Freres. (Summary ofFrench Patent in Rev. Gen. de l'Elec., 28thJan., 1933, Vol. 33, No. 4, p. 29 D.)

D2


SUBSIDIARY APPARATUS AND MATERIALSCATHODE-RAY OSCILLOGRAPHY [Physical Society

Discourse].-R. A. Watson Watt. (Journ.Scient. Instr., Feb., 1933, Vol. ro, PP. 37-43.)

A summary of this discourse was referred to inMarch Abstracts, p. 172. Dealing, in the latterhalf, with the applications of the c. -r. oscillograph,the writer takes in turn its various " disposablevariables " and shows how use can be made ofthem. The " double linotime " scheme, for over-coming the persistence of vision by spreading thebases into a venetian blind pattern, is describedand demonstrated. The process has been appliedto visual counting at rates of a few thousand persecond. " A similar combination of timing voltagesat higher speeds gives the image -scanning structurerequired in television processes depending on c. -r.oscillographs."

The disposable variable, or group of relatedvariables, provided by the differing propertiesof various fluorescent materials (colour of response,duration of afterglow, etc.) is discussed and demon-strated.

Finally the use of the oscillograph as a relay, theelectron beam being trapped in a Faraday cagethe incidence or release of whose charge can beused to control a train of events, is illustrated by aself -releasing time -base (after Rogowski) and bythe lecturer's directional radio -receiver respondingonly to signals from a selected direction.

APPLICATIONS OF THE CATHODE-RAY OSCILLO-GRAPH [Factors affecting Accuracy of Measure-ments, and Their Compensation : Oscillo-graph with Magnetic Deflection).-R. R.Batcher. (Proc. Inst. Rad. Eng., Dec.,1932, Vol. 2o, pp. 1878-1891.)

The writer concludes that the threshold effectmust be considered when small amplitudes arebeing measured : deflection plate leakage isimportant when high -impedance, low -powercircuits are being measured : the effect of diagonalpath can usually .be neglected : it is of the sameorder as the screen curvature error and is oppositeto this : unbalanced magnetic deflection fieldsshow up on calibration curves and can be correctedby physical changes in the coil assembly : strayfields should be eliminated by external magneticshielding, and the effect of filament field distortioncan be eliminated by the use of a d.c. filamentsupply if the field cannot be balanced out.

In gas -concentrated tubes " tests have indicatedthat somewhat better focusing at high frequenciesis obtained if the large end of the tube is heated,although definite conclusive tests on this effecthave not been completed. Incomplete experimentsalso seem to show a deflection amplitude changeat high frequencies that cannot be explainedentirely by the effect of the finite impedance ofthe deflection plates, possibly also due to changesin the velocity of the stream with rapid lateralmovements due to the varied effects of the spacecharge."

A tube with magnetic deflection coils only, havinga sensitivity of about 28 ampere turns per inchwith an anode voltage of Boo, is discussed andillustrated. The problem of the calibration factorfor magnetic deflections is analysed.

INTENSITY CONTROL OF THE GAS -CONCENTRATEDCATHODE RAY BY WEHNELT CYLINDER.-Kleen. (See under " Phototelegraphy andTelevision.")

tJBER DIE GRUNDLAGEN DER GEOMETRISCHENELEKTRONENOPTIK (The Foundations ofGeometrical Electron Optics).-E. BrUche.(Zeitschr. f. tech. Phys., No. 2, Vol. 14,1933, PP. 49-58.)

BEITRAGE ZUR THEORIE DER GEOMETRISCHENELEKTRONENOPTIK (Contributions to theTheory of Geometrical Electron Optics).-J. Picht. (Ann. der Physik., 1932, Series 5,Vol. 15, No. 8, pp. 926-964.)

This theoretical paper considers the differentialequations of electron beams for regions near theaxis. It shows how to calculate the potentialdistribution corresponding to a given focal lengthof the electrical lens. Formulae are found forcalculating the positions of the focal points of agiven potential distribution.

VBER DIE ELEKTRISCHE ELEKTRONENSAMMELLINSE(The Electric Electron Condensing Lens).-H. Johannson and 0. Scherzer. (Zeitschr.f. Physik, 1933, Vol. 8o, No. 3/4, pp. 183-192.)

ELECTRON LEN SES .-C. J. Davisson and C. J.Calbick. (Phys. Review, 15th Nov., 1932,Series 2, Vol. 42, No. 4, p. 580.)

A letter correcting an error in a formula givenby the writers in 1931.

ZUR THEORIE DER ELEKTRONENOPTISCHEN LINSEN-FEHLER (The Theory of the Lens Errors inElectron Optics).-O. Scherzer. (Zeitschr.f. Physik, 1933, Vol. 8o, No. 3/4, pp. 193-202.)

THE CATHAUTOGRAPH : AN ELECTRONIC PENCIL[using a Fluorescent Screen with a DecayPeriod of about 3o Seconds].-A. B. Du Mont.(Electronics, Jan., 1933, p. 7 and frontcover.)

Suggested applications include communicationbetween aircraft and ground stations and betweensmall ships and land stations ; office inter-communication ; police cars ; noiseless instructionsto broadcasting artists ; advertising purposes ;and for Chinese and Japanese services.

STUDY OF THE DISCHARGE IN AN IONIC TUBE[working on the Low Voltage of r 000 vwith High Intensity] with the help of aCinematographic Apparatus.-V. Dolejlekand K. Drab. (Comptes Rendus, 3oth Jan.,1933, Vol. 196, No. 5, pp. 334-336.)

EIN EINFACHES VERFAHREN ZUR BEOBACHTUNGHOCHFREQUENTER SPANNUNGSKURVEN (ASimple Method of Observing High -FrequencyVoltage Wave Forms [Cathode -Ray Oscillo-graph and Rotating Mirror]).-U. Bab.(Zeitschr. f. tech. Phys., No. r, Vol. 14,1933, pp. 18-19.)

The method consists in applying to the two pairs


of deflecting plates the high frequency underinvestigation and a sinusoidal high frequencydiffering from the first by an audible frequency,and using a rotating mirror to view the resultingfigure. The conditions for a faithful reproductionof the wave form are shown to be a = ?thA and

= A(fi- f2)12a, where a is the distance ofmirror from screen, k the number of surfaces onthe mirror, n the rotations per second, A the half -breadth of the image on the screen, and fi - f2the audio -frequency ; it is found, however, thatthese conditions need only be obtained approxi-mately, by adjustment in operation, so that themethod is very practical.

A SPECTROSCOPIC METHOD OF DETERMINING THEEXCITATION OF LUMINESCENCE [for theRapid Choice of the Ultra -Violet Wavelengthsuitable for Various Fluorescent Substances].-G. Heyne and M. Pirani. (Zeitschr. f.tech. Phys., No. 1, Vol. 14,1933, pp. 31-33.)

A MERCURY SEALED GAS VALVE.-S. F. Essig.(Review Scient. Instr., Dec., 1932, Vol. 3,pp. 762-763.)

A FILTER FOR AN X-RAY POWER SUPPLY [giving30 IBA at 20 kv with a Ripple of less than0.02% : Greinacher Rectifier Circuit andBalanced Pi -Type Low -Pass Filter].-J. B.Hoag and V. J. Andrew. (Review Scient.Instr., Dec., 1932, Vol. 3, pp. 750-752.)

X-RAY TUBE INPUT CONTROL.-W. R. Harper.(Journ. Scient. Instr., Jan., 1933, Vol. Jo,pp. 10-12.)

It is shown that the fluctuations of a.c. mainscannot be counteracted by a control confined to theprimary side of the h.t. transformer, owing tochanges in wave form introduced by the controlitself. The correct methods of control are described.

EIN ZEITTRANSFORMATOR ZUR AUTOMATISCHENREGISTRIERUNG KURZER ZEITEN (A TimeTransformer for the Automatic Registrationof Short Times).-M. Steenbeck and R.Strigel. (Archiv f. Elektrot., 9th Dec., 1932,Vol. 26, No. 12, pp. 831-84o.)

Authors' summary :-A time transformer hasbeen designed for the measurement of short times.The principle on which it is based is that a con-denser is charged up by a constant current during theperiod of time to be measured and then dischargedby a small, constant current. Thus the dischargeperiod is proportional to the period of charging.Measurement of the discharge time enables timesto be measured which are too short to be measureddirectly. Two arrangements are described, ofwhich one simply measures the mean value of alarge number of single times, while the otherregisters the magnitude of each single time. Twodesigns are developed, one measuring times downto io-4 sec. and the other times from 10-4 sec.upwards. In the arrangements described, chargeand discharge of the condenser each take pia&via a valve. It would also be possible to arrangeohmic resistances in the two circuits, to limit thecurrents, i.e., in general to use homologous circuitsfor charging and discharging the currents.

KLEI NZEITFORSCHUNG IN DER TECHNIK (TheAnalysis of Processes occurring in VeryShort Times [by High -Speed Cinematographyup to 8o 000 Pictures per Second]).-W.Ende. (Zeitschr. V.D.I., 7th Jan., 1933,Vol. 77, No. 1, pp. 10-12.) See alsoAbstracts, 1932, p. 598 (von Ohnesorge :Thun) and February, p. 112 (Ende).

