RCA Review 1937 Jan

RCA REVIEW
A Quarterly Journal of Radio Progress

Published in July, October, January and April of Each Year by
RCA INSTITUTES TECHNICAL PRESS
A Department of RCA Institutes, Inc.
75 Varick Street New York, N. Y.
VOLUME I January, 1937 NUM BER 3
CONTENTS
PAGE
Three Decades of Radio 5
DAVID SARNOFF
Partial Suppression of One Side Band in Television Reception 19

W. J. POCH AND D. W. EPSTEIN
Equipment Used in the Current RCA Television Field Tests 36

R. R. BEAL
Automatic Alarm 49
I. F. BYRNES AND H. B. MARTIN
Reproducing Equipment for Motion Picture Theatres 65

M. C. BATSEL AND C. N. REIFSTECK
Some Notes on Ultra High Frequency Propagation 76

H. H. BEVERAGE
Frequency Assignments for Television 88

E. W. ENGSTROM AND C. M. BURRILL
Behind the Scenes of Two Notable Broadcasts 94

GEORGE MCELRATH AND G. O. MILNE
Applications of Visual- Indicator Type Tubes 111

L. C. WALLER
Our Contributors 126
SUBSCRIPTION: $1.50 per Year. Foreign, $1.85. Single Copies 50 cents.

Special Rate to Public and College Libraries.
Copyright, 1937, by RCA Institutes, Inc.
Entered as second -class matter July 17, 1936, at the Post Office at New York, New York,
under the Act of March 3, 1679.
Printed in U.S.A.
www.americanradiohistory.com
BOARD OF EDITORS
Chairman
CHARLES J. PANNILL
President, RCA Institutes, Inc.
RALPH R. BEAL CHARLES W. HORN
Research Supervisor, Director of Research and Development,
Radio Corporation of America National Broadcastng Company
H. H. BEVERAGE WILLSON HURT
Chief Research Engineer, Assistant General Solicitor,
R.C.A. Communications, Inc. Radio Corporation of America
IRVING F. BYRNES DR. CHARLES B. JOLLIFFE
Chief Engineer, Engineer-in- Charge,
Radiomarine Corporation of America RCA Frequency Bureau
LEWIS M. CLEMENT FRANK E. MULLEN
Vice President in Charge of Manager, Department of Information,
Research and Engineering, Radio Corporation of America
RCA Manufacturing Company,
Victor Division CHARLES H. TAYLOR
Vice President in Charge of Engineering,
DR. ALFRED N. GOLDSMITH R.C.A. Communications. Inc.
Consulting Engineer,
Radio Corporation of America ARTHUR F. VAN DYCK
Engineer -in-Charge,
HARRY G. GROVER Radio Corporation of America
General Patent Attorney, License Laboratory
Radio Corporation of America
J. C. WARNER
O. B. HANSON Vice President,
Chief Engineer, RCA Manufacturing Company
National Broadcasting Company Radiotron Division
W. S. FITZPATRICK
Secretary, Board of Editors
Previously unpublished papers appearing in this book may be reprinted,

abstracted or abridged, provided credit is given to RCA REVIEW and to
the author, or authors, of the papers in question. Reference to the issue
date or number is desirable.
Permission to quote other papers should be obtained from the publica-
tions to which credited.
HAROLD H. BEVERAGE
Through the election of Harold H. Beverage as President of the
Institute of Radio Engineers the Institute gains a highly capable and
popular leader for 1937. The membership has had opportunity of
noting his energetic activities in committee work and other contribu-
tions in furtherance of I.R.E. aims and principles. I.R.E. honors were
also bestowed upon Mr. Beverage fourteen years ago when he was
awarded the Morris Liebmann Memorial Prize for outstanding contri-
butions in the development of a "wave" antenna system known as the
"Beverage Antenna ".
Other noteworthy achievements in his line of work are marked to
the credit of Mr. Beverage, who is an alumnus of the University of
Maine. He spent five years in the General Electric Company with Dr.
Alexanderson and during the World War was engaged in research work
at the Marconi high -power station at New Brunswick and at the Otter Cliffs
Naval radio station. Later he became Research Engineer for RCA, and in
1929 was appointed to his present post of Chief Research Engineer of
R.C.A. Communications, Inc.
EDWARD WASHBURN KELLOGG
At the banquet of the Society of Motion Picture Engineers, held at

Rochester, N. Y., October 14, 1936, a scroll was presented to Edward W.
Kellogg, whom the Journal Award Committee and the Board of Governors
of that Society had selected to receive the S.M.P.E. Award for 1936. Mr.
Kellogg's paper entitled "A Comparison of Variable- Density and Variable -
Width Systems," was designated as the most outstanding paper originally
published in the Society's Journal during the preceding year.
In the "Citation of the work of Edward Washburn Kellogg ", by Dr.
Alfred N. Goldsmith, Past-President of the Society, reference was made
to Mr. Kellogg's "career of industrial research and engineering, notable not
only for the importance of its accomplishments, most of which were either
directly or indirectly related to the sound motion picture art, but also for
its versatility."
In outlining his professional career, Dr. Goldsmith included mention of
Mr. Kellogg's several noteworthy accomplishments and closed the citation
with the remark: "I am doubly pleased that this honor should come to one
who combines, in such outstanding degree, original thinking, scientific
thoroughness, technical skill, and manly candor and courage."
Thirty years ago Mr. Kellogg was graduated as a civil engineer from
Princeton University, later studying mechanical and electrical engineering
at Cornell. From his first employment with a public utility company in
Chicago, he turned to instruction in electrical engineering at the Universities
of Missouri, Texas, and Ohio. Shortly after the entry of our nation into
the World War, he left the academic world and joined the Research Labora-
tory of the General Electric Company, at Schenectady, to assist in sub-
marine- detection work. Since the advent of sound motion pictures, Mr. Kel-
logg has been concerned with nearly every technical phase of the art, at
the laboratories of the RCA Manufacturing Company at Camden. He is now
in charge of the Advance Development Section in the Photophone Division
of that Company.
THREE DECADES OF RADIO*
BY
DAVID SARNOFF
President Radio Corporation of America
HISTORY in the making is always out of focus. But time and

memory have the saving trick of discarding the trivial and
emphasizing the significant. We need the perspective of the
years to judge the importance of past events with a true sense of
proportion. And -since our decimal system places an accidental em-
phasis upon numbers divisible by ten-the conclusion of a thirty -year
period seems to be an appropriate time to make a survey of the past.
It was my good fortune to go to work in the infant industry of
radio thirty years ago this Fall, and it has been my privilege to be
actively and continuously associated with it ever since. I mention this
fact, not because my remarks tonight are to be autobiographical -for
they are not-but simply to establish my excuse for being here; which
is that I have observed at first hand and participated in the phenomenal
growth of radio, almost from the day it began to be something more
than a scientific curiosity.
Marconi, it is true, had in December, 1901, achieved the first radio
communication across the Atlantic Ocean. It could hardly be called
a "message": just three tiny clicks -the letter "S" in Morse code
and so feeble in strength that one can well imagine the mingled hope
-
and fear written on Marconi's face as he stood on that bleak hillside
in Newfoundland. He handed the earphones to his assistant, Mr. G. S.
Kemp, and said, "Can you hear anything ?" Mr. Kemp did hear the
feeble clicks. After that first meager but momentous transmission
over the Atlantic, a whole year elapsed before a single coherent sen-
tence came across. It was not until the World War that anything
which approached dependable transatlantic wireless communication
was realized.
But we would have to go back many years before Marconi's time
were we to attempt to give credit to all the men of science who have
shared in the development of radio. To name but one, our own Ben-
jamin Franklin in the middle of the 18th Century made invaluable
* A lecture before The Franklin Institute, Philadelphia, Pa., November
18, 1936.
5
6 RCA REVIEW
contributions both to our knowledge of electricity and to the ability
of all science to express itself in clear and unmistakable terms. Radio
is a true descendant of this learned society's godfather and the first
great American scientist. It can trace its lineage to Franklin's meth-
ods and discoveries. Even his immortal kite had its direct successor
in the early experiments with antennas used in wireless telegraphy.
THE GROWTH OF RADIO

As we now look back upon the past thirty years of radio history,
they divide themselves into three fairly distinct periods.
The first was the marine decade. During this time radio was
utilized mainly in ship -to -ship and ship -to -shore communication. It
was a period during which wireless telegraphy was used almost exclu-
sively for maritime purposes. But it was a decade of active research
and experimentation, in which were sown the seeds of important
future devices and services. This is a statement that can be made
also of both the ten -year periods which followed. Every decade in
radio is an experimental one.
The second period was the war decade. In this period, I include
the early post -war years during which the tremendous advances made
by radio as an arm of the military and naval services were consoli-
dated for the uses of peace. This decade was marked by expansion
of radio from the field of telegraphy to that of telephony, an expan-
sion which made possible the achievements of the next period.
The third decade is marked preeminently by the development of
broadcasting. The National Broadcasting Company, the first organi-
zation to establish and operate a nationwide radio network, was
created just ten years ago this month.
Simply to recite the chronology -item by item, year by year,
decade by decade -of the technical progress of the radio art is to unroll
a fascinating tapestry which grows ever more brilliant in color, ever
more intricate in design. Such a recital, however, is merely a story
of "art for art's sake." It omits the very motif of radio's history
and of its future, the keynote in the approach to every problem that
confronts the industry.
For radio has never been an esoteric art. It has never dwelt in
an ivory tower. It is "of the people, by the people, for the people."
Its real history is the adjustment of revolutionary scientific instru-
mentalities to the changing needs and emotions of humanity. By the
success or failure of that adjustment radio must be judged, no matter
how much homage we may pay our scientists for their mastery of
galloping electrons captured in a tube.
THREE DECADES OF RADIO 7
THE MARINE DECADE

First, let us survey the marine decade of radio. Marconi's early
achievements had caught the public imagination, but they seemed
remote from the affairs of every -day life. Quietly and slowly, a few
shore stations were established and a number of larger vessels were
equipped with wireless apparatus. Then, in January, 1909, in a thick
fog off the Island of Nantucket, the transatlantic liner Republic was
-
rammed and sunk by another ship. Her wireless call for help
the CQD of that day brought vessels to her aid. All but six were
-
saved, and the Marconi station at Siasconset, on Nantucket Island,
relayed to the press the messages from the sinking ship and the rescue
ships as fast as the words came through the fog. Here was a drama
as ancient as the sea itself, with an unexpectedly happy ending, and,
for the first time, with all the world in breathless attendance. Wire-
less leaped from obscurity to fame. Congress passed the Radio Act of
vessel carrying more than fifty persons -

1910, prescribing radio equipment and an operator on every deep -sea
legislation which quickly
expanded the facilities and personnel of radio.
That act was amended and amplified in July, 1912. It now called
for a constant radio watch on shipboard, with two operators and an
independent auxiliary source of power. This action was the result of
the horrifying disaster to the Titanic in April of that year. Her SOS
had been heard over a wide area - -
the Carpathia rushed to the scene
of the tragedy and picked up 706 survivors but 1,517 perished, many
of whom might have been saved. A ship much nearer than the Car -
pathia did not hear the Titanic's call. She was equipped with wireless,
but her single operator was off duty and in bed when the call was sent
into the air.
It almost seemed as if the gods of progress had demanded human
sacrifies. An enormous price in human lives had to be paid before
marine radio attained its full importance in the public consciousness.
THE WAR DECADE
- -
Then to the accompaniment of an infinitely greater cost in
human suffering the war put radio into the armament of nations,
and through the achievements of the war decade, wireless became an
instrument of worldwide communication.
For the purposes of this survey, we are concerned, not with what
radio did in the war, but with what the war did to radio. However
far the world conflict fell short of achieving the idealistic outcome
demanded of it - to "make the world safe for democracy" - it did
8 RCA REVIEW
make radio an instrument of democracy. It took it out of the hands

of the few and gave it to the many.
The World War brought to fruition the important but obscure
discoveries of the preceding decade. High-vacuum tubes and regenera-
-
tive circuits ; improved detectors and amplifiers; continuous wave gen-
erators of much higher frequencies all became radio's stock-in- trade.
Nor was it by any means entirely a development of isolated, individual
inventions that was promoted by the war ; of even greater importance
was the effective coordination of all known radio devices into a single,
integrated structure.
Radio is the product not of one mind but of many; not of one dis-
covery, or a handful of related discoveries, but literally of thousands
of inventions originating in many diverse fields of science and of indus-
try. At the time the United States entered the World War the radio
industry was hopelessly deadlocked in a snarl of patent litigation,
actual or potential. A's invention was useful only in combination with
B's; B's invention was useless without A's; and A and B would not
or could not get together. Multiplied many times over, this was the
wartime situation.
The United States Navy Department ended the impasse by declar-
ing, in effect, a moratorium in the operation of the patent laws, as
applied to radio, for the period of the war. It assumed financial re-
sponsibility for all infringements of radio patents, and ordered Amer-
ican manufacturers to produce radio equipment embodying every
needed principle or device known at that time, regardless of patent
ownership. Thereupon, the radio industry threw away its crutches,
stood erect, and walked for the first time. Fortunately, the industry
has never since then suffered a return of the disease of patent
paralysis.
The bulky, stationary equipment of the marine decade became
transformed into the compact, portable field equipment of the war
decade. The very difficulty of quickly building a large technical per-
sonnel out of unskilled men forced the development of relatively fool-
proof apparatus. What had been complicated was simplified; what had
been difficult to service was made easy; what had been intermittent
and unreliable in operation became regular and dependable.
In historical importance, the one development that overshadowed all
- -
others was the gradual perfection of the radio telephone. Ability to
transmit the spoken as well as the recorded word the addition of
radio telephone, however, presented one serious difficulty -

voice to code gave to radio a new function and a new meaning. The
especially
serious in time of war. It offered none of the privacy of the ordinary
wire telephone. What one had to say was for everybody to hear.
Plenty of critics had pointed out this fatal limitation ever since the
days of the earliest experiments in radio-telephony.
THE BROADCASTING DECADE

But there were others who saw in this very limitation radio's great-
est opportunity. They envisaged the possibilities of a new system of
mass communication by radio, which would enable a single voice to
be heard at the same instant by countless millions everywhere. In other
words, they looked upon radio not only as a means of narrowcasting
but also as a method of broadcasting. And that broadcasting gave its
name to the third radio decade.
It was a method by which a liability was converted into an asset.
The very non -privacy of the radio telephone gave to it universality
of application. Had our telephone messages over the air been confined
to secret and private use, we might never have had any radio broad-
casting.
But while the war was on, the custodians of radio were too pre-
occupied with international communication and military equipment to
develop broadcasting. It was not until 1920, when the Harding -Cox
election returns were broadcast to an audience of several thousand
amateurs with home -made sets, that the idea took hold. This "stunt"
marked the birth of radio broadcasting.
Radio promptly proved that an old adage was out of date. We used
to say, "Necessity is the mother of invention." Radio demonstrated
that the opposite is true. Invention has become the mother of neces-
sity. Create some attractive new instrumentality, and even though
dreamed of such a thing - -

there has been no demand for it even though the public has never
millions of people suddenly discover that
they can't get along without it.
BROADCASTING IN AMERICA
Harding was elected and the word "normalcy" became part of the
everyday language. But it was a word that had no meaning in the
world of radio. Like derricks in a new oil field, broadcast transmitting
-
stations sprouted all over the United States in less than two years
after that first broadcast there were 500 of them. With no effective
allocation or regulation of wavelengths, the resulting interference and
confusion were indescribable. It was as though all those broadcasters
had conspired to build a new Tower of Babel.
It was not until 1927, when the Federal Radio Commission was cre-
ated by an Act of Congress, that the unique problems of broadcasting
10 RCA REVIEW
were first recognized by law. Provision was then made for assignment
of wavelengths and regulation of stations from a viewpoint of service
to the public. This purpose was emphasized by Congress in the Act
of 1934, when it placed the supervision of radio in the hands of the
Federal Communications Commission, and directed it to regulate radio
by the standards of "public interest, convenience and necessity."
In localities where broadcasting stations were established, there
followed an overwhelming demand on the part of the public for re-
ceiving sets. At first, the nature of the programs the broadcasters
were sending out was of little moment; the only thing that mattered
was to get hold of a set that would catch the signals of as many stations
as possible, and to build up an imposing -looking list of those caught.
Naturally, this phase passed quickly. The listening public soon
became conscious of the quality of broadcast programs. Then the sta-
tion owners, absorbed at first in mastering a new engineering tech-
nique, found that they had another big problem on their hands
double problem, of showmanship and economics. To build a good radio
-a
show cost money; the money came from advertisers; but advertisers
were interested primarily in attracting an audience, not in buying a
show. To the advertiser, the circulation cost of broadcast advertising
had to justify itself in some sort of relation to the circulation cost of
magazines and newspapers. No single station could supply a sufficiently
large audience to warrant the expense of providing first -class radio
entertainment.
It was a vicious circle. Until enough radio sets were in use to
make radio advertising profitable there would be no radio advertising,
and consequently no radio programs to make receiving sets worth
buying. That is why we organized the National Broadcasting Company
in November, 1926: to give the public programs on the air that would
encourage it to buy radio sets. These programs were expensive, but
by using telephone lines to connect a series of radio stations across
the continent, it was possible to broadcast the same program at once
to listeners in many states of the Union.
The demand for receiving sets was instantaneous and nationwide.
All radio stations profited by the new listeners, and their advertising
clients made better programs possible. Better programs sold more
radio sets; stations installed finer equipment; improved transmitters
meant better reception, further sales of sets, more advertising revenue,
and even additional radio chains.
This, in a nutshell, was the genesis of the nation's broadcasting
networks. They were dictated, not by the soaring imagination of radio
technicians, but by the everyday business requirements of business

men, who wanted value received for money spent.
The nation -wide network has enabled radio in the United States to
introduce to great masses of people the foremost artists in the world
of music and entertainment, and to do so with frequency and regu-
larity. Their services have been paid for by the goodwill of their
listeners-a goodwill voluntarily expressed through purchases of
radio -advertised products. The networks in turn have been enabled
to put on at their own expense a great number of sustaining programs
of high artistic and educational merit, and to broadcast important
news events, crop and weather reports, and the messages of our leaders
in public life.
To summarize: we enjoy in America radio broadcasting services
that are truly comprehensive in scope of programs, truly lavish in
talent, truly national in coverage. The rural listener is served on a
-
parity with the city dweller the voice of the President of the United
States may be heard at one time by the entire electorate. Yet after
we have said all this, we still have not named the most important, the
most vital aspect of our American broadcasting method. It is this:
that in the United States radio broadcasting is a truly free and demo-
cratic institution. No license is required and no fee is charged for
the privilege of owning a broadcast receiver and listening to the finest
radio programs in the world.
BROADCASTING IN EUROPE
This is a significant statement, for it cannot be made concerning

broadcasting in any nation in Europe today. During a recent visit to
Europe, I had the pleasure of hearing some beautiful concerts, broad-
cast from various cities on the continent. But I also listened to some
very poisonous nationalistic propaganda. If it is true that "music hath
charms to soothe the savage breast," it is also true that reckless
political propaganda has the power to make the human breast savage.
Under our system programs are supported by public goodwill vol-
untarily expressed. The American public can, and not infrequently
does, vote a program off the air by its collective and voluntary twist
of the dial. In those countries where radio programs are provided
through a governmental tax on receiving sets, and sending stations and
programs are under direct control of the government, broadcasting is
neither free nor democratic. Its programs are subject to bureaucratic
dictation and censorship. Radio has been made a slave of governmental
politics; a perquisite of the party in power; a mouthpiece for the
propaganda of autocrats. There, the public possesses no franchise -
12 RCA REVIEW
-
as it so emphatically does in this country to elect the programs it
desires to hear, and to reject those it does not.
BROADCASTING IN TUNE WITH ITS TIMES
point message service to a wide popular usage

mass communication
-
The expansion of radio from a specialized, commercial, point -to-
- from individual to
reflects many of the social characteristics of
the period during which it took place. It was the same period which
saw the greatest development of the automobile, the airplane, the mo-
tion picture, and innumerable time- and labor -saving devices in the
home. It was the period during which our daily lives were speeded ulf
and given unprecedented hours of leisure. Radio broadcasting filled a
newly -awakened need and was therefore attuned to the spirit of the
times. More than that, it drew the most distant places, the most for-
gotten lives, within the orbit of modern civilization.
It should be remembered that when international radio communica-
tion was perfected, the ship -to -ship and ship -to -shore services which
gave it birth were in no measure lessened ; on the contrary, they pro-
gressed consistently, both in magnitude and precision of operation.
In the youthful aviation industry, the ships of the air have also found
in radio an indispensable ally, quite as thoroughly as the ships of the
seven seas. And radio broadcasting, in turn, did nothing to slacken
the steady forward march of all the earlier radio services. Instead, it
gave to them a new advertising and a new impetus. Radio at all times
consolidated its gains and abandoned none of the territory it had con-
quered.
Only a week ago - on Armistice Day - it was my privilege to
take part in a four -way radio telephone conversation which illustrated
the present state of radio's technical development. The participants
-
were Senatore Guglielmo Marconi, two distinguished gentlemen from
Europe now visiting in this country Monsieur Robert Jardillier,
French Minister of Communications, and Monsieur Maurice Rambert,
-
President of the International Broadcasting Union and myself.
Marconi was on his yacht in the Mediterranean off the coast of Italy,
our European visitors were travelling in two separate airplanes be-
tween Buffalo and Washington, D. C., and I was seated at my desk in
the RCA Building in New York. We chatted together for ten minutes.
Each of us could hear all the others perfectly. This unique experiment
was broadcast by the National Broadcasting Company and heard by
listeners in America, Europe and in other parts of the world. The
land, the sea and the sky were linked across thousands of miles by radio.
RCA Is ORGANIZED
Within its own organic structure, radio did as other major Amer-
ican industries have done : it moved out of individual hands into those
of the organized business group. It turned to the investing public for
its capital funds. Almost midway in the thirty -year period we are con-
- -
sidering in October, 1919 the company which I have the honor
to represent, the Radio Corporation of America, came into existence.
The history of this company is so inseparably linked with that of the
entire industry that any review of this kind, however brief, would be
incomplete without reference to it.
The war made obvious the vital importance of radio to the future
made in Washington to nationalize the industry

neither then nor at any time since were in harmony
-
development of international communications. Abortive efforts were
with
efforts which
the American
temper and tradition. Congress rejected the idea. Then, on the recom-
mendation of the Navy Department, the war -time custodian of radio,
the Radio Corporation of America was organized.
time -
Two objectives were sought and - within a comparatively brief
attained first, to insure American ownership and control of
:
radio facilities in this country; second, to perpetuate the benefits to

the public of an integrated patent structure, originally achieved purely
as a war emergency measure. Definite and controlling considerations
of public policy have guided the conduct of the Radio Corporation ever
since its inception. It is today the largest radio organization in the
world, and the only one actively engaged in every field touched upon
by the radio industry, and in no other.
I think it has become fairly well established that the great modern
industrial organization bridges the gap between pure and applied
science far more effectively than does the individualistic, "one -man"
type of organization. It reduces the temptation to sacrifice the value
of future services for the sake of a quick profit. It substitutes the long-
range for the short -range viewpoint. It can afford to approach its
problems from the angle of public interest rather than from self-serv-
ing opportunism, and it does so, not by reason of any fanatical moral
attitude, but because the most enlightened self- interest dictates such
an approach.
In the domain of applied science, the trend from individual to col-
lective industrial research has been of unquestionable benefit to the
public- in the quantity, quality, and rapidity with which useful new
products and services have been brought forth.
14 RCA REVIEW
THE FUTURE OF RADIO
This brings me to the second part of this discussion -the future of