AN OSCILLOGRAPH FOR TEN THOUSAND CYCLES.-A. M. Curtis. (Bell S. Tech. Journ., Jan.,1933, VOL 12, No. I, pp. 76-90.)

This paper describes a method of attack of theproblem of extension of frequency range of oscillo-graphs, not by increasing the natural frequency ofthe vibrating element, but by employing a vibratorwith only a moderately high natural frequency and" equalising the response of the string by electricalcircuits both up to and beyond the fundamentalresonance frequency."

MESSUNGEN AN SPERRSCHICHTGLEICHRICHTERN(Measurements on Barrier -Layer Rectifiers[New and Aged : Bergstein's Classification :Wagner's Theory]).-W. Meyer and A.Schmidt. (Zeitschr. f. tech. Phys., No. 1,Vol. '4,1933, pp. 11-18.)

One of the primary objects of this experimentalinvestigation was to test Bergstein's division(Abstracts, 1930, pp. 522-523) of barrier -layerrectifiers into two groups, based partly on theirelectrical behaviour and partly on their mechanicalstructure: one group including the copper -oxiderectifier and the other the iodine and the seleniumrectifiers. As a result of their work the writersconclude that this division has only a " formal "significance, and that there is no fundamentaldifference between the two groups. They also con-clude that Wagner's attempt (1931, p. 625) toexplain the rectifying action without referenceto a barrier layer, by means of a special conductingmechanism of the ion -lattice -forming semi -con-ductor, cannot be accepted : they find most helpfulof all the theory put forward by Schottky (1930,pp. 636-637 ; 1931, pp. 103 and 451 ; 1932, p.232 ; etc.) and developed by van Geel and by Nord-heim (1931, p. 513 ; 1932, pp. 420 and 597 ; alsoWilson, 1932, pp. 108 and 597) of a forced electrondischarge through the barrier layer.

The tests were carried out on plates taken fromsmall commercial rectifiers such as are used foraccumulator charging. The ageing eff.icts in th?.seare chiefly due to the high current densities em-ployed and the consequent comparatively hightemperatures endured : by good cooling or lightloading they can be practically avoided. More-over, the changes-some of them large-found inthe individual resistances as a result of service arefor the most part of little importance as regards d.c.output, owing chiefly to the large leakage of thetransformers used with these small rectifiers. Therectifiers tested were of the following types :lead/copper iodide copper, copper/copper oxidegraphite, copper/copper oxide copper, and lead/selenium nickel, the stroke representing the barrierlayer in each case. Measurements were taken ofthe equivalent series and parallel resistances and ofthe capacity, for new and old plates of the various

1


types, and were plotted against various polarisingvoltages.STEUERKURVEN UNSTET IG GESTEUERTER GASENT-

LADUNGEN (The Control Curves of Inter-mittently Controlled Gas Discharges [in GridControlled Rectifiers, etc.]). -E. Liibcke.(Zeitschr. f. tech. Phys., No. 2, VOL 14, 1933,pp. 61-64.)

THE " STABILISATOR " [Glow -Discharge PotentialDivider and Voltage Regulator].-Stabi-lovolt Company K6r6s. (Hochttech. u.Elek:akus., Nov., 1932, VOL 40, No. 5,pp. 18 5-187.)

Illustrated description of the commercial articlewhose theory was dealt with in 1932 Abstracts,p. 655.DAS ATZNATRON-ELEMENT (The Caustic Soda

Primary Battery with Depolarisation byAtmospheric Oxygen). -Le Carbone Com-pany. (E.T.Z., 9th Feb., 1933, Vol. 54,No. 6, pp. 131-132.)

" The smoothness of the discharge curve forprolonged discharge is much the same as that givenby an American copper -oxide cell . . . , but the LeCarbone caustic soda element has a workingvoltage about 80% higher " [1.2 to I v for a 0.5to 3 A discharge].

TESTS ON THE PERFORMANCE OF DRY CELLS. -U.Ruelle. (L'Elettrotec., i5th Dec., 1932, Vol.19, No. 35, pp. 853-86o.)

SUL CALCOLO DELLE BOBINE DI AUTOINDUZIONE ANUCLEO DI FERRO (The Calculation of Iron -Cored Chokes [for Telephonic Frequencies]).-F. Marocchi. (Alta Frequenza, Sept., 1932,Vol. 1, No. 3, pp. 376-411.)

The writer derives the following simple formulafor the losses in the iron plates or wires, taking intoaccount the non -uniform distribution of the mag-netic flux and the eddy currents, and assuminglimited values of B:- Pre = 14132f" h0 B2f2.5.The coefficient ki and k, characterising the lossesby hysteresis and by eddy currents, can readilybe obtained by measurements of inductance andresistance on a sample of the material, and arerelatively constant in sufficiently wide zones ofvariation of B and of f. Making use of this ex-pression, and of certain geometrical considerations,the properties of any iron -cored choke can besummarised by means of two characteristic para-meters R,. and Q, functions of a core dimension andof the number of turns and the diameter of the wireof the winding. It can thus be shown that thelogarithmic decrement 8 = R/2fL can be given aminimum, by varying either for L (when the lattervariation is produced by altering the size of the airgap). When both quantities fulfil the conditionssimultaneously, a " minimum minimum " decre-ment is obtained. The parameters R,. and Q areeasily calculated for a choke for a definite frequencyor band of frequencies, and from them valuabledesign data can be derived.SCREENING OF A CORE -LESS OIL -COOLED CHOKING

COIL BY MEANS OF AN AUXILIARY WINDING.PART .- J . Hak. (E.T.Z., 9th Feb., 1933,Vol. 54, No. 6, pp. 123-125.)

TRANSFORMERS WITH PEAKED WAVES [e.g. forNeon Lamp Stroboscopes]. -0. Kiltie. (Sum-mary in Review Scient. Instr., Jan., 933,Vol.4, No. 1, p. 41.)

BEITRAG ZUR VORAUSBERECHNUNG DER G UNSTIG-STEN DIMENSIONEN VON ELEKTRISCHENT RAN SFORMATOREN UN D MASCHINEN (Con-tribution to the Calculation of the OptimumDimensions of Electrical Transformers andMachines). -J. La Cour. (Ingeniors Veten-skips Akadentien, Sweden, Handlingar No.120, 1933, 165 pp. In German.)

LIGHT -WEIGHT TRANSFORMERS FOR AIRCRAFT [forTransmitters and Receivers]. -D. W. Grant.(Bell Lab. Record, Feb., 1933, Vol. 11, No.6, pp. 173-177.)

For transformers working at low power levels(e.g., input for transmitter, output for receiver)permalloy cores are used : for those working athigher levels (e.g., output for transmitter, " re-tardation coil " for receiver -in smoothing circuit)silicon steel is retained and other steps are takento reduce the weight. As a result, a complete re-ceiving set weighs only 17 lbs.

NEW APPARATUS AND APPLIANCES [Special Trans-former and Output Choke for " QuiescentPush -Pull "].-(Electrician, roth Feb., 1933,Vol. 1 IO, No. 2854, pp. 200-201.)

ELECTROMAGNETIC SHIELDING AT RADIO FRE-QUENCIES. -King. (See under " Propagationof Waves.")

NEUES VERFAHREN MAGNETISCHER MESS UNGENAN BLECHSTREIFEN (A New Method forMagnetic Measurements on Strips of SheetMetal). -P. C. Hermann. (Zeitschr. f. tech.Phys., No. 1, Vol. 14, 1933, PP- 39-44.)

DIE FLUSSVERDRANGUNG IN EISENBLECHEN INWECHSELWIRKUNG MIT EINEN LUFTSPALT(Magnetic Skin Effect in Laminated Ironwith an Air Gap). -H. Kaden. (Zeitschr. f.tech. Phys., No. 2, Vol. 14, 1933, pp.69-72.)

The writer's analysis shows that as a generalrule the reluctance of such a circuit at all fre-quencies can be calculated with sufficient accuracyas the sum of the air -gap reluctance for a uniformfield distribution and the iron reluctance, withoutattention to their mutual effects.

PERMEABILITY OF. IRON AT ULTRA -RADIO FRE-QUENCIES. -J. B. Hoag and H. Jones.(Phys. Review, i5th Nov., 1932, Series 2,Vol. 42, No. 4, pp. 571-576.)

WAS IST DAS ANOMALE VERHALTEN DER PER-MEABILITAT BEI HOCHFREQUENZ ? (What isAnomalous Behaviour of Permeability atHigh Frequencies ?)-W. Arkadiew. (Zeit-schr. f. Physik, 1932, Vol. 79, No. 7/8, pp.558-561.)


INSULATING SUSPENSIONS OR SUPPORTS FOR SHORT-WAVE COMPONENTS, TO AVOID DIELECTRICLOSSES : THE USE OF METALLIC WINDINGSOF HIGH INDUCTANCE AS INSULATORS. -Telefunken Company. (Summary of GermanPatent in Hochf:tech. u. Elek:akus., Nov.,1932, Vol. 4o, No. 5, p. 183.)

CAPILLARY ACTION IN IMPREGNATED PAPER IN -SULATION.- . B. Whitehead and E. W.Greenfield. (Physics, Dec., 1932, Vol. 3,No. 6, pp. 324-330.)

INSULATING MATERIALS -SOME RECENT, DEVELOP-MENTS [Elephantide Press -Papers, AlkydResins, etc.]. -A. R. Dunton. (Electrician,13th Jan., 1933, Vol. Iv), No. 285o, pp.33-34.)

PLASTIC MATERIALS IN ENGINEERING. -W.(Zeitschr. V.D.I., 17th Dec., 1932, Vol. 76,No. 51, pp. 1233-1239.)

COMMERCIAL TESTING OF SOLID DIELECTRICS[Survey of Methods in Various Countries]. --G. Caviglia. (L'Elettrotec., 15th Dec., 1932,Vol. 19, No. 35, pp. 860-864.)

A SUMMARY OF YEAR'S INSULATION RESEARCH . -J. B. Whitehead and others. (Elec. Engin-eering, Feb., 1933, Vol. 52, pp. 99-106.)

VDE TEST SPECIFICATIONS FOR CONDENSERS FORBROADCAST RECEIVERS AND INTERFERENCEPREVENTION.-Verband Deutscher Elek-trotechniker. (E.T.Z., 9th Feb., 1933, Vol.54, No. 6, pp. 140-142.)

POTENTIOMETERS : A SURVEY. -H. B. Brooks.(Summary in Review Scient. Instr., Jan.,5933, Vol. 4, No. r, p. 42.)

A PEN RECORDER FOR D.C. MILLIVOLTS AND MICRO-AMPERES AT ENERGY LEVELS OF 4-5 MICRO -WATTS. -H. L. Bernarde and L. J. Lunas.(Elec. Engineering, Feb., 1933, Vol. 52, p.124: summary only.)

THE MEASUREMENT OF THE PHASE ANGLES OFSHIELDED RESISTORS. -L. J. Berberich.(Physics, Dec., 1932, Vol. 3, No. 6, pp.296-313.)

" The small, but not always negligible phaseangles of ' non -inductive ' shielded resistors areoften ignored because of the supposed extreme dif-ficulty of their measurement." A substitutionmethod is described which needs no very specialapparatus and gives an accuracy sufficient for themost precise measurements in which such resistorsmay be used. Time constants as low as ro-8 sec.were measured with less than 10% error.

A NEW CONTINUOUSLY VARIABLE RESISTANCE.-(E.T.Z., 8th Feb., 1933, Vol. 54, No. 6, p.133.)

A flexible stranded constantan wire lies freely inthe spiral groove of a rotatable cylinder and alsopasses in a figure -of -eight round a capstan -shaped

April, 1933

" contact -wheel " sliding on a sprung shaft parallelto the cylinder axis. The device is made for variousresistances and for currents up to 15 A ; it can benon -inductive if so desired.

I00 000 -OHM WIRE -WOUND RESISTANCES [forNoiseless Operation of Volume Control, etc.].-Clarostat Company. (Electronics, Dec.,1932, p. 379.)

UN NOUVEAU MODE DE REALISATION DES PHANO-MANES THERMOELECTRIQUES (A New Way ofMaking Use of Thermoelectric Phenomena[Many Thermojunctions in. Series, with Un-broken Continuity in case of Failure of Oneor More Junctions]). -A. Egal. (Comptes'Rendus, 30th Jan., 5933, Vol. 196, No. 5, pp.332-334.)

Two practical designs are shown carrying outthe general principle that the more resistive metal(e.g., constantan) forms a continuous spiral, whilethe second metal (e.g., iron) is made to short-circuitalternate half -turns of this spiral. In the seconddesign the iron is actually deposited on the half -turns.

CONTROL BY " NON-LINEAR " CIRCUITS [Trans-former Combinations, etc.]. -General Elec-tric Company. (Electronics, Jan., 1933, p.

No dtails are given. One circuit produces slowoscillations which will turn on and off an incandes-cent lamp several times per second or as slowly asonce in thirty seconds. It was known for sometime that high frequencies could be produced intransformers under certain conditions, but thesefrequencies are always multiples or integral frac-tions of the frequency of the power supply. Thenew oscillator will produce any frequency in itsrange of usefulness ; vibrations of 5 000 c/s alreadyhave been observed. . . Still another applicationof non-linear circuits has been in resonance relays[working on current and voltage resonance]. . . .

A voltage change of as little as r/ moth of 1% willcause a current change of r00%."

GLOW -DISCHARGE RELAY ACTUATED BY INCREASINGTHE ELECTRODE GAP. -Landis and Gyr.(Summary of French Patent in Rev. Gen.de l'Elec., 28th Jan., 1933, Vol. 33, No. 4, p.32 D.)

At rest, the electrodes are so close that the firstcathode dark space cannot form and no current canpass. If a magnet approaches the end of the dis-charge tube, it moves one electrode so as to increasethe gap, and allows the glow discharge to form, theconsequent current being used in an external circuit.Thus by mounting the magnet on the spindle of anelectricity meter, the device can be used for distantregistration of the meter.

RELAY CIRCUIT FOR POINTER INSTRUMENTS, RE-PLACING POSITIVE CONTACT BY SPARK ORBRUSH DISCHARGE WHICH TRIGGERS A GLOWDISCHARGE RELAY. -(Summary of FrenchPatent in Rev. Gin. de l'Elec., 31st Dec.,1932, Vol. 32, No. 26, p. 205-206 D.)


THE HEAVY -CURRENT GLOW DISCHARGE AT ATMOS-PHERIC PRESSURE : A NEW TYPE OF DIS-CHARGE.-W. Krug : Thoma and Heer.(Zeitschr. f. tech. Phys., No. r, Vol. 14,1933,p. 26.)

A letter prompted by the paper referred to inFebruary Abstracts, p. 112, and referring to thewriter's own results (1932 Abstracts, p. 658.)

A DIRECT CURRENT AMPLIFIER WITH GOOD OPERA-TING CHARACTERISTICS [using RadiotronValves : practically Free from Effects ofFilament and Anode Battery Drifts].-A. H.Taylor and G. P. Kerr. (Review Scient.Instr., Jan., 1933, Vol. 4, No. I., pp. 28-32.)

THE APPLICATION OF A THERMIONIC AMPLIFIERTO THE PHOTOMETRY OF THE STARS.-A. E.Whitford. (Ibid., p. 37 : summary only.)

" A marked improvement in stability has beenobtained by evacuating the space around thephotoelectric cell and amplifier tube." Ampli-fication factor is about 2 200 000.

A D.C. AMPLIFIER FOR PHOTOMETRIC PURPOSES.-P. Donzelot and J. Divoux. (Long sum-mary in Rev. Gin. de l'Elec., 4th Feb., 1933,Vol. 33, No. 5, p. 150.)

Analysis of the relations between the photo-electric current and the three parallel resistancesopen to it (Rm, the high external resistance ; Ro,the resistance of the space between filament andgrid ; and R,, the leakage resistance of socket,etc.) shows that for great amplification R,11? mustbe as small as possible ; but R, must for the samereason be very large, so Ro must be very largeindeed (though " if R, were negative and equalto R, the amplification would be infinite ; butas the grid currents are not stable when R. isnegative, this part of the characteristic can hardlybe used "). Another condition is that the slopemust be steep : all these conditions do not appear tobe satisfied in any one valve on the market. Thewriters therefore use three separate stages : (x) aPhilips electrometer valve with Ro very large ;(2) a stage with very high coefficient of voltageamplification, using Jouaust and Decaux' deviceof a two -grid valve with united grids ; and (3) astage with very steep slope, using two valves inparallel. For the writers' special method of com-pensating the current of repose in the outputcircuit, see February Abstracts, p. 113.

A DIRECT CURRENT AMPLIFIER.-P. A. Macdonaldand J. T. Macpherson. (Phil. Mag., Jan.,1933, Series 7, Vol. 15, No. 96, pp: 72-81.)

The current sensitivity So of a direct currentamplifier is given by So = G,SR, where G. is themutual conductance of the valve used, S the sen-sitivity of the meter in the anode circuit and R theresistance employed to connect the control gridand filament of the valve. The writers describe acircuit in which G, has been increased and S, keptconstant by using the unidirectional conductivityof an auxiliary vacuum tube to alter the potentialof the grid of a triode when its grid current changesdirection. The overall amplification obtained was108, the stability was good. Greater sensitivenesswould be obtained by increasing both G. and S.

A TWO -STAGE MAINS -DRIVEN AMPLIFIER ENABLINGA PHOTOCELL TO OPERATE A COMMERCIALRELAY.-H. L. Kagamaster. (Electronics,Jan., 1933, p. 19.)

THE AUDIO TRANSFORMER AS A SELECTIVE AMPLI-FIER.-Pawley. (See under " Acoustics andAudio -frequencies.")

A FILTER FOR THE INTERMEDIATE -FREQUENCYAMPLIFIER [of a Superheterodyne Receiver].-Piesch. (See under." Reception.")