radio. While its progress during the past thirty years has been im-
pressive. I look for even greater accomplishments during the next ten.
As our knowledge and employment of ultra -short waves increases
progress will be speeded up. The most conspicuous of these advances
during the next ten years will be the addition of sight to sound in the
service of radio.
FACSIMILE
Radio transmission of pictures and reproductions of printed or
written material has been an accomplished fact for several years. It
is now in daily service between Europe and America. The broadcasting
of a facsimile newspaper into every business office and home
hourly installments if desired - - in half -
is perfectly feasible. The establish-
ment of such a service is now an economic rather than a technical
problem. I feel reasonably certain that American ingenuity will pres-
ently discover some way to make this potential national service avail-
able to the public.
On June tenth of this year, Chancellor Harry Woodburn Chase of
New York University and Vice -President W. Chattin Wetherill of The
Franklin Institute exchanged pictures and greetings by radio facsimile.
Their messages inaugurated the public service of RCA's ultra -short
wave radio circuit connecting New York and Philadelphia, with two
automatic radio relay stations at intermediate points.
This event was a new milestone of man's progress in conquering
time and space. It represented not only a marked advance in the speed
with which facsimile messages are transmitted, but a new achievement
in multiple transmission. The new circuit made possible simultaneous
employment, on the same radio wave, of two facsimile channels, two
-
automatic typewriter channels, and a telegraph channel five channels
in all. The association of New York University and The Franklin
Institute in the opening ceremonies was a happy and appropriate one,
since these are the two institutions which, a century ago, first publicly
recognized the genius of Samuel F. B. Morse and the importance of
his electric telegraph.
TELEVISION
When it comes to transmitting and receiving the instantaneous,

flowing image of moving objects the problems are enormously in-
creased. These are the technical problems of television.
The Radio Corporation of America has just concluded an elaborate
series of preliminary field tests to bring television from the laboratory

into the sphere of practical use. These tests began only on June 29th
of this year. That date marked the beginning in this country of organ-
ized television experiments between a regular transmitting station and
a number of homes. Since then we have advanced, and we are contin-
uing to advance, simultaneously along the three broad fronts of tele-
vision development : research and engineering, which must point the
road to effective transmission and reception, and then translate this
progress into practical receiving sets for the home ; program produc-
tion tests to develop a television service suitable for network syndica-
tion; and economic studies to create a sound business basis that will
support so costly an enterprise as television programs are bound to be.
On all these fronts our work has made definite progress and has
brought us nearer the desired goal.
We have been transmitting experimental programs from our tele-
vision station on top of the Empire State Building in New York City,
by remote control from the NBC television studios in the RCA Build-
ing. We have observed and measured these transmissions through a
number of experimental receivers installed in the homes of RCA engi-
neers and located in the metropolitan area and adjacent suburbs. The
results thus far have been encouraging and instructive. As we antici-
pated, many requirements that must be met by a commercial device
have been made clear by these tests.
We have successfully transmitted through the air, motion pictures
as well as living talent before the televisor. The distance over which
these television programs have been received has exceeded our imme-
diate expectations. In one favorable location, we have consistently re-
ceived transmissions as far as 45 miles from the television studios.
The tests have been highly instructive. We have learned much that
is new about the behavior of ultra-short waves and how to handle
them. We know more about interferences, most of which are man -made
and susceptible of elimination. The difficulties of making apparatus
function outside of the laboratory have been surmounted. We have
confirmed the soundness of the technical fundamentals of our system.
Theory has been put into practice, and the experienced gained through
these tests enables us to chart the needs of a practical television service.
In our present field tests we are using 343 -line definition. In cooper-
ation with the industry, we have recommended to the Federal Com-
munications Commission the adoption of 441 -line definition as a stand-
ard for commercial operation. Our New York transmitter will be
rebuilt to conform to the recommended standards. That also means
building receivers to conform to the new standards of the transmitter.
16 RCA REVIEW
The necessity of synchronizing transmitting and receiving equipment
carries with it serious responsibilities. On the one hand, standards
cannot be frozen prematurely or progress would be prevented; on the
other hand, frequently changing standards would mean rapid obso-
lescence of television equipment.
Another major problem in television is that of network syndication.
Our present facilities for distribution of sound broadcasting cover the
vast area of the United States and serve its 128,000,000 people. Similar
coverage for television programs, in the present state of the television
art, would require a multiplicity of transmitters and network inter-
connection by wire or radio facilities still to be developed.
From the standpoint of research, laboratory development, and tech -
nical demonstration, television progress in the United States continues
to give us unquestioned primacy. We lead in research which is daily
extending the radio horizon, and in technical developments which have
made possible a transmitting and receiving system that meets the
highest standards thus far obtainable.
The chief distinction between television in this country and abroad
is the distinction between experimental public services undertaken
under government subsidy, and commercial development undertaken by
the free initiative, enterprise, and capital of those who have pioneered
the art in the United States.
TELEVISION PROGRAMS
The National Broadcasting Company is making an intensive study

of television studio technique. During the next few months we will
expand the engineering field tests into a series of dress rehearsals of
various types of programs. Ultimately television will create its own
individual art form -a fresh and unique world of illusion. It will,
to be sure, borrow from the older arts of stage, motion picture, and
sound broadcasting. It will supplement them all, and supersede none.
The character of the new medium, with its inherent advantages and
limitations, provides new scope for the imagination.
In the world of creative and expressive art the hardest question
which television propounds is that of supplying talent. It suggests a
reversal of the usual comparison between American methods of pro-
duction and of distribution. Industry, it is said, has learned how to pro-
glamorous of industry's children - -

duce efficiently and cheaply, but distributes its wares inefficiently and
wastefully. Here is television, on the other hand youngest and most
preparing to deliver its programs
with the speed of light into the centre of every home. Perfect distribu-
tion! But television's problem of program production is a different
matter. It is still unsolved, and much work must be done before the
solution has been achieved.
Television broadcasting, even more than sound broadcasting, will
-
be the great consumer of art. It will constantly demand more and bet-
ter writers, musicians, actors, and scenic designers new thoughts,
new words, new songs, new faces, new backgrounds. Unlike a play on
the stage or a motion picture which may run for a year, the television
program, once it has been shown to a national audience, is on the scrap-
heap. It is finished. Television will call for a whole new generation of
artists. It should help materially to solve the unemployment problem.
The way things look today, it is not improbable that in a few more
years a man with three sons may train one for business, one for gov-
ernment service, and one to be an artist. Perhaps this thought comes
to my mind because I have three sons still to be trained for a useful
life.
We have lately heard in our own homes the voices of the Pres-
idential candidates. Political campaigns will take on added interest
when people can see as well as hear the speakers, with television sets
in the home. Each Presidential year since radio began to participate
in the campaigns, the number of voters has increased by many millions.
Whereas 27,000,000 people voted in the election of 1920, the vote in
1936 was 45,000,000. Television will increase the usefulness of radio
in the cause of popular government.
While the problems of television are formidable, I firmly believe
they will be solved. With the establishment of a television service to
the public which will supplement and not supplant the present service
of broadcasting, a new industry will have been created.
RADIO CONTRIBUTES TO OTHER FIELDS
I have touched upon the major developments of radio and its pros-
pects for further development in its recognized fields of service. I
shall not attempt to review those collateral contributions with which
radio science has helped to quicken the pulse of other industries. It
is a fact worthy of mention, however, that radio tube technique and
photo -electric controls not only are being applied more and more in
the engineering branches of industry, but are welcome adjuncts to the
tool -kit of pure science. The cyclotron which disrupts the atom -
come true -
suggestion that the alchemist's dream of transmutation may some day
utilizes magnets originally devised for radio arc trans-
mitters. Astronomers and meteorologists, motion picture studios and
electric power plants, doctors of medicine and electric welders, all have
been beneficiaries of the radio research laboratories.
18 RCA REVIEW
Radio has also pointed our attention upward. Until our own gen-
globe -
eration the wealth of the world came from below the surface of our
from the mines and waters and fertile soils. It is only in the
last thirty years that humanity has raised its eyes from the ground
and commenced to look upward for new wealth - into the air, into
the stratosphere. It is only a small beginning we have made, and it is
fascinating to speculate upon the potential resources that still lie
untouched in ultra -short waves, in sun- energy, and in the stratospheric
lines. Once we faced the frontiers of geography. Today we face the
infinite frontiers of science.
RADIO, A SCIENCE AND AN INDUSTRY
If science has taught us anything it is to be humble; to realize that

our knowledge weighs lightly in the scale against our ignorance.
Toward the close of his long and illustrious career, Sir Isaac Newton
said:
"I seem to have been only like a boy playing on the seashore and
diverting myself, in now and then finding a smoother pebble or a pret-
tier shell than ordinary, whilst the great ocean of truth lay all undis-
covered before me."
That statement reflects the view of the true scientist and phil-
osopher.
Pure science acknowledges only truth for her master, and no matter
where that master leads, science must follow. Industry translates the
truth revealed by science into products and services for humanity. For
enlightened industry is the servant of humanity, even as science is the
servant of truth.
Radio is both a science and an industry; it exists to serve both
truth and humanity. As long as it retains its independence it will con-
tinue to do so. The freedom of radio is inseparable from the freedom
of speech, freedom of the press, freedom of education and of worship.
Here in Philadelphia, you treasure one of America's earliest instru-
ments of broadcasting. Upon it are inscribed these words:
"Proclaim liberty throughout all the land, unto all the inhabitants
thereof."
No other broadcasting instrument ever sounded forth a more
momentous message for the human ear to hear, and the human heart
to cherish. May the message of America's Liberty Bell ever remain
the inspiration of radio.
PARTIAL SUPPRESSION OF ONE SIDE BAND
IN TELEVISON RECEPTION
BY
W. J. POCH and D. W. EPSTEIN
RCA Manufacturing Company, Inc., Camden, New Jersey
INTRODUCTION
EARLY television development followed the precedents estab-

lished in sound broadcasting. A radio carrier was amplitude
modulated by the video signals resulting from scanning and
the transmission included both side bands. In the receivers the selec-
tivity or bandwidth was made such as to pass both upper and lower
side bands when the carrier was modulated with the highest desired
modulating frequency. Progress in television development has been
marked by a continual increase in the number of scanning lines and
requiring, in turn, increases in the communication band. This race,
as it became, between the terminal apparatus -ability to increase reso-
lution, i.e., number of lines -and the communication portions of the
system-ability to increase band width in the amplifiers and circuits
exhibiting selectivity characteristics -first found one element in the
lead and then the other. At times when the receiver band -pass char-
acteristics were more limiting than other elements, it was early deter-
mined experimentally that a better picture was obtained when the
receiver was slightly detuned. Thus, by detuning, the picture carrier
was placed near one edge of the selectivity characteristic.
Later when this condition was more thoroughly appreciated, an
analysis was made of its importance and usefulness. Suppose we
deliberately design a receiver so the resulting intermediate frequency
is placed near one edge of the intermediate -frequency circuit selectivity
characteristic and so that carrier and all of one side band but only a
small portion of the other side band is accepted, with the over-all selec-
tivity being insufficient to remove entirely the second side band. We
shall term this a selective side -band receiver. An immediate advantage
is that we nearly double the modulation frequency range that the
receiver will pass. This is of great importance where the band width
for one side -band approaches the limits of circuit and tubes and where
it is inadvisable to reduce gain or selectivity.
19
20 RCA REVIEW
It is a well -known fact that for circuits passing broad bands, the
gain per stage is inversely proportional to the band width. This means
that n intermediate -frequency amplifiers having the same number of
stages, one for selective side -band and the other for double side -band
operation, will have a difference in gain of 2n, where n is the number
of stages. For six stages this means a difference in gain of 64 to 1.
If the gain per stage of the selective side -band receivers were 8, the
double side -band receiver must have three additional stages to have
the same over-all gain.
Before taking the important step of making this change in the
receivers, it was thought necessary to make a further investigation
of this problem. An experimental transmitter and receiver system,
whose condition of operation could be controlled and upon which
measurements could be easily made, was set up. This apparatus was
arranged so that it would be used either as a double or a selective
side-band system with a simple and quick changeover arrangement.
It was also arranged so that part of the suppression of one side band
was done in the transmitter, to determine whether this would intro-
duce any special difficulties. The data taken on this system were also
verified by a mathematical investigation.
Because of the profound influence selective side -band suppression is
likely to have on practical systems of television, it is considered of
interest and importance to describe these early tests and to outline
the mathematical verification.
APPARATUS USED IN EXPERIMENTAL WORK

Fig. 1 is a block diagram of the transmitter and receiver equip-
ment. The only adjustment necessary for changing from double side -
band to selective side -band operation was to shift the master oscillator
frequency from 4.25 to 4 megacycles. Suppose that the master oscil-
lator was generating 4.25 megacycles, the condition necessary for
normal double side -band operation. The modulator then delivered an
8.5- megacycle modulated carrier at the input of the transmitter inter-
mediate-frequency amplifier. Care had been taken to make the mod-
ulation amplifier and the modulator itself with a fidelity characteristic
flat to 1000 kilocycles. The output of the intermediate-frequency am-
plifier, still an 8.5- megacycle carrier but with side bands trimmed to
500 kilocycles on each side, was used to modulate another oscillator
operating at 63.5 megacycles. Only the resulting lower side band was
used. This was a carrier at 55 megacycles with side bands extending
to 500 kilocycles on both sides. The receiver was also tuned in such a
PARTIAL SUPPRESSION OF ONE SIDE BAND 21
way that the incoming carrier was located in the center of the receiver
selectivity characteristic, so that again both side bands were treated
alike. The second detector and video frequency amplifier which were
adjusted to have a fidelity characteristic good to 1000 kilocycles,
brought the modulated signal to the grid of the `{Kinescope. *"
Now suppose that the frequency of the master oscillator was shifted
from 4.25 to 4 megacycles, the condition for selective side -band oper-
ation. The carrier output of the modulator doubler was now at 8 mega-
cycles which brought the carrier to one edge of the transmitter inter-
mediate-frequency pass-band characteristic. The output of this ampli-
KINESCOPE
TRANSMITTER RECEIVER
425 MC. FOR DOUBLE 53.0 MC CARRER FOR
SIDE BAND DOUBLR ODE SAND
4.00 MC. FOR SELECTIVE 55.5 MC. CA .BIER OR
SIDE BAND SELECTIVE SIDE BAND
MAS
MASTER
CDVIL- CARRIER AT 8.5 MC. FOR CARRIER AT 715 MC:FOR
LATOR DOUBLE SIDE BAND
VIrI.I(I1. P
DOUBLE SIDE BAND
CARRIER AT 8.0 MC. FOR CARRI RAT 6.65 MC FOR
SELECTIVE SIDE BAND SELECTIVE SIDE BANO
L F. AMPLIFIER IF AMPLIFIER
MODULATOR 2ND IST 2NU
PASS BAND PASS BAND
DOUBLER MODULADOR DETECTOR DETECTOR
8 MC. TO 9MC. 6.65MCTO 765MC
MODULATION CSCILLATOR OSCILLATOR LE AMPLIFIER 2ND

AMPLIFIER 63.5 MC. 62.15 MC AIS MC. DETECTOR
AF
0 56 MC
AMPLIFIER
501.100 LOUD
TRANS-
MIT TER SPEAKER
Fig. 1 -Block diagram of the transmitter and receiver equipment.
fier was still a carrier at 8 megacycles but with the upper side band
extending to 1000 kilocycles and the lower side band greatly atten-
uated, except at low modulation frequencies. Similarly, in the receiver
whose tuning adjustment had not been altered, the carrier was also
moved to one edge of the selectivity characteristic, causing one side
band to be reduced still more. At low video frequencies normal demod-
ulation of a carrier and both side bands occurred at the second de-
tector. At the higher video frequencies only the carrier and one side -
band were present. In between was a range of frequencies in which one
side -band was being rapidly attenuated. This problem of detection will
be discussed in more detail later.
The sound transmitter and the sound channel of the receiver which
* Trade Mark Registered U. S. Patent Office.
22 RCA REVIEW
had a sharp selectivity characteristic compared with that of the picture

channel, were used to check the tuning of the receiver. The frequency
spacing between picture and sound transmitters was checked by tuning
a broadcast receiver to the difference frequency.
SELECTIVITY MEASUREMENTS
Figs. 2 and 3 show the selectivity characteristics of the transmitter

and receiver intermediate frequency amplifiers. These were taken in
FREQUENCY IN KILOCYCLES
]tl09 0900 0900 0.00 9000 9200 0400 9600 moo
5'
J
s
0
F. 5E L`STI nr OF
sv E BAND Su o7R05505 55 TE
Fig. 2-IF selectivity of side band suppressor system.
the usual manner with a calibrated oscillator and vacuum tube volt-
meter. Rejector circuits were used in both of these amplifiers which
increased the attenuation of the unwanted side band. In the receiver,
the rejector circuits were tuned to the sound intermediate frequency
to prevent interference from the sound transmitter in the picture
channel. Note that these curves show a generous 1000- kilocycle band
width. An over -all selectivity measurement was not made but should
correspond to the product of the two curves shown since the radio -
frequency output circuit at the transmitter and the input system of the
receiver did not have sufficient selectivity to affect the other curves.
MEASUREMENT OF FIDELITY AND PHASE CHARACTERISTICS
Measurements of fidelity and phase characteristics were made in

the video frequency range between 10 and 1000 kilocycles since no
effects due to suppressing one side band were found below 10 kilocycles.
The fidelity characteristics were taken with a beat frequency oscillator
and vacuum tube voltmeter having an upper frequency limit of 1000
kilocycles. A cathode -ray oscillograph also having a 1000- kilocycle
frequency range was used in conjunction with the beat frequency oscil-
FREQUENCY IN KILOCYCLES
Iwo .4000 boo 6200 6400 4b3 4000 1000 770o Iwo Ism 1!W e00D YID $400
s
7
Z
) V
w
2 >
W
C
i
RE CEIV tR_ F. ELE 'TI" ITY
Fig. 3- Receiver IF selectivity.

lator to obtain the phase characteristics. This was done by the familiar
method of connecting the output of the beat oscillator to the horizontal
deflecting circuit of the oscillograph and the output voltage of the
circuit being tested to the vertical deflecting circuit. From measure-
ments of the resulting ellipse the phase angle between the input and
output voltages can be calculated. To avoid wave shape errors, mod-
ulation on the transmitter was kept below about 25 per cent. The most
desirable characteristics are, naturally, to have a flat frequency re-
sponse and to have the phase shift proportional to frequency.
Fig. 4 shows the resulting fidelity curves. The over -all curve for
24
Z 40
C
r0
70
60
50
30
20
;/
11\'i
1111"
iiiiñ1111ii6
N
1111C11111111.11
RCA REVIEW
double side -band operation shows the expected loss in response above
500 kilocycles due to trimming of the side bands. The over -all curve
for selective side -band operation is perhaps better than might be an-
i00
\1\!a/_M9
ii0i11615©
11i
:::11:::=C:s:11:::::
o 2
fidelity of receiver.
.3 .4
FREQUENCY
.5 .6
IN MEGACYCLES
.7 10
Fig. 4-Measured fidelity. 1. Over-all fidelity double side-band operation.

2. Over -all fidelity selective side-band operation. 3. Picture amplifier
ticipated. Since at low modulation frequencies both side bands are

present at the second detector and at the higher frequencies only one,
we would expect the fidelity curve to drop down approximately 50 per
cent at a fairly low frequency and then continue to about double the
frequency limit for double side -band operation before dropping down
2 3 .4 .5 6
FREQUENCY 7N. MEGACYCLES
Fig. 5-Measured phase delay. 1. Over -all phase delay using double side -
band transmitter. 2. Over -all phase delay using selective side -band
transmitter. 3. Receiver picture amplifier alone.
again. This effect, however, is also dependent upon the exact position
of the carrier on the edge of the selectivity curve. The farther down
we put the carrier on the side of the curve, the less will be the first dip
downward in the response curve. It is possible, of course, to carry

this procedure so far that the high- frequency response will actually
be greater than the low- frequency response. In this particular case,
the over-all selectivity curve of the system and the position of the car-
rier relative to it were such that the response curve shown in Fig. 4 was
produced.
Fig. 5 shows the phase delay characteristics corresponding to the
previous frequency characteristics. The phase shift, as measured by
I,
1111111111111111111111M111
MIIIIIIIIMI1111111111111111
Erulummumaim
mermimummonm
.I.....
....'1
I,..,
.!I,
180' 90 1111111111111SINIMINI
160° 19111111111111/11111M11111
140' 70
I20' 60
100° 50 11111=1111=11111111
80° 40
li
40'
30
20
EIM.MMIINIll
20° IO 1111111,MIMMIIIM"11
1íll..M..
0° 0
1'l.®,.....
111M111111,I111111M.N
ESIIIIIIM11111111111111111111
Fig. 6- Selectivity and phase of intermediate -frequency transformer.
the ellipse method, was converted to phase delay by the following

equation:
phase angle
Phase delay =
360° X frequency
This expression gives the actual time required for a cycle of a given
frequency to go between input and output of the amplifier system being
measured. Obviously, if this time is the same for all video -frequencies,
there will be no phase distortion. This condition for a constant phase
delay is equivalent to the previously mentioned condition for having
the phase shift proportional to frequency. The variation in phase delay
over the video frequency band may be taken as a measure of the phase
distortion. Note that this variation is greatest for selective side -band
operation but that the difference in variation between double side -band
and selective side -band operation is not very great.
26 RCA REVIEW
Fidelity and delay characteristics for conditions similar to those in

this experimental work were also calculated and are given in the next
section.
CALCULATION OF FIDELITY AND DELAY CHARACTERISTICS
.
In order to calculate these characteristics it is necessary to know
the over -all selectivity and phase characteristics of the system. Ac-
cordingly, the selectivity and phase curves for one of the coupling
transformers used in the intermediate -frequency system were calcu-
esin
lated. These curves are shown in Fig. 6. Assuming no radio -frequency
Fig.
....
e
8.m
7Zo' 90
!!!
IIE :o
"
.......
....1
.. .. ..
rl.
,1I.
sso' -ro
5
n
60
40
E,
3U
eo° ta
o'
"!9°EIE
Ise
M
!!!
o
1.,,
IM/I0
..1 E
SIM
Il.,I
11111111.
121111
FREQUENCYIN MEGACYCLES
7- Selectivity and phase characteristics of the intermediate -

frequency amplifier.
selectivity and that all the coupling circuits are identical, the selectivity
and phase characteristics of the receiver or receiver and transmitter
may be deducted from Fig. 6 by merely raising the ordinates of the
selectivity curve to the nth power and multiplying the ordinates of the
phase curve by n, where n is the number of identical coupling circuits
used in the system. Fig. 7 shows the selectivity and phase characteris-
tics for four intermediate -frequency coupling circuits is cascade as
obtained from Fig. 6. These curves may be taken as the selectivity and
phase characteristics of the receiver or transmitter alone.
Before proceeding with the calculation of the fidelity and delay
characteristics as obtained from the selectivity and phase curves of the
intermediate frequency, it is worth reviewing, for the sake of clarity,

the action of the intermediate frequency and second dectector on a car-
rier modulated by a single frequency. For this purpose, consider the
simple case of a receiver in which the input to the intermediate-fre -
quency amplifier consists of a carrier of frequency fo modulated with
the video frequency fl. The input to the intermediate -frequency ampli-
fier may then be written as
e = E cos wot -}-

2 cos (wo + wi)t -}-
2 cos (wo - wi)t (1)
where w = 24, E is the amplitude of the unmodulated carrier, and m

is the percentage of modulation. After passing through the selective
circuits of the intermediate -frequency amplifier, the voltage output
becomes
¡- mAu
y = E[Ae cos (wot + cßo) + cos {(wo -}- wi)t + O.}
mAL
+ cos {(wo - wi)t + 4 L }
J
(2)
where A, Au, and AL are the amplitude ratios of output to input of

the intermediate -frequency amplifier for the frequencies fo, fo -I- f1, and
fo - fi, respectively, and cpo, cpu, and cpL are the phase shifts introduced
by the selective circuits of the intermediate frequency for the respec-
tive frequencies.
Equation (2) gives the input to the second detector. The output of
the second detector may be obtained by first determining the envelope
of the modulated carrier given by (2). To determine the envelope at
the input to the second detector it is but necessary to transform (2)
into the form
u = Ve cos (wot -}- (1,)
and then gives the form of the envelope. Performing this transforma-
tion there results that
r m2
- 0L}
Ve = E[A,2 + {Au2 + AL2 + 2Av.AL cos (2wit -}- chu.
L 4
1/2
-E- mAe Au cos (wit

{ + u - 4o) + A L cos (wit + o - 0L) }
J
(3)
Equation (3) thus gives the shape of the carrier envelope at the input
to the second detector.
28 RCA REVIEW
It is worth noting, in passing, that the phase of the modulated car-

rier, cp, is given by
A, sin doo + -sin

mAq
2
(wit + çu) - mAL sin (wit - ckL)
tan - 2
(4)
Ac cos Oa -{-
u
cos (wit + Ou) -i- L cos (wit - (AL)
and that in general when one of the side bands is partially suppressed
tan 4) is a function of time so that some phase modulation exists.
Referring to (3) it may be seen that for low percentages of modula-

tion the output of an n -law detector* is given by
nm
Ve" = E" Ac" -I-.
2
Acn-1 { Au cos (wit + ch. - 'to)
+ A L cos (colt -I-4o - (WI _ (5)
Equation (5) shows that for small percentages of modulation the de-
tector will reproduce only the original modulation frequencies. Assum-
ing, therefore, a small percentage of modulation one may, with the aid
of (5), calculate fidelity and delay characteristics from the selectivity
and phase curves of the intermediate frequency. Thus the output of
the second detector at the frequency fi is, by (5), proportional to
Au cos (271út + l u - Oo) + AL cos (27rfit -f- (ko -'¢LJ
where Au and AL are the ratios of output to input amplitudes for the
frequencies fo -i- f1 and fo -f1, respectively, as obtained from a selec-
tivity curve such as that shown in Fig. 7 and where cpo is the phase of
the carrier and cpu and 4)L are the phases for the frequencies fo -f fi and
fo-f1, respectively, as obtained from the phase curve of Fig. 7. Adding
the above two terms there results that
Au'cos (27fit + Ou - o) + AL cos (27fit -1- cho - L)
= Vfi cos (2vfit - e)
where,
Vfi = VAu2 + AL2 -}- 2A.uAL cos (Y'u + (kL)

and,
tan =
Au sin (4 - 0o) + AL sin (ço (AL)
O
Au cos - 0o) + AL cos - (cho L)
* n = 1 for a linear detector, n = 2 for a square detector, etc.
The fidelity and phase characteristics may then be calculated by using

(6) and (7) and the curves of Fig. 7. As was mentioned previously,
the phase delay in seconds at any frequency f is O/274 where O is in
radians and f in cycles per second.
Fidelity and delay characteristics corresponding to various inter-
mediate-frequency carrier frequencies were calculated with the aid
of (6) and (7) and are shown in Fig. 8. These fidelity curves corre-
spond to those of a receiver with no radio -frequency detector or video
CURVES (1) AND (I)'-FIDELITY AND DELAY MR 6.9 M.C. CARRIER

.. ..
(2) (2Y- . 4
(31 - (3)'- 6.7
(l) (4)'- -
t.a: NNIMMIMMINMENI
V ^ ' " 5.6 "
MI
EgliEMEEMEliall
h
11111111=:!11211111111MlIO((IL
11111_111101
111,
&\
Pr\_a
.
.
Am
PI lóiLI=
SIMIIIIRIVIIIIEB
MNEMIIIIIM
',(4)
\\
11.V'-nsE
, .
., =,o
1 to
,
0 0
10
° WIEN11211 .6
t FREQUENCY IN MEGACYCLES
I
Fig. 8- Fidelity and delay characteristics.
distortion and with the selectivity shown in Fig. 7. Curves (1), (2),
(3), and (4) of Fig. 8 correspond to the fidelity of the receiver for
intermediate-frequency carrier frequencies 6.9, 6.4, 6.2, and 5.6 mega-
cycles, respectively.
In obtaining Fig. 8 it is assumed that the transmitter passes both
side bands. If the transmitter partially suppresses one side band, it
may still be assumed that the transmitter passes both side bands but
that the receiver selectivity has been increased. Fig. 9 gives the selec-
tivity and phase characteristics corresponding to those of receiver and
transmitter. The fidelity characteristics for such a receiver calculated
from Fig. 9 are shown in Fig. 10. Thus Fig. 10 gives the fidelity of a
receiver having the selectivity shown in Fig. 1 when the transmitter
has the same selectivity. Curves (1) and (2) of Fig. 10 correspond to a
carrier at 6.4 and 6.25 megacycles, respectively.
Referring to Fig. 8 it is to be seen that: Curve (1), corresponding
30
Fig.
,
;
In
,AT
.e .
!
Ir!N
fR
-
RI
320e
160° 10
0.
pm
80
10
60
s0
40
30
20
IIIMIIMM
./I.
rI,"
/o
..I
I1
,1..'.
1"/ M.II
MEN
RCA REVIEW
to double side -band reception, is practically flat in the frequency range

0 to 0.6 megacycles. No phase distortion is present in the frequency
range 0 to 0.8 megacycles. Curve (2), obtained with the carrier on one
I1Il. MIN
III
8
P2
FgEWENCY
1N
9-Selectivity and phase equivalent to that of receiver and transmitter.

Go
9
MEGACYCLES
10
edge of the selectivity curve, emphasizes the lower frequencies more

than the higher. No phase distortion is present in the frequency range
0 to 1.3 megacycles. Curve (3) is the best fidelity characteristic and is
100%
90
(1)
80
10
60
(z)
50 - --'
40
30
_CO
10
-
0
0 __.2 _4 ____ .6 -.8 -10 -1 2-1.4

FREQUENCY N MEGACYCLES
1 6
Fig. 10-Curve (1) fidelity for 6.4- megacycle carrier, obtained from figure.
Curve (2) fidelity for 6.25- megacycle carrier, obtained from figure.
obtained when the carrier (with two side bands) is located halfway
down the selectivity curve of Fig. 1. The delay characteristic given by
curve (3) of Fig. 8 is flat in the frequency range 0 to 1.5 megacycles,

and hence there is no phase distortion, as all frequencies are delayed
by the same time. The time of delay is seen to be about 1.25 micro-
seconds. The characteristic is practically flat in the frequency range
0 to 1.3 megacycles. Curve (4) shows the low frequencies badly atten-
uated. Phase distortion exists for low and high frequencies.
The curves of Fig. 8, therefore, show that with the assumptions
made and with the transmitter passing both side bands, it is best to
tune in about halfway down on the selectivity curve in order to obtain
the optimum fidelity curve with a given selectivity curve of the re-
ceiver.
The curves of Fig. 10 similarly show that with the transmitter par-
tially suppressing one side band it is best to attenuate the carrier at
the transmitter and receiver so that the total attenuation is down to
50 per cent. Thus if the transmitter attenuates the carrier to 71 per
cent and the receiver to 71 per cent of this the result is about 50 per
cent. However, as is shown by curve (4) of Fig. 8, the carrier should
not be attenuated by transmitter or receiver any further than to 30
per cent, for at 30 per cent and below phase distortion appears. With
phase distortion objectionable transients occur.
RESULTS WITH PICTURE MODULATION
Coming back to the experimental transmitter and receiver setup

again, tests were made on the system using both double and selective
side-band operation. The previous measurements and calculations
should lead us to expect that with selective side -band operation, there
should be much better detail due to the additional high-frequency re-
sponse. At the time these measurements were made, the picture scan-
ning equipment had an upper frequency limit of 500 kilocycles so that
when pictures under both conditions of operation were compared, most
observers agreed that there was very little difference between the two.
Since that time the upper frequency limit of the picture pickup equip-
ment has been increased and the expected increase in detail clearly
demonstrated. Changing from double side -band to selective side -band
operation, therefore, means an approximately two -to-one improvement
in detail which results in a distinctly clearer and sharper picture.
SECOND-DETECTOR DISTORTION
The conclusions drawn from the calculated fidelity characteristics

shown in Figs. 8 and 10 were based on the assumption that the detector
distortion is negligible. This was shown to be true for any detector so
.,
82 RCA REVIEW
long as the percentage of modulation is sufficiently small. The conclu-

sions have to be modified when detector distortion with large percent-
ages of modulation is considered.
Referring to (3) it may be seen that for double side -band reception
where Au = AL and cpu -cpo = epo -4L the form of the carrier envelope
becomes
E Ae - mAu cos (wit + -yL

(8)
2 J
so that a linear detector will introduce no distortion for any percentage

of modulation. If one of the side bands, say the lower, is completely
suppressed then AL = 0 and (3) reduces to
m2
Ve = E[Ae2 + A.' + mAeA4 cos (wit + 4u - 0o)J1/2 (9)
,; fence, the output of a square detector will be J

I¡r- m2
V,2 = E2[Ae2 -}-
4
mAeA, eos (wit + - 001 (10)
so that no distortion is introduced if a square detector is used. If one
...
of the side bands is but partially suppressed then it follows from (3)
that, for high percentages of modulation, there is, in general, no de-
tector which will reproduce only the modulating frequency.
14
1,T .
es
c 10
P i N\G
á g
.
Z,. i
p Hdu
s i2~`.i
4
o. ME
. 20
3pO'
60 80
4.- P9RCENTA6E OF MO DOLAT100OH
1
Fig. 11-Variation of harmonic distortion with percentage of modulation.
Some estimate of the second -detector distortion may be made by

assuming that one of the side bands is totally suppressed and that a
linear detector* is used on the envelope given by equation (9). Figure
* 100 per cent modulation means a carrier modulated to 100 per cent
with both side bands present.
10 gives the per cent of harmonics introduced by a linear detector as

the percentage of modulationt is increased. It is seen that the intro-
duction of these harmonics occurring at high percentages of modulation
would be objectionable in the case of sound reception. However, in
television it is not the frequency per se but rather the wave form of
the signal that is important. The solid line of Fig. 11 shows a single
Fig. 12-Single side -band carrier envelope.
side -band carrier envelope (the upper half of which is the output of a
linear detector) and the dotted line shows the fundamental sine wave.
In television reception the two wave forms would appear as practically
identical.
In discussing detector distortion for large percentages of modula-
tion it is not sufficient to consider the envelope given by (3), but rather
it is necessary to consider the envelope of a carrier modulated with any
number of frequencies. By a method identical with that used in deduc-
ing (3) it may be deduced that the envelope of a carrier modulated with
n frequencies is at the input to the second detector.
Ve = ErAs .+ "'2 E (A L;2 -}- Au;"-)

L 4
{ -E
m2
2
n
;sl
AL Auì cos (2wit + iko - (AL;)
m2
E E
L !
l A L.A L; CoC [(wt---47T L; -o j
L:1
4 ;=1
;oi
+ A.A.; cos [(w; - w;) + Ou; - 0u; l

+ 2ALAu; cos [(ws + wi) + Ou; - 4'L;}
-1- mA, {AL; cos (wit + 4)o -yL;)
)
i =1
/2
+ Aú; cos (wit + cou; - q5o) } (11)
t Since a linear detector is usually used in practice.
34 RCA REVIEW
It is to be noted that for large percentages of modulation any de-
tector will reproduce not only the original modulation frequencies but
also a great many others resulting from various combinations of the
modulating frequencies.
With the effect of the second detector in mind, picture signal was
again put on the experimental transmitter operating to suppress one
side band and effects due to this type of distortion looked for. The pic-
ture modulation was increased to a value where saturation in the modu-
lator began to be noticeable. All observers agreed that up to this value
of modulation no difference in the picture compared to one at a lower
value of modulation could be noticed. This supports the theory that
distortion of this type causes no appreciable hurtful effect in the pic-
ture. It also indicates that the amount and type of distortion which
can be tolerated in a picture signal is quite different than that in a
sound signal.
LOCATION OF THE CARRIER ON THE SELECTIVITY CURVE
The calculated fidelity curves showed how the over-all frequency

characteristic was greatly influenced by the exact position of the car-
rier at the edge of the selectivity curve. When the carrier was tuned at
the 50 per cent response point of the over -all selectivity curve a very
good over -all fidelity curve was obtained. At this point, however, the
selectivity curve is quite steep and slight variations in tuning cause
considerable changes in the over -all frequency characteristic. It also
means that to obtain uniform results from a number of receivers their
selectivity and tuning characteristics must be held to very close toler-
ances. We have found that a reasonable compromise is to have the
carrier approximately 25 per cent down from maximum response. At
this point the selectivity curve has not yet become very steep and the
slight drop in high frequency response can be compensated for in the
video frequency amplifier following the second detector.
TRANSMITTER CONSIDERATIONS
Since the time these measurements and calculations were made a

moderate power test transmitter was installed in Camden to provide
a signal at a receiver location a mile away. No attempt was made to
suppress one side band in the transmitter but all the receivers used
with it were of the selective side -band type. Excellent results were ob-
tained with this system and the difference in detail between double
side -band and selective side -band operation could be easily demon-
strated by tuning from the center to the edge of the receiver selectivity
curve. The suppression of one side band at the transmitter becomes a

very difficult problem at the frequencies which are used for television.
If this can be successfully done then the band width of one channel for
television transmission can be considerably reduced. The power re-
quirements of the transmitter are expected to be approximately the
same, whether the double side -band or selective side -band operation is
used. While the input signal may be thought of as being reduced due
to the absence of one side band, the gain in the input circuits can be
increased due to the smaller band width necessary. This increases the
signal -to -noise ratio on the grid of the first tube which compensates
for the loss of one side band.
CONCLUSIONS
This investigation has shown that no serious difficulties are en-

countered when a television system is operated with the carrier at one
edge of the over-all selectivity curve. The necessity for fewer stages of
amplification in the intermediate -frequency amplifier of the receiver
makes it very desirable to adopt this system. In addition to this, if
one side band can be suppressed at the transmitter there will be a
considerable saving in channel requirements.
EQUIPMENT USED IN THE CURRENT RCA
TELEVISION FIELD TESTS
BY
R. R. BEAL
Research Supervisor, Radio Corporation of America
HE development of the RCA high definition television system

has been advanced by a step -by -step program of research in the
laboratory and tests in the field over a period of more than ten
years. These developments have passed through many stages during
which effort has been continually directed toward producing a system
to provide a standard of performance of lasting value. Many new
devices and methods have been evolved. Mechanical arrangements used
in the early phases of the work have been entirely replaced by elec-
tronic methods through which much higher standards may be achieved.
The RCA all-electronic system employs the "Iconoscope "* as the device
which converts the light image into electrical impulses for transmission
as radio signals, and the "Kinescope"* for transforming these signals
back into visible images.
This system is now undergoing experimental tests in the field in
furtherance of its development by progressive and evolutionary steps.
These tests are being conducted in the New York City area. They are
comprehensive in scope and embrace studies of the functioning of the
equipment under field conditions; propagation studies to determine the
service area; studies to determine the source of and corrective meas-
ures for interference; measurements to determine the necessary signal
levels in Metropolitan New York and the surrounding suburban locali-
ties; experiments in program technique; studies related to receiver
installation and operation ; and observations on receivers in the field by
technical personnel for determinations of standards for an acceptable
and satisfactory system. These field tests began pn June 29, 1936 and
will continue for several months.
The equipment provided for the field tests is installed substantially
as it would be employed in a radio broadcasting service. Studios for
programs in which artists perform and for motion picture film pro-
grams are located in the RCA Building, Radio City. The transmitting
* Trade Mark Registered U. S. Patent Office.
36
CURRENT TELEVISION FIELD TEST EQUIPMENT 87
equipment is installed in the Empire State Building and the trans-

mitting antenna is on top of the building. Ultra short waves are
used for transmitting the "video" or picture signals and the accompany-
ing sound. The picture signals are transmitted on a frequency of
49.75 megacycles and the sound on a frequency of 52 megacycles.
The system is now using standards of which the essential elements
are 343 lines per frame, a frame frequency of 30 per second, a field
frequency of 60 per second (interlaced), negative polarity of trans-
mission, and a video -audio (picture- sound) carrier spacing of 2,250 kc.
In cooperation with the radio industry, RCA has recommended the
..
adoption of standards which include images of 441 lines and a video-
J, .0 .' ':.,.,, VaolleMENT
PNRPT
COOPMENT
I = I
TELEVISION TERMINAL EQUIPMENT FOR RCA FIELD TEST SYSTEM
Fig. 1
audio carrier spacing of approximately 3,250 kc. The RCA field test
system will be changed to conform to these standards at a time
convenient in the experimental program.
The principal groups of equipment and the continuity of the sys-
tem are shown diagrammatically in Figures 1 and 2. Figure 1 shows
the terminal equipment installed in Radio City, and Figure 2 the video
and audio transmitters in the Empire State Building.
The terminal equipment at Radio City, Figure 1, includes three
"Iconoscope" cameras for direct pickup in the artists studio and their
video amplifier, deflecting and control apparatus. Each of these
"Iconoscope" cameras includes a preamplifier for amplifying the video
output of the "Iconoscope ". This output is delivered by cable to ampli-
fying and control equipment from which it is fed to a video line
amplifier for transmission to the Empire State Building,
38 RCA REVIEW
In the film studio, two film projectors of special design are pro-
vided. Two "Iconoscope" cameras are furnished and, as in the direct
pickup studio cameras, these include preamplifiers for amplifying the
video output of the "Iconoscope ". The camera output is delivered to
the video amplifier and control equipment, after which the picture
signals are fed to a video line amplifier for transmission to the Empire
State Building.
The picture signals may be transmitted to the Empire State Build-
ing either by a radio relay channel or by an experimental coaxial
aNRIECTSD V+ Po..
ACCO 1,100411,7001
VIDEO TRANSMITTER FOR RCA FIELD TEST SYSTEM
Pew CPI SI
AUDIO TRANSMITTER FOR RCA FIELD TEST SYSTEM
Fig. 2
cable. The radio relay transmitter and the common generator of

synchronizing impulses which supplies the entire system, are located
at Radio City.
The video and audio transmitters installed in the Empire State
Building, are shown diagrammatically in Figure 2. The frequency
control of both transmitters is provided through conventional temper-
ature controlled crystals. Frequency multipliers are employed to pro-
duce the carrier frequencies. The outputs of both transmitters are
passed into a coupling filter which permits delivering both signals to
a common transmission line and antenna without reacting on each
other in the power amplifiers.
TELEVISION STUDIO
The equipment in the Radio City television studio is shown in
Figure 3 as it is used for picking up programs for transmission. A
light image of the scene to be transmitted is focused through a lens
system on a mosaic composed of a large number of separate photo-
sensitive elements in the "Iconoscope" in the camera. An electron
beam produced in the "Iconoscope" scans the mosaic and converts the
light image into a train of electrical impulses with amplitudes rep-
Fig. 3- Television studio equipment at Radio City
resenting the various intensities of light as it is distributed from

point to point over the image. These electrical impulses are the picture
signals.
In the scene shown in the photograph, two "Iconoscope" cameras
are employed for picking up close-up and distant views to be trans-
mitted in sequence by switching from one camera to the other. The
cameras are movable as units, adjustable with respect to height and
the upper section containing the "Iconoscope is movable horizontally
and vertically for "panning" or following the action. The camera
attendants are equipped with telephone receivers and microphones for
receiving and acknowledging instructions from the studio control room.
40 RCA REVIEW
The lighting equipment is flexible to enable comprehensive studies

of a variety of effects in experimental programs. Incandescent lamps
are used in the equipment shown in the photograph. Studies of light
intensities in relation to the subject to be picked up and of various
types of lights constitute one phase of the engineering work related
to program experimentation. Other phases of the work involve studies
Fig. 4- Television studio control room
of program technique and research and engineering on methods of

expanding the program capabilities of the system.
The sound which accompanies the television scene is picked up
with a boom type of microphone and delivered to the studio control
room after which it is fed to the Empire State Building over high
quality telephone circuits.
STUDIO CONTROL ROOM
The studio control room, Figure 4, is at one end of the studio

and at such an elevation that the engineers have a clear view of the
floor on which the pickup is made. The control console position in
the foreground of the photograph controls the sound from the studio.
The video circuit controls are at the opposite end of the console. Two
video monitors are mounted on the wall in front of the video control
position. One monitor shows the scene that is being transmitted and
the other, the scene picked up by the second "Iconoscope" camera
preparatory to transmission. The engineer at the video control posi-
Fig. 5 -Film studio equipment
tion performs the operation of switching from one camera to the

other.
The racks of equipment behind the engineers include the video
amplifiers and the synchronizing and control equipment associated
with each of the "Iconoscope" cameras.
FILM STUDIO
The film studio equipment, Figure 5, consists of two motion pic-

ture projectors of special design to permit the use of standard 24
frame motion picture film to produce television pictures at 30 frames
per second. In these projectors a changing rate of intermittent drive
42 RCA REVIEW
is used for the picture portion of the film and a constant 24 frame
rate of feed for the sound portion. Pictures from the projectors are
focused on the mosaics of the "Iconoscope" cameras located in the con-
trol room beyond the partition separating the two rooms.
FILM STUDIO CONTROL ROOM

The equipment in the film studio control room, Figure 6, includes
two "Iconoscope" cameras with their video voltage amplifiers and asso-
ciated synchronizing and control equipment and audio equipment, and
Fig. 6 -Film studio control room
controls for the sound from the film. The two "Iconoscope" cameras
are so mounted that they may be shifted from side to side for use
with either of the film projectors.
Adjustments of signal levels and switching are accomplished by
the engineer at the control console. Two video monitors are furnished,
one for each film projector channel, to provide for continuous trans-
mission from film.
SYNCHRONIZING GENERATOR AND TIME AMPLIFIER EQUIPMENT
The panels containing the electronic synchronizing generator equip-

ment and the video line amplifiers which feed the video signal to the
Empire State Building are shown in Figure 7. Synchronization at the

receiver is obtained by transmitted impulses. The horizontal and
vertical impulses have the same amplitude and wave shape selection
is employed.
INTERBUILDING RADIO RELAY
The ultra short wave radio relay transmitter, Figure 8, is installed
on the 10th floor of the RCA Building. It operates on a frequency of
Fig. 7- Synchronizing and video line amplifier panels

177 mc. The circuit has a channel width of 3 mc. to carry the video
frequencies up to 1.5 mc. with double side band transmission. Video
signals from the studio are delivered to this transmitter by coaxial
cable. Video monitoring equipment is provided for the signal at this
point.
The radio link transmitting antenna is located at about the 14th
floor level on the south side of the RCA Building to provide an unob-
structed transmission path to the receiving antenna placed at the
44 RCA REVIEW
85th floor of the Empire State Building. The receiver for the radio
relay circuit is installed in the Empire State Building transmitter
control room. The air-line distance between the two buildings is about
.87 mile.
The overall frequency characteristic of the radio relay circuit is
substantially flat over the range from 20 cycles to 1500 kc. This
Fig. 8 -Ultra short wave radio relay transmitter

circuit is practically free of noise and the picture quality over it is
equal to that obtained over the coaxial cable.
EMPIRE STATE BUILDING CONTROL BOARD
The coaxial cable and radio relay channels and the channel for the
sound accompanying the picture from the studios in Radio City, ter-
minate at the Empire State Building control board (Figure 9). From
left to right, the control board consists of the sound channel panels,
a video monitoring panel, the radio relay receiver panel and battery
and switching panels. The video monitor may be switched either to
the radio relay or the coaxial cable channel. The video signals are
delivered to the transmitter by coaxial cable.
EMPIRE STATE BUILDING TRANSMITTERS
The video and audio transmitters installed in the Empire State

Building are shown in Figure 10. The tubes used in the final power
amplifier of these transmitters are especially suited for the frequencies
employed. Their plate dissipation rating is 30 kw. per tube. The fila-
ment power is sufficient to produce an electron emission of 18 amperes
Fig. 9 -Left to right, sound channel panels, video monitoring panel,

relay receiver, battery and switching panels
per tube which permits of a video carrier power of 8 kw. with a tank
circuit loading to pass the 1.5 mc. sidebands.
The audio carrier final power amplifier is plate modulated in the
conventional manner. In the video carrier final power amplifier, grid
modulation is employed to reduce the video voltage that must be devel-
oped. The complete video modulator is impedance coupled.
EMPIRE STATE BUILDING ANTENNA
A single antenna structure is employed to radiate both the audio

and video signals. In this antenna, the fundamental radiator unit
46 RCA REVIEW
consists of three dipoles arranged in the face of an equilateral triangle.
Three of these units are so positioned vertically as to increase the
concentration of radiation in the horizontal plane. A horizontally
polarized field is produced with an essentially circular pattern in the
horizontal plane. The power gain in the horizontal plane is about 2.1
to 1 or 3.2 db. as measured with reference to a vertical dipole.
The frequency band of the antenna is practically flat over the
upper side band of the video transmitter, namely 49.75 to 51.25 mega-
Fig. 10-Video and audio transmitters
cycles, and its flat characteristic includes both side bands of the audio
transmitter, 52 megacycles plus and minus 10,000 cycles.
Due to the quasi optical properties of the ultra short waves em-
ployed in television the transmitting range increases with the height at
which the transmitting antenna is placed. The Empire State Building,
having a height in the order of 1250 feet, provides a location from
which a maximum transmitting range may be obtained.
EXPERIMENTAL FIELD TEST RECEIVERS