THE PROPERTIES AND CALCULATION OF THE MUL-TIPLE BRIDGE FILTER.-W. Caller : Jaumann.(E.N.T., Dec., 1932, Vol. 9, pp. 502-504.)

A letter prompted by Jaumann's paper (1932Abstracts, pp. 595-596). Cauer refers to patentstaken out by himself, and points out variousprinters' errors. Jaumann replies.

A NEW SUBSTITUTE FOR THE SLIDE RULE.-P.Ravigneaux. (Comptes Rendus, 9th Jan.,1933, Vol. 196, No. 2, pp. 96-97.)

A NEW HARMONIC ANALYSER [Synchronous Discand M.C. Voltmeter].-B. G. Gates. (Journ.Scient. Instr., Dec., 1932, Vol. 9, pp.38o-386.)

A MECHANICAL AID TO THE ANALYSIS OF COMPLEXSPECTRA.-G. R. Harrison. (Review Scient.Instr., Dec., 1932, Vol. 3, pp. 753-759.)

CURVE PLOTTER AND COMPARATOR FOR LABORATORYAND PRODUCTION TEST USES.-S. Isler.(Electronics, Dec., 1932, p. 361.)

Developed by Wired Radio, Inc., for use in testingreceiving and transmitting filters for frequency andoverall output characteristics.

THE DARBOUX-KOENIGS PLANIGRAPH.-E. J. Nys-trom. (Societas Scientiarum Fennica, Com-mentationes Physico-Mathematicae, Vol. 6[Nos. 1-15], 1932, No. 15, pp. 1-29.)

STATIONS, DESIGN AND OPERATIONULTRA -SHORT RADIO WAVES AND THE CARDIFF-

WESTON-SUPER-MARE RADIO LINK [Huttonto Lavernock].-F. E. Nancarrow. (P.O.Elec. Eng. Journ., Jan., 1933, Vol. 25, Part4, pp. 303-306.)

Simple half -wave horizontal dipoles are used,supported by 50 -ft. wooden poles. The use of abeam system was allowed for but found unnecessaryfor this particular service. " High power " modu-lation is used, and both oscillating and modulatingvalves are of the type V.T.24. The transmittingand receiving huts are spaced some Ioo yds apart tominimise cross -talk.

LA LIAISON RADIOTALEPHONIQUE ENTRE LA CORSEET LE CONTINENT PAR ONDES TRES COURTES(The Ultra -Short -Wave Service betweenCorsica and the Continent).-B. de Clejoulx.(Ann. des P.T.T., Jan., 1933, Vol. 22, No. x ,pp 79-85.)

Short description of the equipment used for the


service dealt with in Jan. Abstracts, p. 54. Thevoice -operated device for the duplex working isbriefly discussed. The paper ends : " In spite ofthe great number of listening hours in the past twoyears, no fixed rules regarding propagation havebeen found. It is very variable. It seems, however,that propagation is better in summer than in winter,or more exactly, in dry weather than in wet. Theinfluence of humidity on the strength of reception isperhaps due to causes at the stations themselves(bad aerial insulation, losses in the receivers) ratherthan to propagation. But it is to be noted thatvery good reception was possible one day lastJanuary when La Turbie was under snow : theaerial insulators were transferred into blocks ofice. Even during the melting of the snow thesame good working persisted.

" Selective fading, moreover, has been noticedfairly frequently. The signal minimum, which issometimes absolute silence, occurs on both waveswith some minutes' interval. It is only very rarelythat fading affects one wave only. Its total dura-tion is seldom more than 20 to 25 minutes. It hasbeen noticed at all hours of the day."

SOME PRACTICAL EXPERIMENTS ON ULTRA -SHORTWAVES [7.75 Metres : B.B.C. and MarconiCompany's Preliminary Tests].-N. Ash -bridge. (World Radio, 3rd Feb., 1933, Vol.16, No. 393, pp. 141 and '44.)

Among points mentioned are the following :-The power radiated was about kw, the aerialbeing of the Franklin type with two " stacked "half -wave aerials. If, with a fairly sensitive re-ceiver, moderately good reception is obtained at15 miles, a decrease in distance to 10 or 71 milesdoes not give as much improvement as might beexpected, and the heavy background hiss due tothe carrier wave still persists : this result is ex-plained by the rapid attenuation over the first fewmiles and the extreme weakness of the fields at thegreater distances, giving readable reception wheremuch stronger signals on ordinary wavelengthswould be drowned by static. The merits of a single -valve superheterodyne adaptor, a super -regenerativereceiver, and a detector with reaction followed byone or two a.f. stages, are compared. No advan-tages of a vertical dipole receiving aerial, as com-pared with an ordinary type of broadcast aerial,were found : in fact, efficient aerials of the lattertype gave the best results.

EXPERIMENTEELE RADIO-OMROEP OP EEN GOLF-LENGTE VAN 7.85 METER TE AMSTERDAM(Broadcasting Tests on a 7.85 -Metre Wave-length in Amsterdam).-P. J, H. A. Nord-lohne. (Tijdschr. Nederland. Radiogen.,July, 1932, Vol. 6, No. I, pp. 1-2o.) See1932 Abstracts, p. 657.

LONG versus SHORT BROADCAST WAVES . . F.Byrne. (Electronics, Jan., 1933, p. 19.)

Summary of measurements in Ohio of the fieldstrengths of nine stations on various frequenciesin the broadcast band. Reduced to a commonbasis, the data showed that a 1 kw station onI 450 kc/s would have a field strength of 0.2 MV/111

at 45 miles ; on T 07o, 700 and 55o kc/s the corres-ponding distances would be 75, 110 and 135 miles :thus the distance is roughly inversely proportionalto the frequency. For the same sequence of fre-quencies the costs of providing a field of 0.5 mvimover one square mile are 15.50, 5.88, -, and 2.06dollars (based on an annual charge of 50 000dollars for a 1 kw station). See also Editorialon p. 20, where the importance of Byrne's resultsis emphasised. Frequencies of 1 000 kc/s or overare uneconomical for large coverage and high power:low -power stations are at present " wasting goodlow -frequency assignments that are suitable forhigh power and large coverage."

THE USE OF TELEPHONIC CIRCUITS FOR SPECIALTRANSMISSIONS : BROADCAST RELAYING :PICTURE TRANSMISSION : CONFERENCES BYTELEPHONE .-C C. Hopfner. (Summary inRev. Gen. de l'Elec., 26th Nov., 1932, Vol.32, pp. 687-688.)

THE BISAMBERG (VIENNA) HIGH -POWER BROAD-CASTING STATION WITH REFLECTOR TOWER.-Telefunken Company. (See abstract under" Aerials and Aerial Systems.")

SOME NEW FOREIGN BROADCASTING STATIONS[Prague, Sottens and Beromunster, Wester-glen].-M. Adam. (Rev. Gin. de l'Elec., TothDec., 1932, Vol. 32, No. 23, pp. 771-779.)

ANSAGER UND ANSAGEN (Announcers and An-nouncing).-H. Lebede. (Radio, B.,F. f. Alle, Jan., 1933, pp. 39-42.)

HOW SHOULD THE SOUND ARCHIVES OF BROAD-CASTING BE USED ? -H. Lebede. (Radio,B., F. f. Alle, Feb., 1933, pp. 91-94.)

THE MADRID CONFERENCE : AMATEUR REPRE-SENTATIVES SUCCESSFUL, AS FOURTEEN -WEEKS' MEETING CONTINUES ALL AMA-TEUR BA NDS. -K. B. Warner. (QST, Feb.,1933, Vol. 17, pp. 9-17.)

RADIOELECTRIC COMMUNICATION WITH SHIPS ATSEA.-E. Picault. (Ann. des P.T.T., Dec.,1932, Vol. 21, pp. 1021-1044.)

NORTH ATLANTIC SHIP -SHORE RADIOTELEPHONETRANSMISSION DURING 1930 AND 1931[Analysis of Short -Wave Data : FieldStrength Contour Diagrams, etc.].-C. M.Anderson. (Proc. Inst. Rad. Eng., Jan.,1933, Vol. 21, pp. 81-101.)

SHORT-WAVE TRANSMISSION TO SOUTH AMERICA[Survey of Transmission Conditions : Ab-sence of Seasonal Effect : Characteristicsof Atmospheric Noise, etc.].-C. R. Bur-rows and E. J. Howard. (Proc. Inst. Rad.Eng., Jan., 1933, Vol. 21, pp. 102-113.)

For a summarised comparison with transatlantictransmission, and reasons for the differences, seepp. 109 and 112-113.

April, 1933 230

TRANSOCEANIC RECEPTION OF HIGH -FREQUENCYTELEPHONE SIGNALS [Short -Wave Relayingfor International Broadcasting]. -R. M.Morris and W. A. R. Brown. (Proc. Inst.Rad. Eng., Jan., 1933, Vol. 21, pp. 63-80.)

Dealing with the method used in rating thesuitability of the reception for re -broadcasting, theeffects of magnetic disturbances upon trans-mission, the correlation of magnetic activity withtransmission, and the forecasting of magneticdisturbances and resultant transmission conditions.HIGH POWER TRANSMISSION AND RECEPTION AT

ONE STATION [Beam Station at Salisbury,Rhodesia]. -Marconi Company. (Rad. Eng-ineering, Dec., 1932, Vol. 12, p. 19.)