The field test receivers are of the superheterodyne type and have
a tuning range of 42 to 84 megacycles. They receive both sound and
picture simultaneously. The head end circuits accept both carriers and
one picture side band. Tuning is accomplished by a single knob con-
trolling the radio frequency circuit and the single oscillator which
heterodynes both carriers to produce the intermediate frequencies.
These are separated by the spacing of the transmitted carriers, namely,
2,250 kc.
Figure 11 shows a photograph of the television field test receiver
Fig. 11- Television receiver used in tests
with the cover raised to the viewing position. The television image
is produced on the luminescent screen of a "Kinescope" 9" in diameter
which provides a picture size of approximately 51/2" x 71/4 ". The "Kine-
scope" is mounted vertically and the picture is viewed in the chromium
plated steel mirror mounted inside of the cover of the cabinet.
Of the seven knobs on the front of the receiver, the center knob
tunes both the picture and the accompanying sound. The three knobs
on the right from top to bottom are the sound volume control, the
48 RCA REVIEW
treble tone control and the bass tone control. The three knobs on the
left are the picture contrast control, the detail control and the back-
ground control.
The receiver operates from the usual 110 -volt, 60-cycle power sup-
ply and draws about 350 watts of power. Since the synchronization
is controlled entirely by impulses sent from the transmitter, it is not
necessary that the power supply frequency of the receiver be syn-
chronized with that of the transmitter, although it should have the
same nominal frequency.
The experimental receivers have 33 tubes including the "Kine-
scope". Horizontal dipole receiving antennas are used in the field
installations.
As the field tests have advanced the soundness of the technical
fundamentals of the system have been confirmed. Engineering studies
related to program technique are broadening the program capabilities
of the system. Good progress has been made in adapting the system
to practical operating conditions in the field. Live talent and motion
picture programs have been satisfactorily transmitted and received
over distances of 25 miles from the Empire State Building at typical
apartment house and suburban home locations. The height of the
transmitting antenna has made possible consistently good reception
at one favorably located suburban home over a distance of 45 miles.
Measurements of signal field strength and noise intensity are
in progress to determine the requirements for service in the area
under test. Electrical interference on the frequencies employed con-
sists almost entirely of man -made noise originating in automobile
ignition systems, diathermic apparatus and other electrical devices.
Propagation studies and measurements are under way to obtain data
on the interference range of the ultra short wave signals used for
television. The tests of the system are incomplete, but as they progress
the engineering and technical data, and the experience obtained by
operating the system under field conditions, are expanding its capa-
bilities and pointing the way toward realization ultimately of a satis-
factory high definition television system for broadcasting service.
AUTOMATIC ALARM
BY
I. F. BYRNES and H. B. MARTIN
Engineering Department, Radiomarine Corp. of America
MORE than ten years ago it was recognized that safety at sea
would be improved if some means could be developed to en-
able a ship in distress to summon aid at any time by radio
from a nearby vessel, especially during those periods when the radio
operator on the nearby vessel might not be "on watch ". Early devel-
opment and tests indicated that a special distress signal, international
in scope and known to all radio equipped ships, would be desirable.
Such a signal would then actuate an automatic receiving device on
vessels in the vicinity of the ship in distress and by means of bells
or other means call attention to the fact that a distress call was being
transmitted. This special signal (which supplements and does not
supersede the conventional S 0 S . . . - -- ... signal) is known
as the automatic alarm signal and the special receiving apparatus is
called the auto alarm. The auto alarm signal is transmitted by the
ship in distress just prior to sending the normal S O S signal.
The form of the alarm signal was specified at the International
Radio Telegraph Convention held in Washington in 1927 and later at
the International Telecommunication Convention of Madrid, 1932. It
consists of a series of dashes and spaces, each dash having a duration of
four seconds and each space between dashes a duration of one second.
Twelve such dashes and spaces can be transmitted in a period of one
minute. The auto alarm apparatus is then designed to actuate the
bells when a certain number of consecutive dashes and spaces are
correctly received. European practice provides for auto alarm opera-
tion after three correct dashes and spaces are received, while present
United States practice is based on four dashes and spaces. Reasons
for this difference are discussed later.
An "International Convention for the Safety of Life at Sea" was
held in London in 1929 and attended by representatives of the prin-
cipal maritime nations of the world. At this convention certain regula-
tions were adopted pertaining to radio aboard ship and particularly
with reference to "watches" in the radio room of cargo vessels. It
49
50 RCA REVIEW
was ruled that cargo vessels of over 5500 gross tonnage, not fitted
with an auto alarm, shall, while under way, keep a continuous watch
by means of an operator or operators. On ships which are fitted with
an auto alarm, this apparatus must be in operation at all times when
the operator is not on watch.
Before auto alarms may be installed aboard ships to meet the re-
quirements of the various international conventions, the type of appa-
ratus used must be approved by the Government having jurisdiction
over the vessel. The Government of the United States deposited its
ratification of the 1929 Safety at Sea Convention on Aug. 7, 1936.
In October, 1935, the Federal Communications Commission pre-
pared specifications and test procedures for the guidance of American
manufacturers in the development of a satisfactory auto alarm. Some
of the outstanding technical conditions that must be satisfied and the
manner in which the various requirements are met in the Radiomarine
auto alarm, are discussed in the following paragraphs. Part I covers
the principal design requirements. Part II describes the technical
design of the AR -8600 Auto Alarm.
PART I
There are two basic elements in an auto alarm, namely the radio
receiver and the selector. Each element requires a design quite unlike
that satisfactory for other services. The radio receiver must be
arranged to possess uniform sensitivity over a frequency range from
487.5 to 512.5 kc., or a total band width of 25 kc. The normal distress
frequency is 500 kc., but provision must be made for the auto alarm
receiver to accept signals outside the exact frequency, thereby permit-
ting some variation in the adjustment of the radio transmitter on the
vessel in distress. The band width of the alarm receiver must also
be fixed as an integral part of the design and not be adjustable by
the operator. To meet F.C.C. requirements as to sensitivity the
receiver must function with an input of 500 microvolts applied through
500 micro -microfarads, 20 microhenrys and 5 ohms over the specified
25 kc. band. Then, on th basis of the ship's antenna Y.aving an effec-
?
tive height of 5 meters, a field strength of 100 microvolts per meter

will be sufficient. In actual practice with average shipboard main
antennas it is easy to obtain effective heights greater than 5 meters
so that the maximum sensitivity of the auto alarm is more apt to be
limited by the prevailing noise level.
Selectivity requirements for the auto alarm receiver may be
analyzed as follows : Above 512.5 kc. interference may first be expected
from broadcast stations operating in the channels around 550 kc.
AUTOMATIC ALARM 51
Broadcast stations whether located along coastlines or considerably

inland may be expected to lay down either a strong ground wave
which might interfere with coastwise vessels, or a sky wave at night
which may be a source of difficulty. F.C.C. specifications state that the
auto alarm, when adjusted for 500 microvolts input, must not be made
inoperative with an interfering modulated signal of 25,000 microvolts
at 550 kc. When the interfering signal is 650 kc. it may have an
amplitude of 100,000 microvolts without affecting the auto alarm.
Below 487.5 kc. interference may be expected from marine traffic in
the band from 375 to 500 kc. Selectivity requirements in this band
call for no interference from a 25,000 microvolt signal at 450 kc. and
a 100,000 microvolt signal at 375 kc.
Overload characteristics of the auto alarm receiver are of impor-
tance. When the operator goes off watch he places the auto alarm
in circuit and adjusts the sensitivity control to an optimum value for
the prevailing noise level as explained later. The apparatus must then
be capable of accepting a very strong signal from a nearby vessel,
without blocking or overloading. Automatic volume control used in
the conventional manner is of no assistance for this problem because
the desired incoming signals are completely keyed, that is, the carrier
wave is either on or off. AVC would simply raise the gain of the
receiver each time the incoming signal was cut off, and during this
interval the noise and static picked up by the antenna would tend to
block the receiver. Time delay in an AVC circuit is also inadmissible
since the receiver and selector circuits must recognize extremely short
"breaks" in incoming signals in order to permit the alarm signal to
function through interference on 500 kc. This action is explained
further in the selector design. Satisfactory design to meet F.C.C. rules
must provide operation with a 90,000 microvolt signal at 500 kc., when
the receiver is adjusted to also respond to a 500 microvolt input.
Since the auto alarm when connected in circuit operates as an
unattended device for several hours, means must be provided to indi-
cate when operation is not normal. There is a possibility that prolonged
static of high level will "hold over" some of the relays in the selector
unit. It is therefore necessary to arrange for a warning light or its
equivalent to show on the bridge so that the radio operator may be
instructed to readjust the sensitivity control. Vacuum tubes in the
apparatus may burn out. If this occurs a "no current" relay is used
to energize the warning bells which are located on the bridge, in the
radio operator's cabin and in the radio room. Failure of the source of
energy which rings the bells is shown by continuous burning of the
warning lights, located alongside the bells.
52 RCA REVIEW
After the radio signal passes through the receiver it controls the
selector mechanism. To allow for reasonable variations in the timing
of the alarm signal the selector must be designed to accept dashes
having a duration of 3.5 to 4.5 seconds and spaces from .1 to 1.5
seconds. The question of operation through interfering signals in the
487.5 to 512.5 kc. band as well as the possibility of false alarms must
also be considered when determining the selector timing tolerances.
The alarm must function through a reasonable amount of interference
on the same frequency as the distress signal. This interference may
produce two effects on the selector. One effect is to prolong the normal
four -second alarm signal in case the interference appears at just the
correct time to add to the desired signal. The second effect of inter-
ference is to "fill in" the normal one -second spaces. If the spaces
are completely filled in at the correct time the selector functions to
reject the signals and it will do the same if interference unduly pro-
longs the desired dashes.
Ordinary telegraphic traffic does not interfere, even though on the
same frequency, with normal operation of the auto alarm, especially
if the selector unit is responsive to extremely short spaces. The mini-
mum U. S. requirement is one -tenth second and if the design provides
performance equal to or better than this value, then the alarm signal
will pass through the selector as long as some "break" takes place in
the interfering signal when the "space" occurs in the alarm signal.
The possibility of false alarms is determined by a combination of
three main factors, namely fortuitous or accidental combinations of
signals or noise equivalent to the alarm signal, timing tolerances of
the selector and finally the number of dashes and spaces which are
selected to ring the bells. As mentioned in the first part of this paper,
European practice is to arrange for the bells to ring when three con-
secutive dashes and spaces pass through the receiver and selector.
F.C.C. requirements are based on ringing the bells after four con-
secutive dashes and spaces are correctly received. The four -dash cycle
considerably minimizes the possibility of false alarms since the chances
are quite remote for accidental combination of signals to repeat them-
selves four times to imitate the alarm signal. On the other hand it is
somewhat more difficult under conditions of severe interference for
the auto alarm to accept the four instead of the three -dash cycle. It
may be mentioned that current designs of American auto alarms may
be easily arranged to accept either the three or four -dash cycle and in
any case no international operating difficulties arise since the vessel
in distress always sends twelve or more dashes.
AUTOMATIC ALARM 53
A rugged design of auto alarm is necessary to withstand success-

fully the operating conditions which obtain aboard ships. In typical
cases the auto alarm will be "on the air" for an average of approxi-
mately 5000 hours yearly. All the time the alarm is in circuit the
receiver and certain parts of the selector are continually responding
to ordinary telegraphic signals, static; induction, etc. This entails con-
siderable wear and tear of any moving parts and for this reason mov-
ing parts should be kept to a minimum, while relays and the like
must withstand hundreds of thousands of cycles of operation. The
Mixer- 0sc. Ist. IF.Amp. Znd I.FAmp. Det. D.C. Amp.

6K7 6K7 6N6 RCA-1611
6,18
o
o
. - -
Fig. 1- Simplified receiver schematic circuit diagram
design must also include provision for testing by sending a local
alarm signal through the circuits, together with suitable meters and
controls to enable the operator to check the overall performance. F.C.C.
specifications also call for ability to endure shipboard vibration,
humidity and temperature. A single master switch to place the auto
alarm in service is required which must be so arranged that power
cannot be applied to the alarm circuits unless the main antenna is
connected to the alarm receiver and having an interlocking feature to
prevent the ship transmitter from being keyed unless the auto alarm
receiver is turned off.
PART II
The design of equipment necessary for performance as indicated
in Part I can best be considered by using a natural division of receiver
and selector requirements; that is, the receiver must have sufficient
sensitivity and uniform reception over a band of 487.5 to 512.5 kc.
and a selector system to differentiate between static and ordinary
54 RCA REVIEW
communication and the auto alarm signal. In order to start selector
operation some form of d -c amplifier is indicated. The plate circuit
of this d -c amplifier must have sufficient power and correct character-
istics to work a fast operating signal relay, the contacts of which
initiate selector action. Certain characteristics of tuned circuits may
be used to evolve the receiver design. Band pass intercoupling cir-
cuits are indicated from a consideration of the selectivity and band
width requirements, as well as considerable amplification, in order to
control the grid circuit of the d -c amplifier or signal relay tube. The
coefficient of coupling (k) of individual r-f transformers would be
of the order of .06 to provide the necessary band width.
Band within = k x resonant frequency (approximately).
A circuit over -coupled to this extent would have a severe valley at the
resonant frequency if modern high Q coils were used. High Q circuits
are desirable in order to secure high attenuation outside of the pass
band. The effect of double humps with a severe trough between them
is even more pronounced when it is realized that several r -f stages
must be used in cascade in order to provide sufficient sensitivity. By
using a frequency greater than 500 kc. the coefficient of coupling in
the band pass circuits may be reduced and still give the required
band width, and in addition, the response within the band may be made
uniform since the coupling need not be much greater than the critical
value. Figure 1 illustrates schematically the tube and circuit layout
of a superheterodyne receiver used to produce the required results.
A total of eight tuned circuits is utilized for necessary amplifica-
tion and selectivity. A signal between 487.5 and 512.5 kc., with
attenuation of signals ouside of these limits, is applied to the mixer -
oscillator tube type 6A8 where it is mixed with the local oscillator
operating at a frequency of 1600 kc. The difference beat between sig-
nal and oscillator frequency, or 1087.5 to 1112.5 kc., is amplified by
two stages utilizing super control pentodes type 6K7 and then applied
to diode detector type 6H6. The d -c voltage developed across diode
resistor R1 is used to control the grid circuit of d -c amplifier RCA -1611,
whose plate relay initiates or stops selector action. The selectivity
obtainable from such a layout is considerably in excess of require-
ments. For example, with the sensitivity control set for 500 micro-
volts, the signal necessary at 450 kc., to produce selector action is
well in excess of 100,000 microvolts, which is four times the required
amount. Likewise, the same values hold for a frequency of 550 kc.
An interfering signal of 25,000 microvolts at the antenna -ground ter-
minals of the receiver, could be as close to 512.5 kc. as 533 kc. and on
AUTOMATIC ALARM 55
the other side of the auto alarm band, the same strength signal could
be as close as 467 kc., before interference would result. A wave trap
directly in the antenna lead is provided for the attenuation of strong
broadcast signals utilizing frequency assignments in the pass -band
accepted by the intermediate frequency amplifiers.
It is desirable to use the ship's 110 -volt line in order to avoid fre-
quent B battery replacements or a dynamotor. This factor is quite
important in selecting the tubes and circuits used. All -metal tubes are
/000
30000
/0000
300
O
]
\ 1000
O
S00
$0 -90 -30 -20 -10 500 +/O +20 +30 +40 +SO
Kilocycles
Fig. 2-- Selectivity curve
used throughout. A 500 -microvolt signal between the limits of 487.5

and 512.5 kc., irrespective of modulation or modulation frequency and
including spark signals, must produce sufficient voltage of the polarity
shown on the diode load resistor R1 (Figure 1) to control the grid of
the d -c amplifier tube. This tube has a "no signal" plate current value
of approximately 7 milliamperes; thus signal relay No. 4 is normally
energized. The 500 -microvolt signal then reduces this relay current
to below its drop-out value of 3 to 4 milliamperes. If the signal
56 RCA REVIEW
remains long enough, selector action obtains as explained later. A
signal stronger than 500-microvolts merely produces plate current
cut-off in the signal relay tube. "Downwards" operation of the signal
relay is desirable inasmuch as a strong signal does not excessively
deflect the milliammeter shown as A in Figure 1, and some sort of
saturation circuit provision is, therefore, not necessary. A saturation
circuit might be subject to a time lag when being restored to normal
and thus limit the high speed operation of the signal relay. This is
quite important, since this relay must recognize breaks of one-tenth
second or less. Sensitivity control R3 is provided for adjustment to
optimum sensitivity, consistent with the prevailing noise level caused
by atmospherics and man -made static. Thus to adjust for proper
operation, the sensitivity control would be turned to its maximum
counter-clockwise position, which gives minimum sensitivity for the
receiver, and then turned clockwise until the plate current reading of
the signal relay tube is approximately one milliampere less than the
former reading. For example, if the plate current which flows through
the signal relay reads 7 milliamperes with the sensitivity control set
to the extreme counter -clockwise point, the control would be turned
clockwise so that the average reading would be approximately 6 mil-
liamperes. Bursts of static would then drive the signal relay current
below the drop -out value and the contacts which initiate selector
action would be occasionally or continually chattering, depending on
existing noise conditions. Quite obviously, a sensitivity control is
necessary; since if the auto alarm was permanently adjusted to
respond to a 500 -microvolt signal, during the heavy static season, and
especially in the tropics, the signal relay would be de- energized most
of the time and the receiver would be "blocked," insofar as ability to
receive the auto alarm signal is concerned. In the event that this con-
dition occurs, that is, where an increase in static level sufficient to
drive the signal relay current below the drop -out value occurs, and
continues for more than 3.5 seconds, warning lights installed beside
each alarm bell will be turned on and will remain lighted until the
static reduces in value, or the sensitivity control is set to the proper
point for the new level.
The characteristics of an acceptable selector unit will next be con-
sidered. Since 16 hours per day or approximately 5000 hours per
year of operation are necessary, the selector unit, to be most reliable,
should have a minimum of moving parts. As previously mentioned,
the standard automatic alarm signal is composed of 12 dashes and
spaces, but the alarm bells are to be actuated after the receipt of four
dashes with tolerances of 3.5 to 4.5 seconds and spaces with tolerances
AUTOMATIC ALARM 57
of .1 to 1.5 seconds. Thus if the space between the first and second
dashes is completely filled by interference, the alarm would be actuated
at the end of the sixth dash. If interference prolongs the length of
any dash beyond 4.5 seconds, or completely fills in the space between
dashes, the mechanism used for selection would be restored to normal.
Thus, of the 12 dashes composing the signal, 4 consecutive dashes
having a space at the beginning and ending of the group, as well as
To all heaters (In cer,e5)

4600004 o
5 /goal
relay Voltage Regulator Warning relay #9
04 o
A
VT- 4-
r=- I
C,
Eg
Warning relay a 9
-^-^-°Oa 01616- O o-
VT-1
r-
I
vl I I
I
H1673150.
i M B
R+ S
Cz
VT-2 Stepping ,twitch

000660ti^ 06
I >o2 -
r--- Xa
Cl
VT- 3
rl- z
000we`-
i13
Ben relay
X,
Rs l 1,
-no +110 +6 -6 -n0

Row
Fig. 3- Simplified selector schematic circuit diagram
between dashes, must be received before the alarm bells will ring.
The obvious advantage of a 12 -dash signal is that it permits more
chances of the alarms being actuated under severe conditions of
interference both from a standpoint of prolonging the dashes and
filling in the spaces. The aforementioned dash and space tolerances
seemingly impose difficult terms for the selector response. For example
four dashes and the intervening three spaces might vary in total
elapsed time between 14.3 and 22.5 seconds. Obviously the selector
58 RCA REVIEW
must check individual dashes and spaces since any attempt to use
the sum would result in the false alarm probability being greatly
increased.
The schematic circuits of the receiver and selector units are shown
separately for explanation purposes, but are combined on one panel
in practice. Figure 3 shows a simplified circuit of the selector unit.
Elapsed time of signal duration is measured by RC circuits connected
in the grid circuits of individual selector tubes. The principle utilized
is the familiar one of current decay in a series RC circuit.
If Ea= steady source of charging voltage

C = capacity
R = series resistance
t = time
i = instantaneous current at time t
= base of naperian logarithms
then i = -
Eg
R
E
r
RC
If ee = Capacitor voltage at time t

e,= Eg iR
therefore e, = Eg E
At time t =RC, ec = 63% of its maximum value Eg
Referring to Figures 1 and 3, the selector action is as follows.

An incoming signal of 500 microvolts or greater, of either A -1, A -2
or B emission, between 487.5 and 512.5 kc. produces a d -c voltage
across the diode resistor R1 of the correct polarity, as indicated in
Figure 1, to reduce lb of the d -c amplifier or signal relay tube below
its drop -out value. Contacts AM (A refers to "armature" or moving
contact, B to "break" and M to "make" contacts when relay is ener-
gized) are normally closed since relay No. 4 (Figure 3) is energized
when no signals are being received. The reduction of lb of the relay
tube therefore causes contacts AB to be closed, which applies charg-
ing voltage E9, developed by the voltage regulator tube, to C1 through
R1. Since the grid- cathode circuit of VT -1 is connected across C1, as
C1 charges, the grid of VT -1 becomes less negative and eventually
Izti begins to flow, gradually increasing to several milliamperes. Thus
if a signal persists for 3.5 seconds Ibr becomes 5 milliamperes and relay
No. 1 closes contacts AM. These contacts apply six volts obtained
from the storage battery to the "notch" coil of the stepping switch
AUTOMATIC ALARM 59
No. 6, which moves up one step and the warning lights go "on ". The
coil of auxiliary relay No. 5 is in parallel with the notch coil of the
stepping switch and, therefore, contacts A1M are made. This starts C2
charging through R2 and if the signal persists up to or greater than
4.5 seconds Ib2 of VT -2 becomes 5 milliamperes, which closes selector
relay No. 2, and its contacts AM, in turn, apply six volts to the
"restore" coil of the stepping relay. As soon as A leaves B of selector
relay No. 2, R8 is inserted in series with the notch coil and auxiliary
Fig. 4 -Front view of Automatic Alarm
relay coil to prevent damage to the low resistance notch coil, due to
an over -long signal or continuous "blocking" by static. The warning
lights continue to glow until the signal stops or has a slight break in
it. A break in the received signal allows signal relay No. 4 contacts
AM to return the grid of VT -1 to a value greater than that required
for plate current cut -off, which allows No. 1 relay contacts AM to
open, thus removing the voltage from the "notching" coil and auxiliary
relay No. 5, in turn allowing the "restore" coil of the stepping switch
60 RCA REVIEW
to return the lever wiper to normal. Contacts A1B of auxiliary relay
No. 5 then bias the grid of VT -2 beyond cut-off and allow selector
relay No. 2 to open contacts AM, which removes the six -volt supply
from the restore coil of the stepping switch. VT -1 thus serves to
check the minimum length of a dash (dashes less than 3.5 seconds do
not actuate the stepping switch), and VT -2 serves to check the maxi-
mum acceptable length of dash (dashes greater than 4.5 seconds ener-
gize the restore coil of the stepping switch which "restores" when
the signal stops) . Four dashes, if greater than 3.5 seconds and less
than 4.5 seconds would then actuate No. 1 relay four times and the
stepping switch wiper would rest on Contact 4. Then if a break occurs
at the end of the fourth dash, Contacts AB of No. 1 selector relay are
closed which applies six volts to the bell ringing relay No. 7. Once
closed, this relay applies six volts to the bells and is held closed by
its auxiliary or hold -in contacts. In order to stop the bells the reset
button on the auto alarm panel must be pressed. The bells are not
allowed to ring until the fourth dash is broken (and within prescribed
limits), since if this was not done three correct dashes and the fourth
one of any length greater than 3.5 seconds would cause an alarm.
VT -3 serves two purposes. If one, two, or three correctly timed
dashes are received, the stepping switch wiper comes to rest on Con-
tact 1, 2 or 3 respectively and would remain in one of these positions
indefinitely until another dash was received, which would then ring
the bells. This is, of course, improper operation. In order to prevent
such a condition the grid of VT -3 is normally connected to the charg-
ing voltage ED through charging resistor R3. The cathode circuit of
VT -3 connects through interlock contacts X1 on the stepping switch,
which are open when the stepping switch is on the zero or "normal"
contact only. Thus after a "notch" has occurred auxiliary relay No. 5
contacts A2M are closed which biases the grid of VT -3 to its below
cut-off value as determined by voltage divider resistors R6 and R7.
At the conclusion of the dash auxiliary relay No. 5 contacts A2B are
made and after an interval of five seconds Ib3 in selector relay No. 3
reaches a value of 5 milliamperes, thus closing No. 3 relay contacts
AM. These contacts in turn apply six volts to the restore coil of the
stepping switch which will then be returned to normal. In the event
that the second or following dash is completed within five seconds
from the end of the previous dash, selector relay No. 3 does not close
contacts AM. This then allows the maximum space between dashes
to be the difference between five seconds and the initial closing time
of relay No. 1, or 1.5 seconds. Such a method is necessary for the
AUTOMATIC ALARM 61
proper checking of spaces, since if an attempt were made to check

spaces directly, any form of interference occurring during the space
would "fill in" and cause an error in timing.
In order to indicate line voltage failure as well as tube heater
burn-out, warning relay No. 8 is connected in series with all tube
heaters which in turn are heated from the ship's 110 -volt line. A
failure of either tube heater or line voltage will cause the alarm bells
to ring. The bells will, of course, stop ringing if the line voltage is
restored. A test button is provided on the auto alarm panel which,
.to 00 w
E; is.
.. ac
h .,
Fig. 5 -Front view of Automatic Alarm--cover open.
when held in, prevents the bridge and operator's room bells from
ringing during routine testing of the auto alarm receiver and
selector.
Warning relay No. 9 is connected across the storage battery through
a series resistor. Failure of the battery supply will allow relay No. 9
to de- energize and turn on the warning lights at each bell location
point. Warning of power failure is therefore obtained except for
simultaneous failure of both the 110 -volt ship's line supply and the
storage battery. The chances of simultaneous failure are very remote.
62 RCA REVIEW
Following is a summary of warnings:

(1) Bells ringing may be caused by
(a) Receipt of auto alarm signal.
(b) Receipt of a false auto alarm signal caused by a for-
tuitous combination of static and keyed interference.
(c) Loss of ship's line voltage.
(d) Tube heater burn -out.
(2) Warning lights burning continuously are caused by
(a) Receipt of a continuous signal from a transmitter whose
key is being held down for a period considerably greater
than 4.5 seconds.
(b) Sensitivity control set too high for the prevailing noise
level.
(c) Loss of 6 -volt battery supply.
(3) Warning lights burning intermittently are caused by
(a) Occasional long bursts of static.
(b) Transmitter testing using dashes slightly longer than 3.5
seconds.
(c) Heavy 500 kc. interference caused by several telegraph
transmitters transmitting at the same time.
Warnings as under (3) are to be expected and indicate that the auto
alarm is functioning correctly. Warnings as under (1) and (2) should
be investigated as to their cause.
The complete auto alarm installation consists of :
(A) The auto alarm receiver and selector chassis.
(B) A cable junction box which also provides mountings for
(1) The radio room bell.
(2) The radio room warning light.
(3) Fuses for both 110 -volt and 6 -volt supplies.
(4) Charging lamps for the 6 -volt battery.
(C) An "antenna -on -off" switch which
(1) When in the "receive" position (operator "off watch")
(a) Connects the main ship antenna to the auto alarm.
(b) Opens transmitter key relay power supply circuit.
(c) Connects the 110 -volt supply to the auto alarm.
(d) Connects the 6 -volt supply to the auto alarm.
AUTOMATIC ALARM 63
(2) When in the "off" position (operator "on watch ")

(a) Connects the ship antenna to the lightning switch.
(b) Closes key relay power supply circuit.
(c) Removes the 110 -volt and 6 -volt tube and relay supply
voltages.
(d) Places 6 -volt battery on charge.
(D) Two bells and warning light boxes, one each to be located on
the bridge and in the operator's room.
(E) One 6-volt storage battery.
Fig. 6- Automatic Alarm with cover and panel open
A small oven, thermostatically controlled and containing the three

selector relays for special protection against the humid conditions
encountered at sea, keeps the auto alarm chassis perfectly dry inside.
All adjustments which affect timing or band width are made relatively
inaccessible by cover plates. Figure 5 shows these plates in place over
the r-f transformer shield cans at the top of the panel and over the
holes in the oven through which the selector relays are adjusted.
Figure 4 shows the auto alarm receiver in its cast cabinet. The
cabinet is made in two sections; the back section, which mounts on
a bulkhead by means of rubber shock mounts, is fitted with hinges
64 RCA REVIEW
on which the panel is hung. When the panel is swung into its cabinet,
the inside is firmly sealed by means of a hollow rubber tube pressing
against the back of the panel. The front section of the cabinet is also
hinged to the back section. The hinge is of special construction so
that the panel may be lifted off the hinge. The front section of the
cabinet covers the upper section of the panel, but leaves the meters,
meter switch, phone pack, test buttons and sensitivity control acces-
sible for observation and adjustment. Figures 5 and 6 show how the
cover and panel swing open for observation and servicing. The cabinet
occupies a space on the radio room bulkhead of 16 x 26 inches. The
weight is approximately 60 pounds.
The apparatus described in this paper represents the present-day
status of an auto alarm development which has been under way since
1928. Acknowledgment and appreciation is expressed to Mr. Chas. J.
Pannill for his cooperation and encouragement during this work.
REPRODUCING EQUIPMENT FOR MOTION PICTURE
THEATRES
BY
M. C. BATSEL and C. N. REIFSTECK
Photophone and Test Equipment Division, RCA Manufacturing Co.
EQUIPMENT for reproduction of sound motion pictures Is lo-

cated in two places in every theatre. The loudspeaking or re-
producer equipment is on the stage behind the sound screen
and the soundhead and amplifiers are in the projection booth, usually
located at the rear of the theatre.
The main elements for successful reproduction of sound accom-
panying a picture are:
1. A soundhead used for translation of the photographic sound
record to minute electrical energy.
2. Amplifying equipment to increase the minute electrical energy
from the soundhead to a value that will operate the loud-
speakers at the required volume.
3. Loudspeaking equipment which translates electrical energy to
acoustical energy or sound.
In addition to these three major elements, there are other associ-
ated parts which are used as part of the complete equipment to insure
proper performance in the theatre. Some of these are: changeover
switching systems to transfer the amplifier from one machine to an-
other for continuous performance, a monitor speaker for the projec-
tionist, and, in some installations, a remote volume control located in
the auditorium.
(A) THE SOUNDHEAD1
The sound reproducing attachment or soundhead is used in con-
junction with a standard motion picture projector. It contains all the
optical, electrical, and mechanical equipment necessary to convert the
variations in opacity of the photographic sound record into electrical
currents.
1 Journal S.M.P.E. -Vol. XXV No. 5, Page 449.
65
66 RCA REVIEW
A recent High Fidelity soundhead, known as the rotary stabilizer
type, is shown in Figure 1. This designation refers to the system em-
ployed for imparting a steady motion to the film at the point where the
translation takes place. This type of soundhead is considered a decided
improvement over previous types used.
Fig.1
The first types of soundheads employed fixed sound gates consist-

ing of stationary guide plates and pressure shoes for holding the film
in the position where it was pulled by a sprocket past the translation
point. Experience demonstrated that these gates required almost con-
stant attention to prevent the accumulation of wax and emulsion on
the polished surfaces of the film guides and pressure shoes to prevent
the film from chattering or being scratched which would result in
REPRODUCING EQUIPMENT, MOTION PICTURE THEATRES 67
noisy reproduction of the sound. The frictional resistance to the film

passage resulted in wear of the guides and pulling sprockets.
A major problem in the first types of soundheads was the achieve-
ment of a constant speed of the film at the translation point. Many
attempts were made to have the pulling sprocket directly below the
gate revolve at a uniform velocity by the use of elaborate mechanical
filters between the driving motor and the sprocket. In the best of
these systems some ripple was present in the sound reproduction from
the film, due to uneven motion of the film resulting from imperfect
mechanical parts. A serious flutter or rapid variation in the motion
resulted from the use of a sprocket to pull the film through the gate
by means of the perforations in the film.
The rotary stabilizer type of soundhead was designed to eliminate
all the objections of the gate type. This was accomplished by the use
of a rotating drum to which is attached the rotary stabilizer elements.
Fig. 1A
Special care was taken to see that the film as it passes through the
soundhead is bent into as wide a curve as possible, insuring that it
would lie flat and all points of the sound track would be in focus. The
free running sound drum shaft is mounted in ball bearings so that
the friction is reduced to such a small amount that it is possible for
the film to drive this drum without appreciable tension. This tension
is so light that the film is never pulled taut except at the start. The
film being in contact with the drum rotates with it. This prevents any
possibility of film scratching. The film assumes a curved path after
leaving the sound drum. The stiffness of the film serves as a compli-
ance which, in conjunction with the mass of the rotating elements,
acts as a filter to eliminate variation in the motion of the film at the
translation point.
The best known expedient for uniform rotation is the fixed flywheel
on a shaft. This, however, is unsatisfactory for control of the drum
shaft because the flywheel would continually hunt or oscillate with the
68 RCA REVIEW
springy film loop in the same manner that a weight suspended from
a coil spring would oscillate under the slightest disturbance.
The theory of the "rotary stabilizer" principle was discovered
several years ago and was later further elaborated and expanded. This
work led to a device for controlling the drum speed which fulfilled the
conditions required for a satisfactory reproducer, namely that the sys-
tem be damped so as to prevent oscillation of the mass of the rotating
system when propelled by a spring for absorbing irregularities in the
motion of the driving mechanism. The practical form of this repro-
ducer soundhead consists of a light case constructed as a short cylindri-
cal casing firmly fastened to the drum shaft. Inside the casing on a
hub forming part of the case, a heavy free floating flywheel is carried
Fig. 2
on a ball bearing. (Fig. 1 -A.) A light oil fills the remaining space
inside the case. The oil acts as a viscous driving medium between the
heavy flywheel and the case. The case is sealed so as to be oil tight.
Any tendency for oscillation between the stabilizer assembly and the
film loop is prevented due to the fact that the energy of the dis-
turbance is dissipated in the oil when there is relative motion between
the casing and the flywheel which does not follow rapid changes in the
motion of the casing because of its inertia. The relative moments of
inertia of the casing and flywheel are approximately one to eight.
To keep the film in proper position on the drum, it passes between
two flanged rollers mounted directly above the drum. To accommodate
film of various degrees of shrinkage, one flange is movable. The fixed
flange is on the sound track side and is known as the guiding roller.
The flange assembly ii also adjustable within limits so that the sound
track may be adjusted to the correct position for being scanned by the
light beam passed through it to a photocell.
The light beam is approximately .001" by .084". It is obtained by
focusing an image of a slit in a diaphragm (five times as large) on
the film. The diaphragm in which the slit is put is illuminated by a
Fig. 3
10-volt, 5 -amp. lamp and a condensing lens. A small collector lens is

mounted in the drum over which the film passes. This directs the light
to a photo -electric cell. This cell is connected through a transformer
of suitable impedance ratio for connecting to a 500 -ohm line to the
amplifier.
To eliminate possible noise due to vibration of the lamps or photo-
cell, the motor is mounted on rubber and the rotating sound drum,
optical system, photocell, photocell transformer and exciter lamp
70 RCA REVIEW
are also mounted on one plate and resiliantly mounted to the main
case of the unit.
(B) AMPLIFYING EQUIPMENT
A current type of amplifying equipment consists essentially of a

fader box (Fig. 2) mounted on the front wall of the projection booth
at each projector, a main amplifier (Fig. 3), a monitor speaker ampli-
fier, a monitor speaker (Fig. 4) and an exciter lamp and loudspeaker
field supply unit (Fig. 5).
The fader system is used to connect the soundhead to the amplifier
at the time the picture changes from one reel to another. It consists
Fig. 4
of a switch for exciter lamp changeover and a 20 -db. variable attenu-

ator pad. This pad serves to preset the volume from the machine to
a predetermined level or to eliminate changes in volume at the change-
over due to different sound level on the film. By this means the pro-
jectionist need not leave his position at the projector to make adjust-
ments of volume after the changeover.
The main amplifier in Fig. 3 is the type used in the larger theatres.
It is so constructed that all parts can be removed from the front for
ease of service. It consists of a voltage amplifier and one or two power
amplifiers. The equipment is completely AC operated and self contained.
Each amplifier is complete in itself including its own rectifiers for power
supply. The voltage amplifier has sufficient voltage amplification to
drive the power amplifiers when the input terminals are connected to
the line from the photocell transformer. Both voltage and power am-
plifiers are novel in construction. Each amplifier consists essentially
of three parts; namely, a vertical panel, a base proper, and a base
support. On the vertical panel attached to the rear of the rack are
mounted the heavy power supply parts such as power transformers,
filter reactors, etc., and the base support. On the base support are
mounted very few parts, as its main function is that of supporting
the main base and the inter-panel cabling. On the main base are
mounted the main amplifier parts and the tubes. This main base is so
constructed that it can be hinged down for inspection and service.
Fig. 5
It can also be completely removed without the use of a soldering iron

as all connections are made to the vertical panel through screw ter-
minals. In the circuits employing large capacitors these are seg-
regated into sections in parallel. Important sections of capacitors and
points of the circuit are fused against possible trouble. Failure of a
section of the filter capacitor which is fused will permit the program
to continue until the end or such time as repairs can conveniently
be made.
An indicator of the Neon lamp type is placed in the plate circuit
of the amplifiers to indicate that this circuit is functioning properly.
The lamp is lighted at all times, failure to light indicates no plate
current to the tubes. The trend in recent years on electrical equipment
72 RCA REVIEW
has been to eliminate all unnecessary controls and meters. AC operated
amplifiers and modern amplifier tubes do not require tube controls or
meters for adjustment. Only one manual control is found on the main
amplifier. This is the master volume control which can be preset and
need not be changed for the entire performance. Experience of projec-
tionists (who are occupied in keeping the show going) with this type
of amplifier in the past years indicates that the design is practical
and capable of meeting the requirements for uninterrupted service.
Fig. 6
The monitor amplifier is self contained with its power supply and
is mounted in some convenient location on the projection booth wall.
It is made to bridge the speech circuit to the stage. It has a sep-
arate volume control to permit adjustment for noise conditions en-
countered in a projection booth. Its output is sufficient to drive
the monitor speaker (Fig. 4) to give adequate sound at each projector
station. Monitor speakers are primarily used to indicate that the
speech circuit to the stage is functioning properly and to give the
operator the proper sound cue for changeover from one machine
to another.
The exciter Iamp and loudspeaker field supply rectifier and filter
unit (Fig. 5) supplies power for the exciter lamp in the soundhead
(10 v. at 5 amp.) and 18 watts at 12 volts for each of the loud-
speakers. Changeover from one machine to another is accomplished by
means of a relay controlled by the control switch at each projector
station.
(C) LOUDSPEAKER EQUIPMENT (FIG. 6)
A recent design of loudspeaker equipment to be installed on the

stage consists of two separate speakers, one of these reproduces the
Fig. 7
frequency range below 300 cycles and the other above 300 cycles. A
dividing network or filter serves to divide the electrical output of the
amplifier accordingly.
The low frequency speaker is made up of two folded exponential
horns, each 40" high by 80" wide mounted one above the other. Each
horn has two 14" cone type driver units. The frequency range is from
40 to 300 cycles as used with the dividing network.
The high frequency speaker is of the exponential horn type. It
consits of a number of small exponential horns, each measuring at the
bell opening approximately 7" x 7 ". These small horns are assembled
into clusters forming the equivalent of a large horn with partitions
74 RCA REVIEW
dividing it into sections. All sections are driven by two speaker

mechanisms through a "Y" throat.
A plain exponential horn has a directional characteristic that
varies with frequency. The higher the frequency the more narrow
the beam becomes. A single exponential horn mounted in place of the
multiple section horn shown in Fig. 6 would give a resultant "bossy"
reproduction to that portion of the audience in seats located well off
the axis, and the reverse would be true for those directly on the axis.
This effect is eliminated if a cluster of small exponential horns are
used. The mouth opening formed by the cluster is spherical in shape.
Fig. 8
Four sizes are used for various types of theatres, each three layers of
the small horns in height, and varying in angle from approximately
521/2° to 105° in clusters of nine, twelve, fifteen and eighteen of the
small horns. The width of the theatre and the acoustical property of
the side walls determine the angle used. The high frequency speaker
operates over a frequency range from 300 to 10,000 cycles. The
300 -cycle crossover was selected as a compromise on high frequency
horn length. Moving it to a higher frequency is a disadvantage from
the standpoint of division of primary speech sounds. The limitation in
depth of the speaker assembly is brought about by the necessity of
flying the speakers on theatre stages where stage presentations are
given and the loss of lines for scenery drops is a problem.
The low frequency driver mechanism (Fig. 7), is a high efficiency

cone type dynamic speaker. Four of these are used per installation.
The high frequency driver mechanism (Fig. 8) is a cone type unit. The
cone is moulded from a fiber sheet and has no seams. It is treated
to make it moisture proof. It has been determined that the strength
per unit weight or mass of the diaphragm for this fiber is greater
than can be obtained with other materials.
The size of the auditorium and its acoustic properties have a
very definite influence on the size of the equipment to be installed and
frequently the acoustical characteristics of the auditorium require
that the characteristics of the equipment be adjusted to compensate
for undesirable effects. It may be necessary, due to high absorption
of the low frequencies to supply additional energy to the low frequency
speakers and vice versa.
It is customary to rate equipment according to theatre seating
capacity and volume of the auditorium. Experience has shown that
the following classification of theatres as to size results in a com-
mercially satisfactory arrangement of equipment:
Cubical contents in cu. ft.
(1) Up to 500 seats 75,000

(2) 500 to 800 seats 120,000
(3) 800 to 1400 seats 200,000
(4) 1400 to 3000 seats 720,000
(5) 3000 up- Special custom equipment
The type of soundhead used is not dependent on the size of the
theatre. For economical reasons the types and sizes of amplifiers and
speaker complements are selected as required for the seating capacity
and size of the theatre. It is desirable that all equipment be installed
under the supervision of factory trained installation engineers who
make the final tests and adjustments of the installation.
SOME NOTES ON ULTRA HIGH FREQUENCY
PROPAGATION
By
H. H. BEVERAGE
Chief Research Engineer, R.C.A. Communications, Inc.
INTRODUCTION
THE propagation characteristics of ultra high frequencies above

30 megacycles have been studied for many years. While many
observations have been made, it is unfortunate that a substan-
tial proportion of these observations have been qualitative only, due to
the difficulty in constructing suitable equipment for quantitative
measurements. It has also been difficult to build transmitters of suf-
ficient power and stability to make possible the quantitative measure-
ment of signals at considerable distances beyond the horizon. It is
the purpose of this paper to review some of the available information
concerning ultra high frequency propagation, including some studies
which have recently been made by engineers of R.C.A. Communica-
tions, Inc. of propagation at various frequencies both within and
beyond the optical distance.
The study of ultra high frequency propagation falls logically into
three divisions, namely, (1) Propagation within the optical distance;
(2) Ground Wave propagation beyond the horizon; (3) Sky Wave
propagation.
PROPAGATION WITHIN THE OPTICAL DISTANCE
The theoretical laws of ground wave propagation over optical paths

are fairly well known. Several excellent papers have been published
on this subject.1 -2 -3 -4-13-14 It has been shown that the received signal
is the resultant of the direct ray and a ray reflected from the ground.
For most practical cases the reflected ray impinges upon the ground
at nearly grazing incidence and is usually reflected at high efficiency
with a 180- degree phase reversal. Consequently, the direct ray and
the reflected ray arrive at the receiving antenna at equal intensity
and nearly out of phase. The phase difference between the two paths
depends upon the location of the transmitting and receiving antenna
and the nature of the intervening ground. For flat ground, Trevor
!6
ULTRA HIGH FREQUENCY PROPAGATION 77
and Carter' have shown that the phase difference for grazing angles
is
4,rah
IF_ (1)
ñ, r
where h is the height of the transmitting antenna, in meters
a is the height of the receiving antenna, in meters
r is the distance, in meters
A is the wavelength, in meters
The direct field E0 from a half -wave dipole is 7(VW/r), where W

is the watts radiated. The received field for grazing angles then
becomes,
7VW 4/rah
E X
r AT
À,
88 VW ah
Volts per meter (2)
A.
r2
From equation (2), it will be noted that for the conditions assumed,
the signal intensity is inversely proportional to the square of the
distance; is directly proportional to the heights of the transmitting
and receiving antennas above ground; and is inversely proportional
to the wavelength. For a given height for the receiving antenna, the
transmitting antenna height must be proportional to the wavelength
to ottain a given signal intensity. For a given wavelength, the signal
intensity will increase directly in proportion to the increase in height
of either the receiving antenna or the transmitting antenna. For
given antenna heights, the signal intensity will be proportional to
frequency. All of these factors are favorable to the use of higher fre-
quencies.
The above simple equation applies only for grazing incidence over
flat land, free from obstructions. If both the transmitting and receiv-
ing antennas are high and fairly close together, as, for example, trans-
mission from the Empire State Building to an airplane, or between
the tops of the Empire State Building and the RCA Building14, the
geometry is such that the difference in path length is no longer a
small fraction of a wavelength, and the simple equation no longer
holds. In general, standing waves occur which may be greatly com-
plicated by reflections from more than one point. The observations
on the signals from the Empire State Building with the receiver in
an airplane over North Beach, Farmingdale and Patchogue, as re-
ported by Trevor and Carter', clearly show and explain these phe-
nomena.
78 RCA REVIEW
If the transmission takes place over sea water, there should be
a marked difference between vertical and horizontal polarization. For
sea water, horizontally polarized waves are reflected nearly 100 per
cent for all angles, and the phase shift changes gradually from 180
degrees at grazing incidence to about 178 degrees at perpendicular
incidence. On the other hand, for vertically polarized waves, the phase
angle and per cent reflection change very rapidly with angle of inci-
dence so that transmission over sea water should be excellent.
Trevor and Carter reported on some propagation measurements
over sea water transmitting from an antenna a few feet above the
water to a motor boat.' They found that the propagation of horizon-
tally polarized waves was extremely poor as compared with vertically
polarized waves, as predicted by the theory given in their paper. On
the other hand, when the antennas are located at considerable eleva-
tions on mountains, as was the case during some observations between
the islands of the Hawaiian group several years ago,5 it was found
that there was no marked difference between horizontal and vertical
polarization, even though the transmission path was mostly over sea
water and distances greater than the optical path were involved.
In applying equation (2), it is obvious that if the transmitting
antenna is directive, either in the horizontal or vertical plane, or both,
the directivity factor should be taken into account, since the equation
was developed on the basis of transmission from a simple half-wave
dipole.
GROUND WAVE PROPAGATION BEYOND THE HORIZON
Comparatively few data are available for determining the laws
of ultra high frequency propagation beyond the horizon. Handel and
Pfister' in a recent paper (published in German) have shown that
the penetration of ultra short wave radiation beyond the range of
optical sight takes place due to both diffraction and refraction. They
state that the field due to diffraction at the earth's surface is inde-
pendent of diurnal and seasonal times. Methods for calculating the
diffraction field together with calculated curves and some measured
values are included in their paper. The calculated diffraction fields
agree very well with the observed fields in most instances, but at
times the observed fields beyond the horizon are shown to be consid-
erably higher than the values calculated from the laws of diffraction.
The authors attribute this to refraction phenomenon, apparently within
the troposphere. The refraction field shows strong variations and
produces an effect similar to fading in short wave reception, whereas,
the field intensities in the diffraction zone are very stable. The authors
point out that the refraction fields appear more frequently and strongly
_--..,
rmm.. .
r=i,ms -C`
ULTRA HIGH FREQUENCY PROPAGATION
PiLìS¡biil®l
ow...
. %
t\
.
79
fee
.e
No
cee
'
Tee
2oe
,
co
.o
111111
90
7e
i ce
MITATME A MIN NI .m
TAYMEEEEM
® -8O .e
[MIMI
Po/ Ts
.19.
.
Le
Z
0
e
ì
,
\ 7
c
s[oA.`%46 r
,
I ' O
.
'
III
i% - .EMEEMEEZy
UIiÏIii3711SJ r
\ailiiiïliiiiriiffl,_iiil
CIi1:/i.
lÌ.7.tFIJlZZz
' J
47eIlLa1
Zï%ImIN
1
E
C
s /
,
.e
é
/c F TH .
9
NEId !Le
-A /0
4 6 7 8 to
T E Ft
30
- e ES
40 50 007000/00 POo
I
Fig. 1
80 RCA REVIEW
in summer than in winter and that the refraction over sea is stronger
than over land.
Ross Hull' has made some very interesting studies of the refraction
field and has shown that there is excellent correlation between signal
intensity and temperature inversion. That is, when warm air masses
exist above colder air masses near the ground, the signals are refracted
down to earth beyond the horizon by the warm air. The existence of
the warm air masses are determined by temperature measurements
with balloons or airplanes. Considerably more data are required before
it will be possible to evaluate these refraction fields or to predict
their frequency of occurrence.
It may be of interest to examine some of the available propaga-
tion data which includes the refraction fields beyond the horizon as
evidenced by fading.
Figure 1 is a typical set of observations made by Mr. G. S.
Wickizer on the signals from the Empire State Building operating on
a frequency of 41 megacycles with about 1200 watts in the antenna.
The transmitting antenna was about 1300 feet above sea level. The
receiving antenna was a dipole mounted on a bamboo pole, the center
of the dipole being 17.6 feet above the ground. By substituting the
above constants in equation (2) with distance as a variable, the curve
marked "Slope 1/D2" was obtained. Beyond the horizon the field inten-
sity falls off faster than the inverse square of the distance. A curve
with a slope proportion to 1 /D3 .s seems to fit the observations fairly
well.
It will be noted that the observed intensities come up to the cal-
culated curve frequently, but seldom exceed it, excepting in a few
cases such as the points taken at Arney's Mount, where the receiver
was on a high hill unobstructed in the direction of the transmitter.
Most of the observations lie between the calculated curve and a similar
parallel curve drawn at 10 per cent of the intensity of the calculated
curve. The average intensity appears to be about one -third of the
calculated intensity. Beyond the horizon, the scattering is probably
largely due to fading of the refraction field, as it was found to be
difficult to check the readings by returning to the same observation
points at different times.
Within the optical distance, the attenuation is probably relatively
high due to large buildings and other obstructions. Burrows, Hunt
and Decino3 have shown that the average fields in the City of Boston
on a frequency of 34.6 megacycles follow the inverse square law
but
average 10 or 12 db below the calculated level terrain values.
Holmes and Turners, on the other hand, have shown that under
some conditions, the observed attenuation in urban areas does not
\ ME;;;.
hy wk..
, ú
cSl3PE' -
4
[á
P...1-
sc.,
Ts'
b
0
F 30o
7e
1 i 1
CUiPVF CoDLCfK r7TE0 POR AN