SYSTEM FOR THE ELIMINATION OF EFFECTS OFFADING AND ATMOSPHERICS IN HIGH-SPEEDTELEGRAPHY, USING SIGNAL REPETITIONMETHOD.-Telefunken Company : Schroter.(Summary of German Patent in Hochf:lech.u. Elek :akus ., NOV., 1932, Vol. 40, No. 5,p. 184.) See Figs. 19 and 20.

DIE DEUTSCHEN ANLAGEN FUR DEN DRAHTLOSENT.IBERSEEVERKEHR (The German Stationsfor Oversea Service [Nauen, Beelitz, and theCentral Offices in Berlin and Hamburg]). -W. Hahn. (E.T.Z., i5th Dec., 1932, Vol. 53,No. 5o, pp. 1197-1201.)

RADIO -TELEPHONY [Survey for 1932, including the" Compandor " Device]. -H. Faulkner.(Electrician, 27th Jan., 1933, Vol. ix o, No.2852, pp. 109-1I0.)

" A new device known as the ' Compandor ' isbeing tried out, principally on the long -wavetransatlantic circuit, which is found to give aconsiderable improvement in signal/noise ratio.The device is arranged to amplify the weak syllablesof speech to a greater extent than the stronger at thetransmitting end, and to reverse the process at thereceiving end. Thus noise entering the circuit at apoint beyond the transmitting unit is -so long as itis of lower mean level than the speech -attenuatedalong with the weaker speech elements before beingpassed on to the listener. Since the weaker speechelements, however, are only attenuated to theircorrect proportionate strength, a net reduction innoise is achieved, while speech is received in itscorrect proportions. This device has been in opera-tion now for several months and has been foundto effect a real improvement in the circuitconditions."

GENERAL PHYSICAL ARTICLESTHE ANALYSIS OF COMPOUND WAVE FORMS. --J. S.

Thompson. (Phil. Mag., Jan., 1933, Series 7,Vol. 15, No. 96, pp. 1-15.)

This paper classifies the forms which may arisein the analysis of a wave form compounded of a pairof sine waves, and provides a convenient methodfor performing the analysis.ON SOME APPLICATIONS OF THE ABSOLUTE DIF-

FERENTIAL CALCULUS TO PHYSICS [to Hooke'sLaw in Elasticity Theory and the Constitu-tive Relations in Electromagnetic Theory]. -C. Kaplan. (Phys. Review, i5th Jan., 1933,Series 2, VOI."...43, NO. 2, pp. 137-142.)

THE WIRELESS ENGINEER &

TIBER DIE GRUNDLAGEN DER GEOMETRISCHENELEKTRONENOPTIK (The Foundations ofGeometrical Electron Optics). -E. Brilche.(Zeitschr. f. tech. Phys., No. 2, VOL 14,1933, PP- 49-58-)

DIE AUSBILDUNG DER MAXWELLVERTEILUNG IMLANGMUIRSCHEN PLASMA (The Formation ofthe Maxwell Distribution in a LangmuirPlasma [Theoretical Investigation]). -D.Gabor. (Physik. Zeitschr., rst Jan., 1933,Vol. 34, No. 1, pp. 38-45.)

QUANTUM MECHANICS OF COLLISION PROCESSES,PART IL -P. M. Morse. (Reviews of Mod.Phys., July, 1932, Vol. 4, No. 3, PP. 577-634-)

ZUR ERKLXRUNG VON LICHTERSCHEINUNGEN BEIMELEKTRISCHEN GASDURCHSCHLAG (The Ex-planation of Luminous Phenomena [inAbsence of Appreciable Voltage Drop] inSpark Discharge through Gases). -W. Fucks.(Zeitschr. f. tech. Phys., No. 2, Vol. 14, 1933,PP. 59-61.)

CONTEMPORARY ADVANCES IN PHYSICS, XXV.HIGH FREQUENCY PHENOMENA IN GASES,SECOND PART. -K. K. Darrow. (Bell S.Tech. Journ., Jan., 1933, Vol. 12, No. 1, pp.91-118.)

For the first part see March Abstracts, p. 174.

DIELECTRIC CONSTANT AND PARTICLE SIZE. -J. W. Williams and J. L. Oncley. (Physics,Dec., 1932, Vol. 3, No. 6, pp. 314-323.)

MISCELLANEOUS

PROBABILITY THEORY AND TELEPHONE TRANS-MISSION ENGINEERING. -R. S. Hoyt. (BellS. Tech. Journ., Jan., 1933, Vol. 12, No. I,PP. 35-75-)

This paper " contributes methods, theorems,formulas and graphs to meet a previously unfilledneed in dealing with certain types of two-dimen-sional probability problems -especially those re-lating to alternating current transmission systemsand networks " with two dimensional variables.

ON THE PRIOR PROBABILITY IN THE THEORY OFSAMPLING. -H. Jeffreys. (Proc. Camb. Phil.Soc., Jan., 1933, Vol. 29, No. r, pp. 83-87.)

REMARK ON THE PAPER BY G. KOHLER AND A.WALTHER : ON THE FOURIER ANALYSISOF FUNCTIONS WITH DISCONTINUITIES, COR-NERS AND SIMILAR PECULIARITIES. -R.Feinberg. (Archie f. Elektrot., 7th Jan.,1933, Vol. 27, No. I, pp. 15-18.)

For reference to the paper named see 1932Abstracts, p. 113. The writer of this note finds thatinfinite series occurring in the Fourier analysis ofdiscontinuous functions, etc., may diverge andcause the method to fail. He also gives a graphicalmethod of analysis. In a succeeding note (pp. 19-20)Walther shows the divergence to be unimportantas far as applications are concerned and gives thehistory of the method.


DEVELOPMENT AND PROGRAMME OF THE DIFFER-ENTIAL ANALYSER.-V. Bush. (Summary inScience, 2nd Dec., 1932, Vol. 76, No. 1979,P. 519.)

Results achieved with, and probable developmentof, the machine referred to in 1932 Abstracts, p.113.

DISSEMINATION OF SCIENTIFIC KNOWLEDGE.-E. Leifson. (Science, 3oth Dec., 1932, Vol.76, No. 1983, p. 624.)

It is suggested that short " previews " of scien-tific papers should be " submitted to the journal ofpreviews after the full reports have been accepted forpublication by a standard journal. Within onemonth the previews should appear in print."TABLE OF ORGANISATION AND PROGRAMME OF

HIGHER EDUCATION IN ELECTRICAL ENGIN-EERING IN DIFFERENT COUNTRIES.-(Rev.GM. de l'Elec., 31st Dec., 1932, Vol. 32,No. 26, pp. 86o-863.)

UBER DAS ORDNEN TECHNISCHEN SCHRIFTUMS(On the Classification of Technical Papers[The Brussels Decimal Classification Sys-tem]).-F. Moench. (E.N.T., Dec., 1932,Vol. 9, pP 496-500.)

THE U.S.A. PATENT " POOL."-( Wireless Engineer,Jan., 1933, Vol. 1o, No. 112, p. 18.)

NEW APPLICATIONS AND THE EXPANSION OFRECENT USES OF ELECTRICITY : RECENTAPPLICATIONS OF ELECTRICAL METHODS.-G. F. Sills : D. H. Bishop. (Journ. I.E.E.,Jan., 1933, Vol. 72, No. 433, pp. 45-5051-55.)

LOOKING AHEAD AS 1933 OPENS.-(Electronics, Jan.,1933, P. 2-3.)

New aids to television : Rochelle salt crystalsto modulate the light beam : Myers mercury -vapour lamp (March Abstracts, p. 168) : ultra -shortwaves (field -strength measurements on signals fromEmpire State Building aerials 128o ft from ground).Photo -sensitive cells (speed -traps for motorists :restaurant " magic doors " save enough dish -breakage to offset the carrying cost of the instal-lation, and speeding -up of service amounts toequivalent of 2 500 dollars worth of rental space).Power uses of rectifiers : new type of metal -envelope tube (indirectly heated cathode) rectifyingloo A at usual voltage and frequencies. Beautythe new note in broadcast receivers.

RADIO -TELEPHONY IN MINES.-F. Noack. (Sum-mary in Electronics, Dec., 1932, p. 377.)

RADIO -TELEPHONY [Survey for 1932, including the" Compandor " Device].-Faulkner. (Seeunder " Stations, Design and Operation.")

THE THEORY OF HEATING BY INDUCED H.F.CURRENTS.-G. Ribaud. (Journ. de Phys.et le Rad., NOV., 1932, Vol. 3, pp. 537-554.)

THE PRODUCTION OF THE MERGET PHENOMENON[Gaseous Thermodiffusion] BY D'ARSONVAL-ISATION BY ULTRA -SHORT WAVES.-H.Bordier. (Comptes Rendus, 23rd Jan., 1933,Vol. 196, No. 4, pp. 255-257.)

April, 1933

" RADIO KNIFE " EMPLOYING INDUCED CURRENTS[thus dispensing with Electric Cord Con-nection]. (Electronics, Jan., 1933, p. zy.)