De
/300 FEET
TRH{Y-sM/TTER hbYTáNN/r /itFA?NT
REçE/VeR
WS r73 RA/7iRTCO . /RO!)
n . /7.6 - ) d°
7e
ce
ñ
PO/Y/73 OBaR v VEO 0 ro
40
.S:APE Os.s 30
th
ÿ 20
/O
D
8
7
6
6
4
P/rv,w9 0RT/ OA' 3

OP
4/ /'lc. z
//Y THE
/REGT/O/Y OEl OOVE.4. oEL .
TRSq/Y SM/T YEiR

/ 7 T . o
1
FMP/RE 54.9r BLDG. /Y. Y.G a
VE.BT/CAL POLIlA?/Z BT/OH .... 7
- 6
.f
_ -9
3
OPTK'NL %fIT// 6.1
.274S7-.9NCE PPOH TRR/YSM/TTER -- /s1/4ES

/00 30a
3 9 ,s 6730/0
I
2 Po .i3 7b .'A 6378 00 ,Zo3
Fig. 2
82 RCA REVIEW
seem to fit the inverse square law very exactly. They also show that
the attenuation increases markedly with frequency, so that 100 mega-
cycles is considerably inferior to 30 megacycles in an urban area,
whereas, equation (2) indicates that the optical path transmission
should be better on the higher frequency in the absence of obstructions.
It should be noted, however, that the transmitting antenna used by
Holmes and Turner was only 190 feet above the ground. Somewhat
different results might have been obtained if the transmitting antenna
had been at a greater elevation so as to clear the obstructing buildings
more effectively. Nevertheless, the survey reported by Holmes and
Turner indicates very definitely that in the presence of obstructions,
the attenuation increases greatly as the frequency is increased.
Figure 2 is a set of observations for an entirely different trans-
mitting condition, but with nearly the same frequency as Figure 1.
The transmitter was located on top of the RCA Building at 30 Rocke-
feller Plaza in New York City. The antenna was about 980 feet above
sea level and was horizontally polarized. The power in the antenna
was about 80 watts. The calculated curve for inverse square law up
to the horizon and inverse 3.6 power law beyond the horizon, again
seems to fit the maximum signal intensities fairly well.
Figure 3 shows the results of observations made by B. Trevor
and R. W. George on the much higher frequency of 91.8 megacycles.
The antenna was a simple half -wave dipole on the roof of the Conti-
nental Bank Building at 30 Broad Street, New York City. The an-
tenna was about 600 feet above the street level. The power in the
antenna was about 50 watts. The antenna was readily adjusted to
radiate either horizontally or vertically polarized waves. The receiving
antenna was a dipole rigged on the roof of a car, the center of the dipole
being about ten feet from the ground.
Substituting the above data in equation (2) gives the curve marked
"Slope 1 /D2." Beyond the horizon a curve with a slope of 1/D5 seemed
to fit the maximum points with the exception of the points measured
on top of hills, as indicated. There was apparently no consistent dif-
ference between the transmission characteristics of horizontal and
vertical polarization over land.
Observations on transmission from the top of the RCA Building
with a frequency of 25.7 megacycles indicate that the signal beyond the
horizon falls off about as the 3.2 power of the distance.
Airplane observations on a frequency of 411 megacycles reported
by Trevor and George9 indicate that the signals fall off approximately
as the 9th power of the distance beyond the horizon.
Figure 4 shows the observed rates of attenuation beyond the hori-

zon plotted against frequency. The factors determined for the four
frequencies fall on a smooth curve, but the data are too few to warrant
much confidence being put in this curve. The curve does indicate,
however, that the attenuation beyond the horizon increases rapidly
for the higher frequencies. Perhaps the attenuation law changes for
/0 /yiL ES /oo
/`// 0i7 TRHNSM/TTE/¢

Fig. 3
increasing distance beyond the horizon, but there are insufficient data
to indicate whether this is so or not. When higher powered trans-
mitters become available, the attenuation laws beyond the horizon
can be determined more accurately.
SKY WAVE PROPAGATION

As the frequency is increased, a point is eventually reached where
the sky wave is not bent sufficiently to come back to earth. This is
an advantage for certain services, such as television, as there is
essentially only one path and multiple images are absent. It is also
possible to duplicate the frequencies at moderate distances without
84 RCA REVIEW
fear of interference. The lowest frequency that will just fail to have
the sky wave returned to earth depends upon several factors. In gen-
eral, the higher frequencies are returned to earth in the early after-
noon. Contrary to what one might expect, there is some evidence that
the high frequency sky waves are transmitted better in winter than in
summer, particularly over the north Atlantic path10. The transmission
also is apparently associated with the 11 -year sunspot cycle. For
example, the high frequency sky waves were getting through quite
frequently during 1927 and 1928 when extensive observations on fre-
quencies above 30 megacycles were first made. Subsequent to 1928,
high frequency sky wave transmission was relatively poor until the
on, pow
iiiiiiiiiiiiiiiiiiiiiiiiiiriiiiiiiiüiii
- _°=UlUUcir;UrcUUiUiUáiUi
=_;i_=_-_ __1®ili%/®11MM®®
ííiíiiìiiiii MOíiííiííiiMOMMOM
morW77V411t!
Immommumn
=w/.íi000E
.îg
N
Fig. 4
spring of 1935. Accordingly, there is relatively little information avail-

able concerning the transmission of very high frequency sky waves.
The long distance transmission on these frequencies is very irregular,
which adds to the difficulty of obtaining consistent data. As the fre-
quency is raised, the sky wave transmission becomes more and more
erratic. Above about 45 megacycles, the sky wave transmission appears

to rarely occur, and when it does occur, it tends to appear as a "burst."
That is, the signal comes up very suddenly, remains fairly strong for
a few seconds or even minutes, and suddenly disappears, perhaps not
to be heard again over that particular path for months. This "burst"
phenomenon was apparently first observed by W. I. Matthews on 50
megacycles over a distance of about 240 miles.'
In May and June, 1935, amateurs reported hearing 5-meter signals
over distances of 900 miles. Amateurs in the vicinity of Chicago
heard several New England amateurs working in the 50 -60 mega-
cycle band, and at least one two -way contact was established." A
similar effect was reported for May 9, 1936.12 On this occasion, sev-
eral two -way contacts were made between East Coast and Middle
West amateurs. Many signals were heard over a period of some three
hours, beginning at approximately 8:30 P.M. Several of the amateurs
reported severe selective fading, with part of the signal dropping out
and the rest remaining.
It seems probable that the transmissions on these occasions were
due to sky waves and not refraction. Probably the suggestion that
the signals were bent down by "unusually heavy sporadic E- region
ionization" is the correct explanation.
From the available information, it would seem that little sky wave
transmission takes place above about 45 megacycles, and that such
transmissions that do occur above 45 megacycles are produced by
unusual ionization conditions which probably will rarely occur. Ama-
teur operators spread over a wide area are in an excellent position to
observe these sporadic transmissions and their observations should be
extremely valuable for indicating the location, duration and frequency
of occurrence of these sky wave transmissions, as well as diurnal and
seasonal effects.
CONCLUSION
There are apparently four mechanisms which may be involved in
ultra short wave propagation. These are (1) combination of the direct
ray and the ray reflected from the ground; (2) diffraction at the
earth's surface; (3) refraction in the troposphere; (4) sky wave
transmission.
The first mechanism is the principal effect within the optical path.
It shows that the signals are attenuated according to the inverse
square law of the distance for grazing incidence within the optical
range. The signal intensity can be calculated by simple equations,
although scattering and absorption, even in open country, tend to
reduce the average intensity to something in the order of 30 to 60
86 RCA REVIEW
per cent of the calculated value. In urban areas, the scattering and
absorption due to buildings, increases the attenuation considerably,
particularly at the higher frequencies. The amount of this increase
in attenuation on the higher frequencies probably depends to a great
extent on the height of the transmitting antenna in relation to the
obstructing objects in its vicinity. More data are required before
the relative performance of various frequencies in urban areas can
properly be evaluated and compared.
The diffraction factor becomes important beyond the horizon. The
diffraction field can be calculated by the methods indicated by Handel
and Pfister'. In addition to the diffraction field, which is believed to
be constant and stable, the refraction field is important beyond the
horizon. The refraction field is variable and produces fading. There
is insufficient information available to calculate the refraction field.
In general, it is definitely known that the attenuation beyond the
horizon increases rapidly as the frequency is increased. The slope of
the attenuation curve that fits the available observations best is indi-
cated in Figure 4. An approximation of the attenuation for a given
circuit may be calculated by using equation (2) up to the horizon
and plotting the calculated points on log -log paper. These points will
lie on a straight line having a slope of 1/D2. At the horizon, another
straight line having a slope determined from Figure 4 should be
drawn. This line will indicate the order for the attenuation beyond
the horizon for a particular frequency. In general, the signal intensity
determined in the above manner should represent the maximum, except-
ing for unusual conditions such as locating the receiver on a mountain
top, abnormal refraction fields, sky wave transmission, etc. The avail-
able data are based on overland transmission, for which case there
seems to be little difference between vertical and horizontal polariza-
tion. Over sea water, vertical polarization is superior to horizontal
polarization, at least for moderate distances with relatively low
antennas.'
The optical distance for flat ground is easily calculated from the
equation:
Distance in miles = 1.22 -/ Height in feet
If the receiving antenna is also at a high elevation, the same equa-
tion may be applied to determine the horizon for the receiving antenna,
and this added to the horizon for the transmitting antenna gives the
total optical path for that particular set of conditions.
Too little is known about sky wave transmission on the ultra high
frequencies. From available information to date, it would seem that
sky wave transmission above about 45 megacycles is too spasmodic

to give much concern. However, it is possible that more frequent
sky wave transmission may be observed at some more favorable phase
of the sun spot cycle.
BIBLIOGRAPHY
"'Notes on Propagation of Waves Below Ten Meters in Length"

Trevor and Carter, Proc. I.R.E., March 1933.
-
2 "Ultra- Short -Wave Propagation"
Proc. I.R.E., March 1933.
- Shellung, Burrows and Ferrell,
3 "Ultra- Short-Wave Propagation : Mobile Urban Transmission Char-

acteristics" Burrows, Hunt, and Decino -Bell System Tech. Journal, 1935.
4 "Some Results of a Study of Ultra- Short-Wave Transmission Phe-
nomena"- Englund, Crawford and Mumford, Proc. I.R.E., March 1933.
s "Application of Frequencies Above 30,000 Kilocycles to Communica-
tion Problems " -Beverage, Peterson and Hansell, Proc. I.R.E., August 1931.
s "Ultra Short Wave Propagation Along the Curved Earth's Surface"
-Paul V. Handel and Wolfgang Pfister, Hochfrequenztechnik Und Elec-
troakustik, Vol. 47, No. 6, June 1936.
7 "Air -Mass Conditions and the Bending of Ultra -High Frequency
Waves" -Ross Hull, QST, June 1935, Page 13.
8 "An Urban Field Strength Survey at Thirty and One Hundred
Megacycles " -R. S. Holmes and A. H. Turner, Proc. I.R.E., May 1936.
9 "Notes on Propagation at a Wavelength of Seventy-Three Centimeters"
-Trevor and George, Proc. I.R.E., May 1935.
19 "The Propagation of Short Radio Waves Over the North Atlantic"
C. R. Burrows, Proc. I.R.E., September 1931.

-
11 "Five -Meter Signals Do the Impossible. Signals Swapped Over 900-
Mile Path; WICBJ Contacts W8CYE ", QST, August 1935, Page 17.
12 "Five Meters Again Shoots the Works ", QST, July 1936, Page 9.
13"A Study of the Propagation of Wavelengths Between 3 and 8

Meters," L. F. Jones, Proc. I.R.E., March 1933.
14 "Ultra High Frequency Transmission Between the RCA Building and
the Empire State Building in New York City " -P. S. Carter and G. S.
Wickizer, Proc. I.R.E., August 1936.
FREQUENCY ASSIGNMENTS FOR TELEVISION
BY
E. W. ENGSTROM and C. M. BURRILL
RCA Manufacturing Co., Camden, N. J.
Foreword-This article is not a report of original work, but is a correla-

tion or synthesis of information pertinent to the subject, available to the
authors within the RCA Services or through published papers. Since the
results of all have been taken into account it has not seemed feasible or
desirable to give credit to individual sources except to mention the article
by H. H. Beverage entitled "Some Notes on Ultra Short Wave Propagation"
appearing in this number of RCA REVIEW, and the bibliography forming a
part of that article. Much credit is due collectively to the many workers in
this field, who have made possible the drawing with reasonable certainty of
the conclusions here stated. The basic plan of any new service must always
be determined by the work of such pioneers, before commercial experience
has made everything plain. Because fundamental plans for broadcast
television are now in the making, it is hoped that this brief article will be
found both timely and interesting.
INTRODUCTION
IN TELEVISION we are concerned with a broadcast type of service in

which adequate coverage of a service area is the prime considera-
tion. The frequency of the radio carrier must be high enough to
permit the relatively wide side bands which are required, and so high
that reflections from the ionosphere will not regularly occur. The
second requirement is necessary to prevent multiple images caused by
reflections from the ionosphere, and to permit duplication of frequency
assignments with reasonable geographical separation of stations on
the same channel. Further, in television we are concerned with a serv-
ice requiring high signal strength at the receiver locations. Thus the
radio carrier must be sufficiently low in frequency to permit generation
of adequate power. The carrier frequency must be low enough so that
the attenuation caused by obstacles on or near the transmission path
is not too great, and so that the shadows cast by obstacles are not too
sharp or too dense. All these considerations point to frequencies above
40 megacycles, and not greatly above 40 megacycles, as necessary if
television is to be practicable in the near future.
A more specific discussion follows. It is very difficult to make posi-
tive statements because our knowledge of the many relevant factors
88
FREQUENCY ASSIGNMENTS FOR TELEVISION 89
is still so incomplete. Much information has been obtained and corre-

lated, but this is only a small sample of what will be required for a
complete and specific analysis. It is believed, however, that this sample
is reasonably representative, and therefore that general conclusions
convincingly deduced from it are reliable. The following analysis is
therefore confidently believed to be correct in its general and major
implications, although time may indicate some error in detail.
PROPAGATION WITHIN THE HORIZON WITH RESPECT

TO THE TRANSMITTER ANTENNA
The theory of direct transmission over a flat earth, taking into ac-
count reflection from the earth, indicates that under certain conditions
the field intensity is inversely proportional to the square of the dis-
tance, is directly proportional to the heights of the transmitting and
receiving antennas, and is proportional to the frequency. For these
relationships to be theoretically valid, the reflection from the ground
must be at grazing incidence, that is, such that the conductivity and
dielectric constant of the ground are not significant factors. The angles
of incidence which may be called "grazing" in this sense- depend on
the frequency. Thus, in any case, the validity of these simple propaga-
tion relationships depends on the geometry of the antennas and earth,
on the constants of the reflecting earth, and on the frequency. Meas-
urements, made for the most part under conditions such that these
relationships were not strictly applicable, have nevertheless indicated
that they are approximately correct for these practical cases, although
scattering and absorption, even in open country, tend to reduce the
average field intensity to somewhat less than would be calculated from
the theory.
In urban areas the scattering and absorption due to buildings re-
sults in additional attenuation and this additional attenuation is the
predominant factor, as far as propagation effects are concerned, in com-
paring the suitability of different frequencies for a local broadcast
service. It increases rapidly as the frequency is increased, so that, for
one set -up investigated, the average field intensity at 30 megacycles
measured at a distance of 5 miles was 4.5 times that obtained at 100
megacycles for the same antenna power. The transmitter antenna was
relatively low over a flat urban area, but was not overshadowed by
higher structures, so that this case was typical for a broadcast service.
A higher transmitter antenna might be thought more favorable to the
higher frequencies, but the high antenna would only remove the region
of high attenuation of the higher frequencies to a greater distance.
A factor likely to be overlooked is that the effective heights of
90 RCA REVIEW
simple and practical receiving antennas tend to be proportional to the

wavelength, i. e., inversely proportional to frequency. Thus, to obtain
a given voltage at the receiver input, the necessary field strength is
proportional to frequency. When this is taken into account, the trans-
mitter power needed is at best the same for higher frequencies, and
on the average, for broadcast coverage, higher frequencies will require
higher powers because of greater attenuation in transmission over
urban areas. In the specific investigation already cited, it was indicated
that several hundred times greater power would be required at 100
megacycles compared with 30 megacycles.
This tendency to lower effective height of antennas for the higher
frequencies may be partially overcome by using arrays which may
occupy no more space than a simple antenna for a lower frequency.
However, the gain due to the directional characteristics of an array is
never as great as its increased physical expanse, so the offsetting effect
is only partial. Furthermore, the selectivity of an array increases with
its gain, so that the use of extended arrays is not likely to be practicable
for television reception. A further objection to a very selective re-
ceiving antenna is that such an antenna could not be used for efficient
pickup of signals over a band of television channels.
For broadcast service in urban areas the signal path to the receivers
becomes complicated. Shadows are cast by obstacles and signal re-
flections occur. It has been shown that shadows become sharper and
more defined as frequency increases. Thus, complete coverage becomes
more difficult with increasing frequency, and a more nearly optical
path is required.
Reflections cause the signal for any particular receiver in a service
area to arrive over a variety of paths of differing lengths and therefore
the corresponding times of arrival will be different. It has been shown
that the effect of this on the reproduced image is determined by the
multiple path structure and the range of video frequencies (width of
side bands) and is independent of the frequency of the radio carrier.
PROPAGATION BEYOND THE HORIZON WITH RESPECT TO THE

TRANSMITTER ANTENNA
Comparatively few data are available for determining the laws of
propagation beyond the horizon. It is well known that frequencies
above 40 mc. fade at points beyond the horizon and that this fading
increases as the distance increases. At and beyond the horizon the
signal intensity falls off faster than the inverse square of the distance.
Such limited data as are available indicate that this increase in rate
of attenuation with distance also increases with frequency from ap-
proximately a 3.6 power at 40 megacycles to approximately a 5 power

at 100 megacycles and to approximately an 8 power at 300 megacycles.
In television it is important that there be no sky wave propagation
primarily because this would produce multiple images and secondarily
because this would affect duplication of frequency assignments at rea-
sonable distances. American amateurs and others have, on a number of
occasions, established communication over long distances by fre-
quencies up to 60 megacycles, apparently as a result of sky wave prop-
agation caused by some sporadic condition in the ionosphere. Because
of practical difficulties, quantitative measurements of such propagation
have not been made. Since there have been no widespread transmis-
sions at frequencies above 60 megacycles, there has been no opportunity
to determine whether such sky wave propagation occurs at these higher
frequencies or not. However, it is our opinion, based on such experi-
ence as is available, that as the frequency is increased, there is a
gradual transition from normal sky wave propagation at about 20
megacycles to a condition of no sky wave propagation at any time at
some frequency above 60 megacycles, and that between these limits the
time of sky wave propagation becomes less and less and more and more
sporadic as the frequency is increased. It is further believed reason-
able to assume that sky wave transmission will not occur at frequencies
above 40 to 45 megacycles in any but very sporadic instances and that
these need not cause concern in the establishment of a television
service.
REQUIRED SIGNAL LEVELS
For frequencies above 40 megacycles natural static is not of prac-

tical importance. Man -made interference is, however, very serious in
urban districts. The major sources of such interference are ignition
systems and apparatus for diathermy. In quiet suburban or rural
districts, noise generated in the receiver may determine the minimum
useful signal. The signal to noise ratio required for satisfactory visual
reception has not been very definitely established by experience, but
for noise not synchronous in any way with the picture system and not
too continuous in character, it is certainly less than is required for
sound broadcasting.
It has been found experimently, using frequencies of 40 to 50 mega-
cycles, that a signal of 1 millivolt is required at the receiver input to
produce an image satisfactorily free from noise generated in the re-
ceiver input circuits. A signal strength of 5 millivolts is required to
overcome ignition interference and to produce a satisfactory image in
an average residential location. Proportionally higher signal levels
are required as the noise interference increases, particularly in areas
92 RCA REVIEW
near disturbing sources, and in congested urban districts. Locating
the television broadcasting station in the center of the area to be
served is a favorable condition in consideration of noise interference.
Corresponding figures for higher carrier frequencies are not avail-
able, because no actual television experience has been had with such
frequencies. Effective noise field strengths are somewhat less at higher
frequencies, but little quantitative information is available with re-
spect to this. Under assumptions as favorable to the use of higher fre-
quencies as could be made, this could only mean that receiver noise
would be the limiting factor, and this would still necessitate an input
of over 1 millivolt.
Few experimental data are available as, to the magnitude of the
interference from another television station on the same frequency
channel which can be tolerated in television reception. Assuming that
the two carriers are sufficiently spaced to prevent audible beats in the
sound channel, say 30 kilocycles, the tolerable signal to interference
ratio will be determined by visual conditions, especially image contrast.
Such measurements as have been made are not directly useful, since
a system having a high inherent noise level was used, but they do in-
dicate, on a conservative basis, that it will be reasonable to assume a
carrier signal intensity ratio of about 100 to 1 for allocation purposes.
GENERAL CONSIDERATIONS
Signal is propagated out to the horizon for broadcast service in a
manner such that for the areas of most interest, the attenuation will
increase with increasing frequency. At and beyond the horizon the
attenuation rises sharply for all frequencies and more sharply with
increasing frequency. Thus the practical limit of service area is the
horizon. Power increases will be useful to the point of providing the
desired signal at the horizon (and naturally at locations of high noise
interference within the horizon). For a given frequency the power re-
quired to produce a given signal input to a receiver at the horizon with
respect to the transmitter antenna is approximately constant for all
transmitter antenna heights assuming flat unobstructed ground.
A desired condition is a radiation pattern circular in the horizontal
plane (or of proper directivity characteristics for the transmitter
location and service area) . Since sky wave transmission is not desired
and is not present it is important to concentrate the energy into low
angles in the vertical plane so as to obtain a power gain in transmis-
sion. This may be done for either vertical or horizontal polarization,
but present known simple structures are most effective for horizontal
polarization. For broadcast service, in terms of signal intensity, there
appears to be no advantage in one polarization over the other (ex-

cluding transmission over sea water). The matter of signal reflec-
tions in and around building has not been fully investigated, but again
there appears to be no advantage in one over the other. In considering
noise interference, complete data are lacking, but some experience has
indicated a slightly lower noise level for horizontal polarization.
APPARATUS CONSIDERATIONS
Present vacuum tube and transmitter circuits place limits on the

power levels obtainable for television transmission. The conditions of
band width and high signal intensity are severe. Greater powers call
for larger dissipating surfaces in the output tubes-affecting dimen-
sions already too large for best efficiency even at 40 megacycles. As the
frequency increases smaller dimensions are necessary resulting in lower
powers, whereas propagation conditions call for greater powers to pro-
duce the same signal intensity. For the present experimental tele-
vision band of 42 to 86 megacycles, higher powers may be obtained
over the lower portion of the band and lower powers over the upper
portion. The present specific power limitations will be modified as the
technique advances. However, the time when power in tens of kilo-
watts at frequencies over 100 megacycles will be practicable appears
a long way off. Coupled with this is the uncertain practicability
of a broadcast service of the present basic type above 100 megacycles
in view of the higher attenuation in urban territory which may indi-
cate impracticably high output powers. These higher frequencies ap-
pear more suited to point -to-point service than to broadcast service
in urban areas.
CONCLUSIONS
From the foregoing it is concluded that the television band should

start at a frequency between 40 and 45 megacycles. A frequency of
42 megacycles was recommended by Radio Manufacturers' Associa-
tion as the lower limit, and is satisfactory from considerations of lack
of sky wave, modulation with video frequency band, and propagation
characteristics. The upper frequency limit is determined by the num-
ber of six megacycle channels required. A proposal of 42 to 90 mega-
cycles has been made. This is satisfactory with respect to propagation,
apparatus, and distribution of channels at reasonable distances.
BEHIND THE SCENES AT TWO NOTABLE
BROADCASTS
Nov. 8, 1936 -GEORGE MCELRATH
Operating Engineer
Nov. 11, 1936-G. O. MILNE
Eastern Division Engineer
WO of the most interesting special events programs in the

history of radio were broadcast on November 8th and 11th as
features of NBC's Tenth Anniversary celebration week. Nearly
every technical development introduced since broadcasting started in
1922 was utilized.
The November 8th program included pick -ups from two streamline
trains in motion on different continents, from a submarine, Pikes
Peak, a modern tunnel, a Coast Guard ship, Police radio, a six -day
bicycle race, the Golden Gate Bridge and a squadron of Navy planes
flying over San Diego. The November 11th program demonstrated the
feasibility of communication between two airplanes in flight and indi-
viduals on the ground, connected by transoceanic radio facilities.
These two programs demonstrated the usefulness of radio over short
and long distances and furnished a splendid example of the complete
coordination of facilities at widely separated points.
NOVEMBER 8TH PROGRAM -3:15-4:00 P.M., E.S.T.

Equipment was tested time and again during the three weeks preced-
ing the program and when the broadcast started at 3:15 P.M. the
engineers were sure that every technical item was functioning properly
and everything was in order to present a perfectly timed program to
the radio audience. Printed herein is a "Cue and Timing Sheet" show-
ing the program location, duration of time on the air, the scheduled
time to turn control over to the following point and the all important
"cues" by which the technical facilities were rearranged to feed that
portion of the program to all stations on the network. These instruc-
tions also served to give announcers complete details on the next portion
94
BEHIND THE SCENES AT TWO NOTABLE BROADCASTS 95
of the program ; to start the program instantly or to await a given num-

ber of seconds to permit the switching of program circuits. This Cue
Sheet is presented for the purpose of giving the reader a backstage
view of a program tool used by broadcast personnel:
David Sarnoff (seated) participating in the four -way broadcast. William

Burke Miller (standing) in charge of program arrangements.
TIME TO TURN
PICKUP TIME ON AIR OVER TO NEXT POINT
1) New York Studios. 4 Min. 3:19:00

Cue by New York Studio Announcer to turn over to Cleveland
will be: "TAKE IT AWAY CLEVELAND." Switching on this
turnover will be instantaneous.
2) Cleveland Police Short Wave. 3 Min. 30 Sec. 3:22:30

Cue from Cleveland Announcer to turn over to Navy Subma-
rine will be: "NOW UNDER THE SEA TO THE NAVY
SUBMARINE S -20."
3) U. S. Navy Submarine S -20. 4 Min. 3:26:30

Cue from Announcer aboard U. S. Navy Submarine S -20 to
turn over to Pikes Peak will be: "FROM THE BOTTOM OF
THE SEA TO THE TOP OF PIKES PEAK." Switching
from the U. S. Navy Submarine to Pikes Peak will be on a
15- second basis. Announcer at Pikes Peak will GO AHEAD
15 seconds after cue is given.
96 RCA REVIEW
TIME TO TURN
4) Pikes Peak. 2 Min. 55 Sec. See Note
Denver Announcer on Pikes Peak will conclude with the fol-
lowing: "OUR SPECIAL TENTH ANNIVERSARY PRO-
GRAM WILL CONTINUE IN JUST A MOMENT. THIS IS
THE BLUE NETWORK OF THE NATIONAL BROAD-
CASTING COMPANY." Denver Announcer on Pikes Peak
should end exactly at 3:29:35 and Announcer in studios will
ring chimes at exactly 3:29:40.
5) Boston Streamline Train and

Berlin Streamliner. 4 Min. 3:34:00
Announcers on Boston -Providence Streamline Train will GO
AHEAD at exactly 3:30- facilities will be all set up so that
at exactly 3 :30 Boston Announcers will start two -way conver-
sation with German Announcers speaking in English aboard
Berlin train. Boston Announcers will start with phrase:
"CONTINUING WITH NBC'S TENTH ANNIVERSARY
SPECIAL EVENTS PROGRAM WE ARE NOW SPEAKING
TO YOU ABOARD THE NEW HAVEN RAILROAD
STREAMLINE TRAIN `THE COMET'." "HELLO GER-
MANY." Then continue with remarks. Cue from Boston An-
nouncer aboard "The Comet" to turn over to Pittsburgh will be :
"WE TAKE YOU NOW TO OUR ANNOUNCERS STA-
TIONED DEEP DOWN IN A COAL MINE IN PITTS-
BURGH." Switching on this turnover will be on a 5-second
basis.
6) Pittsburgh Coal Mine. 3 Min. 3:37:00

Pittsburgh Announcer in mine will GO AHEAD 5 seconds
after cue is given. Cue from Pittsburgh Announcer to turn
over to NBC mobile unit on Fifth Avenue will be: "WE NOW
CONTINUE FROM NBC'S STREAMLINE MOBILE UNIT
WHICH IS SPEEDING UP FIFTH AVENUE, NEW YORK."
Switching on this turnover will be on a 5-second basis.
7) NBC Mobile Unit on Fifth Avenue. 3 Min. 3:40:00

Announcer on NBC mobile unit will GO AHEAD 5 seconds
after cue is given by Pittsburgh Announcer. Cue from New
York Announcer aboard NBC mobile unit to turn over to bike
races in Chicago will be: "NOW TO THE SIX -DAY BICYCLE
RACES IN CHICAGO." Switching on this turnover will be
instantaneous.
8) Six-Day Bike Races in Chicago. 3 Min. 3:43:00

Chicago Announcer at bike races will give following cue to
turn over to U. S. Army tanks: "FROM CHICAGO WE JUMP
TO THE NATION'S CAPITAL WHERE NBC ANNOUNC-
ERS ARE SPEEDING ABOARD THE LATEST U. S.
ARMY TANKS." Switching on this turnover will be
instantaneous.
TIME TO TURN
9) Army Tanks Outside Washington. 4 Min. 3:47:00
Washington Announcer speaking from U. S. Army tanks will
give the following cue to turn over to Coast Guard: "OUR
TENTH ANNIVERSARY SPECIAL EVENTS PROGRAM
CONTINUES FROM A COAST GUARD CUTTER IN LONG
ISLAND SOUND." Switching on this turnover will be
instantaneous.
10) U. S. Coast Guard Cutter. 3 Min. 3:50:00

Announcer aboard Coast Guard Cutter will give the following
cue to turn over to the New York Mid -Town Tunnel: "FROM
LONG ISLAND SOUND WE HOP TO THE NEW YORK
MID -TOWN TUNNEL." Switching on this turnover will be
instantaneous.
11) New York Mid -Town Tunnel. 3 Min. 3:53:00

Announcer in New York Mid-Town Tunnel will turn over to
San Francisco with cue: "WE CONTINUE FROM SAN
FRANCISCO." Switching on this turnover will be on a 15-
second basis.
12) Golden Gate Bridge. 3 Min. 3:56:00

San Francisco Announcer on Golden Gate Bridge will GO
AHEAD 15 seconds after turnover cue has been given. San
Francisco Announcer will give following cue to turn over to
U. S. Navy planes flying over San Diego: "TO CONCLUDE
OUR SPECIAL PROGRAM WE TAKE YOU NOW TO A
SQUADRON OF NAVY PLANES FLYING OVER SAN
DIEGO." Switching on this turnover will be instantaneous.
13) U. S. Navy Planes Over San Diego. 3 Min. Sign Off

San Diego Announcers will sign off the program at exactly
3:59:35 with the cue: "THIS IS THE BLUE NETWORK
OF THE NATIONAL BROADCASTING COMPANY." An-
nouncer in San Francisco studios will ring chimes at exactly
3 :59 :40.
Standby announcers will be located in the studios at the following

places:
Radio City Studios . NewYork
WTAM Studios . . . Cleveland
KOA Studios . . Denver
WBZ -WBZA Studios . . Boston
KDKA Studios . . . Pittsburgh
Chicago Studios . . NBC Chicago
Washington Studios . NBC Washington
.
San Francisco Studios NBC San Francisco

.
Hollywood Studios . NBC Hollywood
98 RCA REVIEW
In case any pick-up on the program fails to come through, the an-
nouncer who is standing -by in his studios for that particular point
will give the cue to turn over to the next point. All points on the pro-
gram should be standing -by throughout the program.
These broadcasts were planned through to completion by the en-
gineer in charge of technical operations and a member of the Program
The streamline "Comet" between Boston and Providence in two -way

conversation with the "Flying Hamburger," D -12, Germany.
Department in accordance with the procedure which is briefly described

in the following outline:
The director of the program first approached various organizations
to ascertain if they could participate in the broadcasts. Before that
inquiry was made, however, the idea was discussed with the Engineer-
ing Department to determine the availability of apparatus and if it
would be practical to connect the program to the network of stations.
The Operating Engineer then contacted the engineers in charge of
each NBC office, advising them of the program requirements and
requesting each one to complete tests to insure the success of the

program. In due time affirmative reports from all points had been
received and technical arrangements were complete. A final check was
made prior to the broadcast confirming all technical arrangements.
This program opened at 3:15 P.M. with an orchestral program in
Radio City studios and was sent over the network to the associated
stations in the conventional manner. At 3:19 P.M. the announcer in
the New York studio gave the cue to switch to Cleveland.
Arrangements had been made with the Cleveland Police Department

for technical tests using their short -wave equipment between head-
quarters and a cruising car in the city equipped for two-way conversa-
tion. A contact was established between the Cleveland and Massillon
Police Radio Stations demonstrating effective communication facilities
with highway -patrol law- enforcement agencies in and between nearby
cities. Three wire -lines were installed between the Cleveland control
room and police headquarters. One was a private -line communication
circuit and one was a feed-back or cue circuit connected to the network.
The announcer at police headquarters received the cue to start his
program over the latter circuit. In this case, the cue originated with
the announcer in the New York studio who signed off the preceding
portion of the program. The remaining circuit was the program line.
The conversations between police and the announcer at headquarters
were mixed together in the police apparatus and an audio tap was
furnished which was fed to the control room in Cleveland over the
program circuit. When the announcer in the New York studios gave the
cue the network circuit was opened in the Cleveland control room and
the program from police headquarters was sent to all broadcast sta-
tions. At 3:22:30 P.M. the announcer in Cleveland gave the cue
transferring the control to the U. S. Submarine S -20.
In New York two radio channels had been established between the
top of the RCA Building and the submarine. A 100-watt, ultra -high-
frequency transmitter sent program information and cues to the sub-
marine. A 25 -watt UHF transmitter aboard the submarine supplied a
program channel to the RCA Building. Wire lines from the Radio City
Control Room to the top of the building connected the cue trans-
mitter to the network carrying the program. One of these circuits,
connected to a receiver, extended the radio channel from the submarine
to the network. At 3:26:30 P.M. when the proper cue was given the
100 RCA REVIEW
network facilities were opened at Radio City and the program from
the submarine was connected to the network. This same 100 -watt UHF
transmitter also furnished cues to the mobile unit on Fifth Avenue.
It was necessary for the announcer and engineering personnel to
board the submarine in Brooklyn Navy Yard at 9:00 A.M. in order
to place the submarine in position for the broadcast and complete
final radio tests.
Fifteen seconds intervened between the end of the submarine pro-
Senatore Marconi's yacht "Elettra" with which communication

was established near Genoa.
gram and the start of the program from Pikes Peak. This time was
required to reverse line repeaters between Denver and Chicago so that
the program could be connected to stations east of Denver. Messrs.
Peregrine and Rorher, with two experienced mountain climbers left
Colorado Springs at 11:00 A.M. Saturday, November 7 and plodded
through four feet of snow to erect a UHF 1 -watt transmitter on Pikes
Peak for the program channel. A 25 -watt UHF transmitter was
installed at Colorado Springs as a cue channel. Receivers at both
locations completed the two -way communication circuit. One might
ask why the higher powered transmitter was not used as a program
channel. Storm conditions and snow drifts near the Peak prevented
cars from reaching the top, and the weight of the 25 -watt transmitter
was too great to be carried through snow drifts by the engineers. Two
wire -lines connected Colorado Springs and the Denver control room;
one carried the network program and the other was connected to
the receiver at Colorado Springs, thus a program link from the
Peak to the network was made. Fifteen seconds after the announcer
on Pikes Peak received a cue from the submarine he started the pro-
gram which was sent to associated stations east and west of Denver.
At the end of this portion of the broadcast another 15- second interval
was required to re- establish the line repeaters between Denver and
Chicago on a normal basis, thus providing stations west of Chicago
with the program service to follow. This was not noticed by the radio
audience as the reversal was made during the 20- second interval at
3:30 P.M. while associated stations were making their local station
announcements in accordance with the Federal Communications Com-
mission regulations.
The most complicated technical portion of the program followed:

a two -way conversation between two streamline trains in motion
The New York, New Haven and Hartford "Comet" in the U. S. and
-
the "Flying Hamburger" D -12 in Germany. The "Comet" was travel-
ing between Boston and Providence and the "Flying Hamburger"
between Berlin and Hamburg. An R.C.A. Communications transoceanic
short -wave channel from Rocky Point, L. I., to Reichs Post Berlin,
Germany provided one side of the two -way radio link between conti-
nents. Another short -wave channel from Reichs Post to Riverhead
provided the other side of the two -way communication circuit. Two
wire -line circuits from Berlin to a point near the speeding "Flying
Hamburger" and two radio links from this point to the "Flying Ham-
burger" completed the facilities on the European Continent.
The radio channel from Reichs Post to the receiving station at
Riverhead was extended directly to WBZ, NBC Boston outlet, via the
control room in Radio City. Existing wire lines from WBZ's control
room to the WBZ and WIXK transmitters completed this channel.
Receivers installed on the "Comet" tuned to either of the above sta-
tions permitted those on board to hear the conversation from the
"Flying Hamburger" in Germany. A 25 -watt UHF transmitter on the
"Comet" sent a signal to four receivers along the 15 -mile right -of -way
between Boston and Providence. These receivers were connected to
a wire line to the Rocky Point transmitter via the control room in
Radio City and this completed the channel from the "Comet" to the
"Flying Hamburger." Thus the two -way channel was established
102 RCA REVIEW
between the two speeding trains on different Continents through radio
and wire facilities.
Both sides of the conversation were mixed in the NBC New York
control room and connected to the network at that point. German
broadcasters were so intrigued by this program that they requested
us to repeat the two -way conversation in German especially for German
listeners. This performance took place as the United States listeners
were hearing the Pittsburgh portion of the NBC program.
At this point after the cue by the Boston announcer aboard the
"Comet" the program was transferred to the Carnegie Institute Model
Senatore Guglielmo Marconi aboard his yacht "Elettra."
Coal Mine in Pittsburgh, Pennsylvania. The engineering staff at

KDKA had installed radio transmitters and receivers in the mine
which were connected to the network via their control room over local
telephone facilities as described in similar cases above. After a descrip-
tion from the mine and interviews with miners, the Pittsburgh
announcer turned over to the NBC mobile unit cruising along Fifth
Avenue in New York City. A 25 -watt UHF transmitter and receiver
on the mobile unit was connected to the network through the receiving
station and the 100 -watt UHF cue transmitter on the RCA Building
in the same manner as the submarine broadcast outlined above.
Next the program was transferred to the Chicago Stadium, Chicago.

The program originated from a .2 -watt microwave pack transmitter
strapped to "Torchy" Peden's back. He was a participant in the six-
day bicycle race. A low powered UHF cue transmitter at the track
side and receiver strapped to the bicycle completed the local two -way
radio circuit inside the stadium. These facilities were attached to
local wire lines and connected to the network through the NBC control
room in Merchandise Mart, Chicago.
From Chicago to Washington the U. S. Army gave the radio audi-
ence a spectacular show. Signals from Army equipment in the new
reconnaissance cars maneuvering in Potomac Park and traveling at
high speed along roads entering the District of Columbia were inter-
cepted by NBC engineers and relayed to the network through the
control room in the National Press Building.
At this point the broadcast hopped immediately to a U. S. Coast
Guard Cutter. Engineers and the announcer, with equipment, reported
aboard the U. S. Coast Guard Ship "Ponchartrain" in New London
Harbor at 9:00 A.M., and the ship proceeded on a routine control
cruise. A 50 -watt intermediate frequency radio transmitter aboard the
vessel transmitted the program to the NBC control room in the RCA
Building through the R.C.A. Communications Receiving Station at
Riverhead, L. I. The program was transmitted via an RCA Rocky Point
transmitter and it was from this source that the announcer aboard
the ship received his cue to start his portion of the broadcast. The
Coast Guard contribution to the program was connected to the net-
work through the Radio City control room. The broadcast now shifted
to the New York Midtown Tunnel. A low- powered UHF transmitter
attached to the mid-town tunnel end of a feedback circuit from Radio
City carried the switching cue from the Coast Guard "Pontchartrain"
to the announcer inside the tunnel. He spoke into a microphone im-
bedded in the front of a box 6x8x12 inches that housed a microwave
transmitter which radiated the program signal to a receiver at the
mouth of the tunnel connected to a program circuit to Radio City
and then to the network. Fifteen seconds elapsed between the closing
cue from the mid-town tunnel and the start of the program from
Golden Gate Bridge, San Francisco. During this interval repeaters
on the transcontinental circuit were reversed so the program could
be broadcast over stations east of Salt Lake City. A UHF 1 -watt pack
transmitter was installed on top of Golden Gate Bridge and the
announcer's description was broadcast to a receiver located in the
104 RCA REVIEW
U. S. Navy Receiving Ship Observation Tower on Yerba Buena Island
in San Francisco Bay. The program signal was then sent to the San
Francisco control room over a wire line where the network connection
was made. KGO, San Francisco, carrying the network program, pro-
vided the announcer on the bridge with a starting cue from the New
York mid-town tunnel.
Next the program was switched to a fleet of new U. S. Navy planes
flying over San Diego, California. The technical equipment on the
Robert Jardillier, French Minister of Communications, in four -way

conversation from airplane.
ground was supplied by station KFSD, and in the air by the Navy.
The cue from Golden Gate Bridge was transmitted over regular net-
work facilities and over the cue transmitter on the ground in San
Diego to the Commander of the planes in the air. His ship was
equipped with transmitting and receiving apparatus and his voice was
transmitted to a receiver on the ground which was mixed with the
announcer's voice and sent to the network through KFSD's control
room. A two -way conversation could be carried on by the KFSD
announcer and the Navy flier by mixing the audio and radio channels
at the field program originating location.
The program ended with a roar of motors from the sky, picked up
by sound -effect parabola microphones on the ground, and the San Diego
announcer gave his cue for the NBC San Francisco office to ring
chimes, placing the network facilities in readiness for the program to
follow and thus ended one of the most intricate special events programs
ever presented by NBC.
NOVEMBER 11TH, PROGRAM 2:15 -2:45 P.M., E.S.T.