RECORDING AND MEASURING BLOOD PRESSUREBY A PIEZOELECTRIC METHOD.-L. Bugnard,P. Gley and A. Langevin. (Comptes Rendus,23rd Jan., 1933, Vol. 196, No. 4, pp. 293-295.)

Further development of the work referred to in193o Abstracts, p. 647 (Boulle and Langevin).

USE OF THE PHOTOELECTRIC CELL IN BIOLOGY[Study of Area and Shape of Blood Cells].-A. Savage. (National Research Council,Canada, 14th Annual Report, 193o-1931,p. 126.)

The method dealt with in 1931 Abstracts, p. 170,has been applied to the comparison of normal andanaemic blood cells.

PHOTOPHONY [Photo -Telephony] WITH VOLTAGEMODULATION OF INCANDESCENT LAMP-VIEWED FROM ITS WAVE FORM.-S. Suzuki.(Journ. I.E.E. Japan, Dec.; 1932, Vol. 52[No. 12], No. 533 : English summaryP. 134.)

The successful use of an incandescent lamp atthe transmitter has already been described byKujirai (1932 Abstracts, p. 540). The presentwriter has previously shown that the wave form ofthe filament temperature takes the integrated formof the square of the applied voltage, if the frequencyis within the range of audible frequencies. In thispaper he deals with the transient state of thefilament temperature with periodic wave form. Adiagram of the photophone is given on p. 961.

VARIATION OF LUMINOUS INTENSITY OF IN-CANDESCENT LAMPS AS A FUNCTION OFVARIATION OF APPLIED VOLTAGE [for VariousTypes of Filament].-H. Pecheux. (Rev.Gen. de l'Elec., 31st Dec., 1932, Vol. 32,No. 26, pp. 867-871.)

MEASUREMENT OF YARN THICKNESS AND EVEN-NESS [by Photoelectric Method].-Stanbury.(Electronics, Dec., 1932, p. 370.) Forsimilar use on silk see February Abstracts,p. 117, top 1. h. column.

CLOTH -DYEING MACHINE WITH PHOTOELECTRICREVERSER : ELECTRIC EYE WATCHES BOWL-ING -ALLEY " FOUL LINE " : SMOKE RE-CORDER AND TRUCK -HEIGHT INDICATOR INHOLLAND TUNNEL.-(Ibid., p. 371.)

METHODS OF MEASURING AND RECORDING RAPIDLYVARYING PRESSURES [as in AeroplaneEngine Cylinders : a Photoelectric Method].-A. Labarthe and M. Demontvignier.(Comptes Rendus, 19th Dec., 1932, Vol. 195,No. 25, pp. 1240-1242.)

A [Photoelectric] SMOKE DENSITY METER.-R. D.Bean. (Journ. Scient. Instr., Dec., 1932,Vol. 9, pp. 391-392.)

April, 1933 232

EXPERIMENTS ON POLARIM ETRY BY PHOTOELECTRICTECHNIQUE : MEASUREMENTS OF THEROTATORY DISPERSIONS OF SOME SUGARS.-G. Bruhat and P. Chatelain. (Journ. dePhys. et le Rad., Nov., 1932, Vol. 4, pp. 501-51r.)

THE IMPORTANCE OF THE BARRIER -LAYER PHOTO-CELL IN ILLUMINATING ENGINEERING. --L. Bloch. (Summary in Rev. Gen. de l'Elec.,lath Nov., 1932, Vol. 32, pp. 619-620.)

TORSIONAL VIBRATIONS IN AXLES AND SHAFTSAND THE RESONANCE OSCILLATIONS INFILTERS [Suggestion of a Method of Testing].-S. Treves. (Alta Frequenza, Sept., 1932,Vol. i, No. 3, pp. 429-431.)

A letter prompted by Sacerdote's paper on ultra -micrometric and other methods (1932 Abstracts,pp. 480 and 6oi).

TESTS ON HIGH-SPEED DIESEL ENGINES BYELECTRICAL [Piezoelectric, Ultra -Micro-metric, etc.] METHODS. -A. Nagel. (Zeitschr.V.D.I., loth Dec., 1932, Vol. 76, No. 5o,pp. 1213-1217.)

MEASUREMENT OF THE DISSIPATION OF VIBRATORYENERGY IN A STEEL BAR BY AN ELECTRICALMETHOD. -A. L. Kimball. (Summary inRev. Gen. de l'Elec., 24th Dec., 1932, Vol. 32,No. 25, p. 826.)

In a series of articles on papers dealing with" Various Applications of Electricity " read at theInternational Electricity Congress.

THE EMPLOYMENT OF THE RECIPROCITY RELATIONSIN THE INVESTIGATION OF MECHANICALVIBRATIONS. -A. Steinheil. (Zeitschr. f.tech. Phys., No. I, Vol. 14,1933, pp. 36-39.)

THE INVESTIGATION OF MECHANICAL VIBRATIONSBY PHASE MEASUREMENT [using a Glow -Discharge Lamp rotating with the Machineand fed through a Contact -Breaker influencedby the Vibrations]. -W. Spath. (E.T.Z.,5th Jan., 1933, Vol. 54, No. r, pp. ro-12.)

THE SICCOMETER : AN ULTRA -MICROMETER FORTHE TESTING OF MOISTURE IN THE MANU-FACTURE OF PAPER. -H. Carsten and C. H.Walter : Siemens & Halske. (E.T.Z.,2nd Feb., 1933, Vol. 54, No. 5, p. 109.)

Cf. Walter, 1932 Abstracts, p. 651, and backreference (Jaekel).

THE [Ultra -Micrometric] MEASUREMENT OF SMALLELECTRODYNAMIC FORCES WITH THE CON-DENSER MICROPHONE. -Agricola. (See under" Measurements and Standards.")

THE SOUNDING STRING AS A MEASURING APPARATUS[as Ultra -Micrometric Device for SmallMovements in Buildings, Hulls, etc., andLathe Investigations]. -H. Haake. (Zeitschr.V.D.I., loth Dec., 1932, Vol. 76, No. 5o,pp. 1228-1229.)

Two electromagnetically vibrated strings are

THE WIRELESS ENGINEER &

used, tuned to the same note. An accuracy of± 1% is given.

MEASUREMENT OF VELOCITY OF PROJECTILES BYMEANS OF KERR CELL.-Cranz, Kuttererand Schardin. (Summary in Zeitschr. V.D.I.,26th Nov., 1932, Vol. 76, No. 48, p. 1183.)

ELECTRONIC DEVICES FOR INDUSTRIAL CONTROL. -F. H. Gulliksen. (Elec. Engineering, Feb.,1933, Vol. 52, pp. 106-11o.)

THE INTERNATIONAL ORGANISATION OF BROAD- .

CASTING. -P. Brenot. (Rev. Gen. de l'Elec.,17th Dec., 1932, Vol. 32, No. 24, pp. 185-186B.)

THE 8TH " DEUTSCHE PHYSIKER- UND MATHE-MATIKERTAG," BAD NAUHEIM, 1932. -(Listof Papers, and short Summaries, in E.T.Z.,29th Dec., 1932, Vol. 53, No. 52, pp. 1247-1250.)

SCIENTIFIC INSTRUMENTS AND AERONAUTICS. -H. E. Wimperis. (Journ. Scient. Instr.,Nov., 1932, Vol. 9, pp. 337-341.)

RECENT RADIO RESEARCH. (Nature, 4th Feb.,1933, Vol. 131, pp. 156-157.)

A note on the Report of the Radio Research Boardfor the Year 1931.

THE ACTIVITIES OF THE PHYSIKALISCH-TECHNISCHEREICHSANSTALT IN THE YEAR 1931.-(E.T.Z.,15th Dec., 1932, Vol. 53, No. 5o, pp. 1203-1204.)

THE PHYSICAL SOCIETY'S EXHIBITION, JANUARY1933.-(Journ. Scient. Instr., Feb., 1933,Vol. ro : Exhibition Number. Also WirelessWorld, 13th Jan., 1933, Vol. 32, pp. 34-35 ;Wireless Engineer, Feb., 1933, Vol. ro,PP. 74-77 ; World Power, Feb., 1933, Vol. 19,PP. 75-79.)

PITTSBURGH'S CONTRIBUTIONS TO RADIO. --S. M.Kintner : Fessenden. (Proc. Inst. Rad.Eng.; Dec., 1932, Vol. 20, pp. 1849-1862.)

NOTES ON WIRELESS HISTORY. -J. E. Taylor.(P.O. Elec. Eng. Journ., Jan., 1933, Vol. 25,Part 4, pp. 295-299.)

ELECTRODELESS DISCHARGE AND SOURCES OFMONOCHROMATIC LIGHT. -F. Esclangon.(Review Scient. Instr., Jan., 1933, Vol. 4,No. I, p. 38.)

Summary of a paper on the work previouslydealt with in a Comptes Rendus Note (1932 Abstracts,p. 422).

THE PERMANENT ELECTRET.-M. Eguchi. (Sum-mary in E.T.Z., 2nd Feb., 1933, Vol. 54,No. 5, p. 112.)

=

EXPERIMENTAL WIRELESS a t3

Some Recent PatentsApril, 1933

The following abstracts are prepared, with the permission of the Controller of ILM, Stationery Office, fromSpecifications obtainable at the Patent Office, 25, Southampton Buildings, London, W .C.2, price I/- each.

RECTIFYING VALVESConvention date (U.S.A.), 17th September, 1930.

No. 379736In order to prevent " arcing " between the anodes

of a double -wave rectifier, the cathode is made inthe form of a strip or ribbon and is mounted spirallybetween two supports, so as to form a shield betweenthe two anodes. The cathode is partly surroundedby a reflector element, which confines the heat andhelps to maintain a unifortn operating temperature.

Patent issued to Standard Telephones and Cables,Ltd.

HIGH -FREQUENCY GENERATORSConvention date (France), 9th March, 1931.

No. 381087An electrically -driven tuning fork F, ther-

mostatically controlled for constant frequency,is used to generate an alternating current which isfed through an amplifier (not shown) to a syn-chronous motor M. This drives a toothed discD which intercepts the light from a lamp L fallingon a photo -electric cell C. The output from thecell passes through an amplifier A, which mayinclude frequency -doubling stages, to the controlelectrodes E, E, of a cathode ray tube K. Aphase -shifting reactance R is interposed in the

cathode ray is swept in a circular path over theperiphery of a toothed disc D1 located at the farend of the tube. The interrupted ray falls on to

K

No. 381087

a photo -electric cell C1, and the resulting currentpulsations are fed to current amplifiers and usedto energise a transmitting aerial. The equipmentis capable of generating a frequency of 21,000kilocycles.

Patent issued to Cie Generale de Telegraphiesans Fil.

PUSH-PULL CIRCUITSConvention date (U.S.A.), 9th July, 1930.

No. 381233A push-pull circuit, used either for high -power

amplification or as an oscillation generator, ischaracterised by the provision of individual grid -biasing resistances R, R1 inserted in the plate -filament circuit between H.T. negative and a

No. 381233

centre -tapping on the input coil L. A high re-sistance R2, common to. each grid, is shunted bycondensers C, C1, and forms a filter circuit toeliminate any " hum " components present in thespace current. Both valves are energised from thesame heating -secondary S, as shown, though withdirectly -heated A.C. valves a separate centre -tappedsecondary is provided for each valve. The arrange-ment permits valves having slightly differentcharacteristics to be accurately " matched," andalso prevents one valve from damage by excessiveplate current should the other valve become in-active for any reason.

Patent issued to Electrical Research Products Inc.

AUTOMATIC TONE COMPENSATIONConvention date (U.S.A.), 29th April, 1930.

No. 378539In order to preserve an " apparent " fidelity in

reproducing speech or music, it is necessary tooffset an inherent defect in the human ear bylowering the proportion of the lower notes tothe higher, as the overall intensity of the volumeof sound is increased. To secure this result theinput to a multi -valve low -frequency amplifieris divided into two parts, one passing through theamplifier, and the other into a regulating or pilotcircuit in which a variable voltage is produced tobias the control electrodes of the main amplifier.The adjustments are such that an increase in signalinput results in a greater amplification of the highnotes than the low in the final output.

Patent issued to Marconi's Wireless TelegraphCo., Ltd.

April, 1933

DETECTORS OR RECTIFIERSApplication date 26th June, 1931. No. 381555A cylindrical body B of ebonite or similar insulat-

ing material is formed with four peripheral slots(of which only two are shown) each containing acolumn of superposed copper -oxide rectifying-

-N1

No. 381555

discs D approximately one -twelfth of an inch indiameter and one -thirtieth of an inch thick. Atthe centre of each disc is a contact zone formedby coating the oxide surface with graphite bymeans of a hard pencil. Interposed betweeneach disc is a soft -metal contact -member F formedas the frustum of a cone, the smaller end beingcoated with tin. The rectifying elements are keptin contact by means of springs N, NI. Prongs P,Pi allow the device to beplugged into a standardtype of valve -holder.

Patent issued to A. V.Tomlinson.

AMPLIFIER CIRCUITSConvention date (U.S.A.),

19th March, 1931.No. 380377

A four -electrode valve isoperated with a high positivevoltage on the grid adjacentto the anode so as to have anegative -resistance character-istic, and is used as a high -frequency amplifier with atuned input and tuned out-put. The electrode potentialsare adjusted so that thevariation of the ratio of theexternal plate impedance tothe internal plate -cathoderesistance is automaticallyreduced as the output circuit is tuned downthrough the broadcast waveband, whereby thenormal gain with frequency is reduced, anda straight-line frequency response is ensured.Alternatively the ratio of plate load impedanceto plate -cathode resistance is made substantially


independent of the frequency to which the outputcircuit is tuned.


AMPLIFYING CIRCUITSConvention date (Germany), 24th February, 1931.

No. 380694When a detector valve is followed by large low -

frequency amplification, stray fields from the supplytransformer induce L.F. voltages in the grid coilwhich are substantially amplified in succeedingstages and give rise to undesirable noise in the loud-speaker. To prevent this the grid coil is dividedinto two equal and .oppositely -wound parts,arranged either in series or parallel, so that anyextraneous voltage pick-up reaches the grid inphase -opposition and therefore produces no effectin the output.

Patent issued to Ideal Werke A -G fur drahtloseTelephonie.

ELIMINATING INTERFERENCEConvention date (Austria), igth December, 1930.

No. 381784Incoming signals are heterodyned by two local

oscillators 0, 0i, to produce two beat frequencies,one F of which is audible, the other Fi supersonic.Both derived frequencies are passed to separatedetector valves D, Di, coupled in opposition to anoutput coil L. In the output from valve D there willbe present the desired signals and the interference.In the output from valve Di only the interference

No. 381784

will be audible, since the beat frequency is super-sonic. Since the two are in opposition, the interfer-ence is cancelled out, leaving the signals alone topass through; to the secondary of the outputcoil L.

Patent issued to J. Fuchs.


DETECTOR -OSCILLATORSApplication date list July, 1931. No. 382286

Relates to means for ensuring sufficient feed-back, as well as efficient rectification, over a widerange of frequencies in a self -oscillating detector,as used, for instance, in a superheterodyne receiver.The output circuit of the valve comprises a variable

235

L.T.+

L.T.-H.T:-

No. 382286

resistance R, of the same order as the internalvalve impedance, in series with an output orcoupling impedance Z of much higher value.Reaction is provided through a variable condenserC, whilst by-passing condensers Cr and C2 areinserted on each side of the impedance Z couplingthe detector valve to the intermediate frequencystages.

Patent issued to The Marconi InternationalMarine Communication Co., Ltd., and S. E. A.Landale.

TELEVISION SYSTEMSApplication date, 24th July, 1931. No. 381254In order to utilise more advantageously the

limited frequency -band available for radio trans-mission, the initial scanning -frequencies are con-densed into a smaller band by means of a series offrequency changers and associated filters beforebeing transmitted. To recompose the picture inreception the incoming signals are applied to filters,the output in each case being either passed throughfrequency -changers or applied to circuits resonantto the original frequencies. The system is alsostated to simplify the synchronisation and phase -control of the scanning means used at the trans-mitting and receiving ends.

Patent issued to J. C. Wilson and Baird Tele-vision, Ltd.

Application date, 27th July, 1931. No. 382326Relates to a television system of the kind in

which the received picture is built up on a closelygrouped bank of lamps, each lamp correspondingto a definite picture element or signal. Accordingto the invention, a secondary " light " frame-of-reference, is superposed on the main surfaceso that the brightness of any particular point onthe picture is judged by contrast with the illumina-tion of the frame of reference, and not by contrastwith the previous picture -signal corresponding tothat point. For instance, if the secondary groupof lamps forming the " reference frame " is suppliedwith current modulated in a direction the reverse

April, 1933

of that carrying the actual picture, the relativeintensity of the picture as judged by the eye willbe doubled.

Patent issued to Marconi's Wireless TelegraphCo., Ltd., and H. M. Dowsett.

SCANNING DISCSConvention date (Germany), loth April, 1931.

No. 379646Scanning discs used in television are usually

cut -away for some distance from the hub so thatthe outer parts are supported only by spokes ofsolid material. When running at high speed thesespokes create friction with the air which tends tocause the disc as a whole to " flutter " transverselyand to produce undesirable noise. According to theinvention such transverse vibrations are preventedby mounting on each side of the scanning disc,and on the same driving shaft, an auxiliary disc ofthin metal, stout paper, or cardboard extendingradially as far as the " spoked " portion of themain disc. The effect is to produce a more or lessquiescent air -space between the main and auxiliarydiscs, which ensures a steady and quiet rotation.

Patent issued to Fernseh A -G.

WIRED -WIRELESS SYSTEMSApplication date 22nd February, 1932.

No. 381449Each subscriber is given the choice of two

programmes over a three -wire distributing network.The output from one receiver A is connected to thelines r, 2, whilst that from a second receiver B istaken to a line 3 and to a mid -point tapping on aresistance S shunting the output from the firstreceiver, so that the programme in this case iscarried by a " phantom " circuit comprising theline 3 and the lines I, 2 in parallel. The circuitsat the receiving end are shown at R, and the

No. 381449

alternative plug -points at P and Ps, the arrange-ment being such that the phantom circuit is auto-matically opened when the loud -speaker is insertedacross the lines r, 2, so as to limit the load on theselines.

Patent issued to J. W. Field and J. E. Ball.


TRANS/JUTTING CIRCUITS

Convention date (Germany), 11th February, 1931.No. 380682.

When renewing or replacing one valve generatorby another in a transmitting installation, it isgenerally necessary to re -calibrate the circuitsover the whole of the working wavelengths owingto the varying interelectrode capacities of differentvalves. According to the invention an auxiliarycondenser is inserted in the tuned H.F. circuit, inshunt with the inherent valve capacity, so as to

c> V V

0-0 0

0

F

0

Fl

aaaa

aaa a

are applied to the grid from a coil L after passingthrough amplifiers V. Synchronising frequenciesare tapped off through filters F, F1 and are appliedthrough amplifiers Vi to a pair of deflecting coils S.Sr which cause the ray to traverse the fluorescentscreen X at the speed necessary to produce per-sistence of vision. The picture may finally beprojected through a lens on to an observationscreen 0. The ray is accelerated towards the endof its travel by a second anode A I formed as acup -shaped meshwork housed in the end of thecathode tube, as shown, and carrying a high

No. 381306

supplement the latter. When an existing valveoscillator is changed for another, it is only necessaryto readjust the supplementary condenser for oneselected wavelength to ensure a complete recali-bration of the circuit for all frequencies.

Patent issued to C. Lorenz A -G.

CATHODE-RAY TELEVISIONConvention date (U.S.A.), 25th September, 1930.

No. 381306Relates to systems of the type in which the

light -and -shade effects of the original picture arereproduced by the varying velocity of impact ofan electron stream upon a fluorescent screen. Theinvention is directed to the provision of means forensuring that the deflection necessary to traversethe ray over the viewing -screen is effected whilstthe stream is moving at a relatively -low speed,the stream being subsequently accelerated to securea bright image. The scanning ray R is producedfrom a cathode gun comprising a heater, cathode,grid G and anode A. The incoming signal elements

positive voltage from a source B. The anode Aimakes contact with a silver lining deposited on theinner walls of the tube.


SCREEN -GRID AMPLIFIERSApplication date, 17th April, 1931. No. 380827In order to increase the permissible voltage

swing applied to the last stage of S.G. amplification,without passing over the " knee " of the char-acteristic curve and so giving rise to cross -modu-lation, the internal electrodes of the valve aremodified by widening the pitch of the control gridand enlarging the mesh of the screening grid, so asto reduce the internal impedance to a fraction ofthe normal value. The valve is then operated withscreen and plate potentials set below the potentialdifference at which secondary emission from theplate occurs.

Patent issued to The Gramophone Co., Ltd., andC. S. Bull.


(RI

"MALL VOLTAGE EFFECT

NISTOBS'Phone: Gladstone 2582.

TAMMYHIGH GRADE

SOUND REPRODUCERSP.A. LOUD SPEAKERS

AND

OTHER EQUIPMENT SUIT-ABLE FOR HOTELS, DANCEHALLS, SPORTS MEETINGS,

ETC.Full Details or Demonstration

1.7, DALTON RT., W. NOR WOOD, 11.1.27Boottbh Agentst-Blesurs. Sottisen (Scotland) Ltd.,

119. George Street, Edinburgh

FOURTH EDITIONHANDBOOK

ofTECHNICALINSTRUCTION

forWireless Telegraphists

ByH. M. DOWSETT,M.I.E.E., F.Inst.P., M.Inst.R.E.

Author of.' Wireless Telegraph, efld Telephony," el,.

THIS work constitutes a completetextbook for the use of wireless

telegraphists. In this new edition thesubject matter has been entirely recastand the scope of the book widened tomeet present-day requirements.

This volume provides a completetheoretical course for the P.M.G. cer-tificate.

A leaflet giving synopsis of chaptersand other particulars will be sent onrequest.

PRICE25/- net. By post 25/9

From all leading Booksellersor direct from the Publishers

ILIFFE & SONS LTD.Dorset House, Tudor St., London, E.C.4

The change of resistance with voltage changes is sosmall in Erie Resistors that it is not necessary tospecify the test voltage except where special precisionis required. Invariably the resistance change is lessthan 2%, as, for instance, with a No. 4 (} watt)Resistor there is a resistance change of less than 1%when the voltage is changed from 100 volts to 150volts. Thus only one size of resistor is needed where thesame resistance value is to be used on different volt-ages. The safety and certainty of Erie Resistors meritcareful consideration. Technical Data Free on Request.

ERIE RESISTOR,LTD.WATERLOO ROAD. CRICKLEWOOD, LONDON,N.W.1

PATENT AGENTS.A. Mathisen, B.Sc. (Elect. Eng.),Specialist : Patents covering electricalradio inventions of high technicalstandard. Preliminary advice free.-First Avenue House, High Holborn,London, W.C.i. [0268

PATENTS, ETC.The Proprietors of British Letters PatentNo. 230,026 for Gaseous Space CurrentDevices, and Nos. 248,721 and 248,722for Electrionic Discharge Devices, aredesirous of entering into negotiationswith interested parties with a viewto the exploitation of the inventionseither by sale of the Patent Rights orby the granting of Licences on reason-able terms. All enquiries should beaddressed to Wm. Brooks & Son,No. r, Quality Court, Chancery Lane,London, W.C.2. [0270

RICH & BUNDYEXTRA HEAVY DUTY

MAINS TRANSFORMERMODEL W.211.

For use with Westinghouse MetalRectifier H.T.11.Secondaries. 300v. 550 rut/a.

4 v.c.t. 2 amps.4 v.c.t. 6 amps.

Full specification on request.Price £3 . 10 . 0

Special Models supplied for Cinemas, BroadcastRelay, and Public Address work.All RICH & BUNDY Mains Transformers forRadio are fitted with Screened Primary.Every Transformer and choke an individual job.We welcome specialist work.

Full details of Mains Transformers, Outputand Filter Chokes, Auto -Transformers andOutput Transformers are given in our

NEW ART CATALOGUE "E."May we send you a copy ?

RICH dr. BUNDY LTD.,Transformer Manufacturers,

New Road, Ponders End, Middx.'Phone: Brute 40777.

dole distributors UFA: HELLY & SAUL Ltd.,47, Fleet Street, Dublin, C.4.

W.E.14.Kindly mention " The Wireless

xxxxxxxx:4xxxxxxxxxxxxxxxCalculation and icc

5c Measurement of xx Inductance and 3

x3 4,

Capacityx> BY W. H. NO TT A G E, B.Sc.X (1924) The more generally useful formula, and methods of measurement of inductanceX and capacity are here brought togetherX in a convenient form. 224 pages. OverXX So illustrations and diagrams.x Price 7/6. By Post 8/-.XX

ILIFFE & SONS LTD.,X Dorset House, Tudor St., London, E.C.4,X and leading booksellers. W.E.rsXXXXXXXXXXXXXXXXXX'XXXXXX

NOW READYPublishers'

BINDINGCASESin Grey Cloth

for

VOLUME IXofrMi

WIRELESSENGINEER


Price 3/- By post 3/3Volume IX, Bound Complete, 42

By post £2 . 1 . 0Earlier Volumes can be supplied.Write for particulars to the

PublishersILIFFE & SONS LTD.

Dorset House. Tudor Street. London E.C.4

Engineer" when replying to advertisers.

W.E.16.

April 1933 iv THE WIRELESS ENGINEE

"THE WIRELESS WORLD"The weekly journal which keeps its readersthoroughly in touch with all that is happeningin wireless. It covers every wireless interestand appeals to every wireless amateur.

It is in every sense a wireless newspaper, forit records and reviews prevailing theory andcurrent practice in design, and also gives,week by week, broadcasting and othergeneral news.Special features are : Theoretical Articlesdiscussing new principles: Detailed TechnicalReviews of Commercial Radio Receivers :Designs for Experimental Receivers speciallyprepared to illustrate the practical applica.tion of new principles in valves, components,or circuit arrangements : The completeweekly programmes of all ForeignStations.

WirelessWorld

EVERY FRIDAY 4dSubscriptions:

Home & Canada, £1.1 .8, other CountriesAbroad, £1.3.10 per annum, poet tree.

ILIFFE & SONS LIMITED DORSET HOUSE TUDOR STREET LONDON E.C.4

Printed for the Publishers, ILIFFE k SONS LTD., Dorset Rouse, Tudor Street, London, E.C.4,by The Cornwall Press Ltd., 1-6, Paris Garden, Stamford Street, London, 8.E.1.

Sole Agents for Australasia --Gordon & Gotch (Australasia), Ltd.Sole Agents for South Africa-Central News Agency, Ltd.

THE WIRELESS ENGINEER · Regulations Controlling Receiving Licences-Broadcasting Stations of...

Documents

Transcript of THE WIRELESS ENGINEER · Regulations Controlling Receiving Licences-Broadcasting Stations of...