Another special broadcast, notable from the standpoint of world
coverage, occurred at 2:15 P.M. on November eleventh. At this time
visiting radio executives, enroute by air in two planes, from Buffalo
to Washington, carried on an exchange of greetings with Mr. David
Sarnoff, President of the Radio Corporation of America, in his office
in the RCA Building in New York City, and Senatore Marconi, on
his yacht "Elettra" near Genoa, Italy. Robert Jardillier, French
Minister of Communications, and Maurice Rambert, president of
the International Broadcasting Union, were the principal speakers
from the air. Dan Russell was the announcer in the air who also
acted as master of ceremonies for the broadcast. The program was
carried by the NBC Red Network in the United States and was also
relayed by NBC and RCAC short -wave transmitters to Europe where
it was re- broadcast over regular broadcast stations in Germany, Den-
mark, Austria, France, Italy, Switzerland, and Czechoslovakia. Radio
representatives of all these countries were aboard the two planes and
after the broadcast had been completed, several of them spoke in their
native language to people in their home countries.
The program, starting at 2:15 P.M. was opened by six minutes of
an international medley played by an orchestra in the NBC Radio City
Studios. This was followed by a five -minute description of the occa-
sion by Dan Russell from American Airways plane No. NC16005. At
the end of this description, Mr. Sarnoff called the plane and spoke
to Mr. Rambert who was aboard. A two -way conversation followed
at the end of which Mr. Sarnoff called in Senatore Marconi. The
next development was a three-way talk between Messrs. Sarnoff, Mar-
coni and Rambert. Then Mr. Sarnoff called in Mr. Jardillier who was
aboard the American Airways plane No. 16030. The peak of the pro-
gram was reached by the four -way contact then held between the gen-
tlemen in the two planes, Mr. Sarnoff in his office and Senatore Marconi
on his yacht. At the conclusion, an orchestra in the studios completed
the balance of the thirty- minute broadcast.
106 RCA REVIEW
Technically, it might appear that such a broadcast would not be

difficult, particularly with the facilities and equipment which are avail-
able today. However, the numerous circuit connections involved and
the many radio links necessary, coupled with the unusual number of
unforeseen complications that developed, made this an outstanding
technical broadcast that could not possibly have been arranged ten
years ago.
The transmitters used on the planes were two standard sets, WIEO
operated on 2102 kilocycles, fifty watts 100 per cent modulated on
Jack Hartley and Dan Russell (announcer) aboard one of the

airplanes in the four -way contact broadcast.
plane NC16005, and WIEW, twenty -five watts, 2758 kilocycles on NC-
16030. These transmitters are crystal controlled and operate from
twelve-volt storage batteries through dynamotors supplying the re-
quired plate voltages. Messrs. Sturgell and Peck, field engineers,
operated WIEO with Messrs. Wies and Whittemore at WIEW. These
men are field engineers assigned to the special events technical group.
Two receivers were placed on the roof of the RCA Building to intercept
the signals from the planes. These receivers and associated amplifiers
were manned by two more men from this special group, Mr. Wilbur
and Mr. Campbell. The output of each receiver and its amplifier was
fed by a separate line to the main control point. The cue transmitter
utilized to talk to the planes from the central control point located in
the master control room at Radio City was W10XGC. This is a new
crystal controlled 25 -watt transmitter operated on 40.6 megacycles and
located on the roof of the RCA Building. This transmitter was used
to line up the plane circuits and also carried the program back to the
planes with WEAF, the local outlet, as a secondary means of hearing
all other points during the actual broadcast.
Aboard each plane were two receivers, one ultra-high tuned to
W10XGC and one tuned to WEAF. The output of either of the receivers
was fed to a multiple with taps at each seat. At each seat two pairs
of headphones were provided for the passengers so that all the men
on each ship heard the complete program. Incidentally, there were
twenty -two passengers on each plane. The ship's antenna was used
for receiving and a special trailing wire for transmitting.
A set of audio equipment, in duplicate, was installed in Mr. Sarnoff's
office on the fifty -third floor of the RCA Building. The output of his
microphone amplifier was fed to the mixing and switching panel in
the master control room. A return circuit was provided to feed back
all the other originating points. This was terminated in several pairs
of phones, one of which was worn by Mr. Sarnoff and one by Mr.
William Burke Miller, who was in direct charge of all program arrange-
ments. Here also were two field engineers, Messrs. Davis and Jackson
to handle the technical facilities.
Senatore Marconi's portion of the show originated from aboard
his yacht in the bay at Santa Margherita near Genoa. His voice was
fed by local radio and wire line via Genoa to Rome and then trans-
ferred to a radio link to R.C.A. Communications at Riverhead. From
here it came by wire line to the central control position in Radio City.
The return circuit was via wire line to RCAC transmitters at Rocky
Point, radio to Rome and wire line and radio back to the "Elettra ".
All preliminary testing on this portion of the circuit was handled by
RCAC. In arranging all these facilities provision was made so that
during the preliminary test period the engineers could work with the
two planes without interrupting the circuit between Mr. Sarnoff and
Mr. Marconi. At five minutes before program time they were all to
be tied togethere. Here the unpredictable completely disrupted our
plans and almost ruined the broadcast.
The trip was planned so that the planes would leave Newark Air-
108 RCA REVIEW
port at 7:30 A.M. and fly to Buffalo where the guests were shown
Niagara Falls both from the air and ground. Then a short trip into
Canada by car for a greeting with Mr. Howe, the Canadian Minister
of Communications, lunch and a return flight via New York direct
to Washington. Because of heavy passenger traffic, the planes were
not made available until after midnight and the installation of all
equipment was not completed on both planes until approximately 6:00
A.M. The men then had about an hour's sleep on some blankets in the
Maurice Rambert, President of the International Broadcasting Union,

speaking from the airplane in the four -way contact broadcast.
hangar. Both planes left at 7:30 A.M. Arrangements had been made
to have both sets put on the air as soon as the trailing wire could be
released. Both planes were to tune in each other and conduct tests
between themselves on the way to Buffalo. Any trouble could then
be cleared while the ships were on the ground at Buffalo. Everything
worked beautifully and all pointed to a successful show. Antennas were
reeled in and everything set for the return.
Because of the planes' speed of 180 miles per hour and the limited
range of our sets on the only frequencies available for this service,
the timing of the ships' departure and therefore arrival at New York
had been very accurately set so the planes would be approaching New
York when they went on the air and still be within range at closing time.
The transmission range was limited to only fifteen minutes because
at 180 miles per hour this gave a usable radius of only about 25 miles
either side of New York. The first upset came when the ships arrived
over New York about 15 minutes too early. When they got up at
Buffalo on correct schedule, they found a 45 -mile tail-wind, which
brought them into New York at over 220 miles per hour. Secondly,
this same high pressure area causing the winds, dropped the tempera-
ture well below the freezing point. Results, another freak to overcome.
The trailing wire antenna was fed through the floor of the ship
from a special reel in each plane. The sudden drop in temperature
had frozen the fish or weight on the end of the antenna, to the ship.
So after getting up in the air on the return trip, neither ship could
get out its transmitting antenna. After about thirty minutes of try-
ing, one fish was loosened, antenna dropped and WIEW came on the
air. WIEO, however, never got on the air until they were almost over
New York City. It was finally necessary to cut a small hole through
the ship's skin, just above the fish, in order to pry that one loose. In
the meantime, WIEW was reporting itself over New York, but couldn't
imagine what was happening to WIEO. Heavy sighs of relief when
WIEO finally came through. It was still only 2:10 and the ships weren't
due on the air until 2 :21. At their rate of speed they would have been
almost to Philadelphia and well out of range before the show was
completed. So about 2:12 they were requested to circle New York a
few times before heading south.
Simultaneously, there was trouble from a different quarter. RCAC
had lined up the circuits between New York and Santa Margherita,
and transmission in both directions was excellent. At 1 :55 P.M. all
tests were completed with that point.
However at 2:00 P.M. one of those intense, irregular magnetic
disturbances which have come to be known as the Dellinger effect,
set in and out went the Rome circuits. RCAC jumped in, put into
service two additional transmitters at Rocky Point and shifted fre-
quencies with Rome until they found two on which speech was still
intelligible. All of this took place between 2:00 and 2:15 P.M. while
we were still concerned about the planes.
Despite these irregularities the program started on schedule with
the orchestra playing in the studio. Testing continued via the cue
circuits, however, and we succeeded in getting Mr. Sarnoff and the
planes together when suddenly, at 2:16, Senatore Marconi's voice was
cut off and Rome advised that the line had failed between Genoa and
Rome Quick orders were passed to the other three points not to call
!
110 RCA REVIEW
in Senatore Marconi in order to eliminate the possible ghastly pauses
on the air which, to the uninitiated, sound like unsatisfactory facili-
ties or improper planning.
The orchestra completed its number and just as Dan Russell started
talking from his plane, at 2:21, the circuit to the "Elettra" came up
O.K. Orders again passed to all points, to that effect and the show
went on as planned. Results -on the broadcast circuits, a perfectly
synchronized performance, marred only by slight unintelligibility on
the Rome circuit caused by the magnetic storm. Thus despite all the
technical difficulties encountered by the engineers behind the scenes,
the public was given another example of complicated broadcasting
technique.
APPLICATIONS OF VISUAL -INDICATOR TYPE
TUBES
BY L. C. WALLER
RCA Radiotron Div., RCA Mfg. Co., Harrison, N. J.
HE visual- indicator tube, known as the 6E5, originally designed

as a visual tuning indicator for receivers, has characteristics
which make it admirably suited for use in many types of elec-
tronic devices where an indication of voltage (or current) maxima or
minima is desired. In addition to indications of maxima or minima, the
6E5 can be utilized for quantitative measurements in devices which re-
quire calibration or which operate on the voltage-balancing principle of
CATHODE
LIGHT
SHIELD
- FLUORESCENT
-----COATING
RAY -
CONTROL
ELECTRODE
TARGET TRIODE
-PLATE
TRIODE CATHODE
GRID-
Fig. 1- Cutaway of the 6E5 illustrating its electrode structure.
the "slide- back" vacuum -tube voltmeter. A new type of visual- indicator
tube, not so well known at present because of its newness, is the 913.
This midget, low- voltage, high- vacuum, cathode -ray tube, shown in
Figure 2, is suited for numerous oscillographic applications where the
use of a larger cathode -ray tube might be impracticable. Applications
of both the 6E5 and the 913 will be described.
OPERATING PRINCIPLES OF THE 6E5

The cutaway drawing of the 6E5 in Figure 1 shows its electrode
structure. Essentially, the tube consists of a high -mu triode above
which is located a cone -shaped fluorescent target and a blade -like ray -
control electrode ; the latter is internally connected to the triode plate.
111
112 RCA REVIEW
The simple circuit of Figure 3 illustrates the operation of the 6E5.

With zero bias on the triode grid, the fluorescent pattern caused by
electrons striking the coated target will have a shadow sector of about
100 degrees, as shown in Figure 4. The "electronic shadow"
is cast
on the target because the ray -control electrode is at a negative potential
with respect to the target; the bias on the ray control is caused by
the
IR drop across the plate resistor (R2) when triode plate current exists.
When the triode -grid bias is changed from zero to some negative
value by means of R1, the triode plate current decreases, the voltage
drop across R2 decreases, the ray-control electrode becomes less nega-
tive with respect to the target, and the shadow sector closes to a smaller
angle. With about 6 or 8 volts of negative bias on the triode grid, all
plate current through R2 is cut off and the shadow sector is reduced
Fig. 2 -The new midget, low-voltage, high- vacuum, cathode -ray tube.
to a narrow, dark line, as illustrated by the second part of

Figure 4.
Slightly more bias on the triode grid may cause the pattern to
close
completely or even to overdose; in the latter case, the dark
line may
change into a narrow, luminous line having greater brightness
than
the remainder of the pattern.
Because the variations of the fluorescent pattern on the target
are
controlled by the negative bias on the triode grid, it is apparent
that
the 6E5 is a voltage- indicating device which draws substantially
no
power and hence can be used across high- impedance circuits
with little
or no loading effect. This feature of the 6E5 makes it valuable
as an
indicating device. The actual variations of shadow angle, plate
current,
and target current with respect to grid voltage are shown in
the aver-
age control characteristics of Figure 5.
APPLICATIONS OF VISUAL -INDICATOR TYPE TUBES 113
6E5 BALANCE INDICATOR FOR A VACUUM -TUBE VOLTMETER

One of the most useful applications of the 6E5 is that of balance
indicator in a slide- back -type vacuum -tube voltmeter. An instrument
of this type having a voltage range of 200 volts is diagrammed in Fig-
ure 6.* A description of its manner of operation follows.
Referring to Figure 6, assume that the input terminals (test prods)
A and B are shorted and that the movable arm of slide -back potentio-
meter R7 is at the positive -voltage end of its range, at which setting
d -c voltmeter "V" will read zero. The plate current of the triode-con-
nected input -coupling tube (Type 6J7, 6C6, 57 or 954) is practically
cut off, due to the biasing action of the 2.0- megohm cathode resistor
(R1).
Because it requires about -14 volts to obtain approximate cutoff
of the 6J7 plate current, the cathode end of R1 is at a potential of +14
volts with respect to the lower, or negative, end of R6. The "zero-set-
= I MEG.
TYPE
6E5
4.5 MA.
I =
APPROX.
r I° III
250 V.
9V.
Fig. 3-Circuit illustrating the operation of the 6E5.
ting" potentiometer (R5) is next adjusted so that the potential of the

6E5 cathode is about +21 volts with respect to the negative end of R6.
This 21 volts, in series opposition to the 14-volt drop across R1, places
a net voltage of -7 volts on the grid of the 6E5, with respect to its
own cathode. Thus, the pattern on the target will be closed to a narrow,
dark line, which is the correct initial setting for all vacuum -tube volt-
meter measurements.
If the test prods A and B are now applied across any d -c or peak
a -c voltage of suitable value (0.5 to 200 volts) which it is desired to
measure, the plate current flowing through R1 will increase by an
amount substantially proportional to the d -c or peak a -c voltage applied.
The action is similar to that of a diode rectifier, except that practically
no power is drawn by the negative -grid input circuit. In the case of
a -c voltages, rectification occurs on the positive half -cycles and the
* This instrument was described in Radio Retailing, Dec., 1935, and in
QST, Oct., 1936.
114 RCA REVIEW
large condenser C1 (shunted across R1) holds the d -c voltage developed

across R1 at practically the peak value of the a -c wave.
Because of the 2.0-megohm circuit across which it is shunted, con-
denser C1 must be of a high -quality, low- leakage type; a good paper
condenser is satisfactory. The capacity value of C1 depends on the
lowest- frequency a-c voltage that might be measured. A value of 4
microfarads is suitable for frequencies of 60 cycles per second or
higher. In the case of d -c voltages, terminal A of the vacuum -tube
voltmeter is always connected to the positive side of the input voltage.
To complete the explanation of the vacuum -tube voltmeter oper-
ation, assume that a d -c voltage of 10 volts is applied to the input
terminals. The IR drop across R1 will now be 14 + 10, or 24 volts. This
-
places a voltage of 24 21, or +3 volts on the grid of the 6E5, causing
the shadow sector to open. The slide -back control R7 will now have to
ZERO GRID BIAS NEGATIVE GRID BIAS

Fig. 4-The pattern on the 6E5 target appears as shown above with zero
grid bias and with grid bias near cutoff.
be moved toward its negative end until voltmeter "V" indicates 10 volts
before the original -7 volts of bias is again restored to the grid of
the 6E5. Thus, when the voltage introduced into the circuit by R7
(read on voltmeter "V ") is adjusted just to cancel the unknown input
voltage, the pattern of the 6E5 will again close to the initial, or narrow-
line setting. The 6E5 is, therefore, functioning as a voltage -balance
indicator.
The use of a protective resistor (R2) in the grid circuit of the 6E5
is important, because any input voltage exceeding 7 volts will drive the
grid of the 6E5 positive. The IR drop across R2, developed by the 6E5
grid current, automatically biases the tube so that the grid current can
not reach a value high enough to be harmful.
ACCURACY OF THE VACUUM -TUBE VOLTMETER

In general, the accuracy of this type of instrument will depend upon
the care with which the fluorescent pattern is adjusted before and after
the unknown voltage is applied, as well as upon the accuracy of the d-c
voltmeter "V ". D -c voltages between about 25 and 200 volts can be
read to one volt or better, depending on the readability of the voltage
scale on "V ". Between 0.5 and 10 d -c volts, the accuracy is plus or
minus 0.1 to 0.2 volt. A -c voltages will give readings which are in error
by a fairly constant value of 0.8 to 1.3 volts, on the low side of the
correct value. This error is apparently due to the reaction of the nega-
tive half -cycle on the static value of the 6J7 plate current. For example,
a peak a -c voltage of 1.4 and 2.8 volts gave a vacuum -tube voltmeter
reading of 0.6 and 1.75 volts, respectively. The percentage error is
naturally smaller for larger values of a -c voltage, so that the higher
a -c readings are quite accurate.
250 100 TYPE
Ef e
6E
6.3 VOLTS
-_
---
CURVE PLATE-SUPPLY SERIES PLATE
(B+)LY SISTOR (R)
200 - 80
--- 250
100
1.0
0.5 / ?A
.u ÿ `JPm
'úI
TARGET CURRENT ,,__-

Q
LARGET
CURRENT
II /
Z
< 40 ._MMIWan
r
o
0 Pil ó I PJ2
?
IOWA
c
5
Y1
20 All y,Z.r Qv%

'
r/ O
3 TO CONTROL-
LING VOLTAGE
B+ -
O
-10
GRID VOLTS 92C-4422R2
Fig. 5- Operating characteristics of the 6E5.
The error on low a -c voltages is not disturbing, because the instru-

ment can readily be calibrated for these voltages by means of a vari-
able a -c source of known voltage. The calibration can be made in terms
of either rms or peak values. It is important to remember, however,
that the voltage indicated by "V" is invariably in terms of either d-c
or peak a -c. If the a -c input voltage has reasonably good wave form,
the peak a -c readings can easily be changed to rms values by multiply-
ing them by the factor 0.707.
OTHER VACUUM -TUBE VOLTMETER CONSIDERATIONS

A voltage calibration of slide -back potentiometer R7 can be made
if it is desired to eliminate d -c voltmeter "V" from the circuit. This
arrangement will not, of course, provide as good accuracy as with a
116 RCA REVIEW
d -c voltmeter having several voltage scales. If many measurements
are to be made in the low- voltage range (1 to 10 volts), a 500 -ohm
potentiometer should be placed in series with R7, as shown in Figure 7.
A single -pole double -throw switch is required to change from the 10
to the 200 -volt range. The switch must be thrown to the low-voltage
scale and the 500 -ohm potentiometer adjusted to zero in order to make
the initial zero setting with R5, even when a voltage higher than 10
volts is to be measured ; the switch is thrown back to the 200 -volt scale
after the zero setting is made with R5.
The mounting of the input tube is quite important. Where r-f
voltages are to be measured, the capacitance of the input circuit must
be kept as low as possible. The well -known "goose- neck" probe con-
struction is recommended for this reason. If very many r-f measure-
ments are contemplated, a 954 acorn tube, connected as a triode, should
be used in place of the 6J7. The loading introduced by the 954 is rela-
tively small.
APPLICATIONS OF THE 6E5 VACUUM -TUBE VOLTMETER

Because the vacuum -tube voltmeter described has a considerable
voltage range, draws no current, and can be used across high- impedance
and high- frequency circuits, it has a multitude of uses. A few of these
may be worthy of mention, although this is to digress somewhat from
the main subject.
The avc circuit of a refractory superheterodyne receiver can easily
be checked. Prod A is connected to the cathode of one of the controlled
r -f or i -f stages and prod B to the ground side of the r -f or i -f trans-
former secondary. The ave bias variation can be measured accurately,
either on a signal or with a test oscillator. The 6E5 will at the same
time serve as a resonance indicator or as an alignment meter, because
the avc voltage varies as the different circuits are adjusted. This
method of i -f alignment is not as good, of course, as the double-image
method employing a cathode-ray oscillograph and its auxiliary equip-
ment.
Tube screen and plate voltages can be checked with precision, even
though a very large series resistance is included in the circuit. The
true plate voltage at the plate of a resistance-coupled a -f amplifier can
quickly be determined, even if the plate load has a value of 0.5 megohm,
or more. If the a -f tube is biased correctly, the voltage at the plate
will usually be about one -half of the plate -supply voltage.
The operation of r-f or i -f stages can be checked roughly by meas-
urement of the r-f voltage across the transformer secondary. The test
lead from prod A should be short and should have very little capacitance
APPLICATIONS OF VISUAL-INDICATOR TYPE TUBES 117
to ground. The 6J7 may, even with very short input leads, place a
capacity load of 5 to 10 micromicrofarads across the transformer; this
will detune the circuit under test more or less, depending on its nature.
The use of the 954 in the input stage of the vacuum -tube voltmeter is
preferable for tests of this kind.
Where an a -f output meter is needed, the test prods can be applied
to almost any part of the a -f circuit. If it is necessary to separate an
a -f voltage from a d -c voltage, in such applications, a 0.1- microfarad
blocking condenser and a 1 to 5- megohm grid leak can be employed at
the input to the vacuum -tube voltmeter. In addition to serving as an
a -f output meter, the vacuum-tube voltmeter can be used to measure
the peak a -c driving voltage applied to the grid of the a -f tube.
TYPE 6J7 +430V. 115 V.
60",
A- OnOn
+
375 1 375
6.3V.' ' 6.3V.
TYPE F-I F-2
5Z4
R5
R1 =2 MEGOHMS, 0.5 WATT

R2= 100000 OHMS, WATT
1
R31 MEGOHM, 0.5 WATT +- C7

R4 =10000 OHMS, 10 WATTS
R5 =1000 -OHM, WIRE -WOUND
LINEAR POTENTIOMETER T
R6 =500 OHMS, 2 WATTS
R7= 10000-OHM, 10 -WATT WIRE -WOUND
LINEAR POTENTIOMETER CI= 4 p f, 400 V. PAPER C4= 0.1 }1 f, 300 V.
(LOW- LEAKAGE) C6= 6yf, 350 V.
V= VOLTMETER, 1000 OHMS PER VOLT,
200 -VOLT RANGE C2,C3,C5= 0.01)1f MICA C7=16j.1f, 500V.
Fig. 6 -This v -t voltmeter circuit shows one of the most useful unconven-
tional applications of the 6E5. The instrument is of the "slide- back" type.
The gain of the audio stages can also be measured. A known peak
a -f voltage (such as a 60 -cycle source shunted by a known voltage
divider) is applied to the grid of the a -f tube under test. The peak a -c
voltage across the plate load is next measured; this value, divided by
the known peak input voltage, gives the voltage gain of the stage, at
the particular test frequency employed.
The determination of the ratio of a transformer is simple. A suit-
able a-c voltage is applied to any winding and that voltage measured.
The voltage of any other winding is also measured; the ratio of the
two peak voltages is substantially that of the two windings.
The power output of an audio power amplifier can be determined
with the aid of the vacuum -tube voltmeter and a. bit of arithmetic. A
118 RCA REVIEW
test signal voltage of (say) 1000 cycles from an audio oscillator having
a reasonably good wave form is applied to the audio system at any
convenient stage. A pure resistance load of the correct resistance and
wattage is shunted across the primary of the audio output transformer.
A value of 7000 ohms, for example, will be used in the case of a single
47 pentode. The secondary load of the output transformer is discon-
nected. The test signal is then increased until its peak value at the
grid of the power tube (or at one grid in a push -pull stage) is the
maximum permissible for the stage under test. For a power tube oper-
ating Class A or Class AB1i the peak signal should not cause d -c grid
current to flow. The peak a -c voltage across the plate load resistor is
then measured. Taking the 47 as an example, we find that the measured
peak output voltage (Epic) is 186 volts. Changing this to an rms value,
for power calculations, we get, Erm8 = (0.707) (186) = 132 volts. From
the relation P = Er7.2 /R, we find that P= (132) 2/7000 = 2.5 watts,
the power output.
The peak plate current of a mercury-vapor rectifier can be measured
as a check on the correct operation of the rectifier tube. A 100 -ohm (or
other suitable value) resistor is placed in the -B lead of the rectifier
between the transformer and the filter system. The vacuum -tube volt-
meter will measure the peak d -c voltage developed across the resistor,
the rectifier being operated under normal load. Ohm's law gives the
peak d -c plate current in the circuit, Ipk = Epk /R.
The ripple voltage of a high -voltage rectifier can be checked readily,
if there is sufficient ripple voltage to measure (0.5 volt or more). A
d -c blocking condenser -and -leak input circuit must be used
for the
vacuum -tube voltmeter, of course, with a condenser of the necessary
high -voltage rating. The peak ripple voltage is measured across the
filtered output of the supply, in the usual manner for measuring small
a -c voltages.
The vacuum-tube voltmeter has a number of useful applications in
the adjustment of radio transmitter stages. It may be used as an ultra-
sensitive neutralizing indicator, merely by employing it as an r -f
output meter across all, or a portion, of the plate -tank circuit of the
stage being neutralized. The voltage measured will be at a minimum
when the point of best neutralization is found. Test prod B should
be
placed at the r-f voltage node on the tank coil, because this input
terminal has a lower impedance to ground, at high frequencies, than
terminal A.
As a modulation meter, the vacuum -tube voltmeter will measure the
percentage of modulation with good accuracy, provided the modulation
is symmetrical and the carrier is not subject to appreciable shift.
A
APPLICATIONS OF VISUAL-INDICATOR TYPE TUBES 119
small amount of r-f voltage, picked up at the modulated r -f stage and

transferred to the vacuum -tube voltmeter through a low- impedance
line (twisted pair), is measured. Assume that this voltage, with no
modulation on the carrier, is 10 volts. At 50 per cent modulation, the
vacuum -tube voltmeter will read 15 volts; at 100 per cent modulation,
it will read 20 volts, or twice the non -modulated value. Only positive
modulation peaks can be indicated by this method.
As an over -modulation indicator, the vacuum -tube voltmeter can be
operated in a somewhat different manner. Slide -back potentiometer
R7 is adjusted just to cancel the non -modulated carrier voltage, picked
up as described above, and the d -c voltmeter reading noted. R7 is then
adjusted so that "V" reads about 95 per cent higher than the first read-
500 1+.
WIRE -WOUND
POTENTIOMETER
5I
- C6
C5
Fig. 7- Partial circuit of vacuum tube voltmeter showing the use of a vernier
control for the measurement of low voltages.
ing. This over -biases the 6E5 and over -closes the pattern. Then, when
the pattern begins to flip open slightly as the carrier is modulated, the
positive modulation peaks are beginning to exceed the 95 per cent
modulation point. A disadvantage of both of these arrangements is
that the indicator does not show the effects of carrier shift, or of over -
modulation on the negative modulation peaks. It is the negative mod-
ulation peaks which create the most disturbance when they reach the
carrier cut -off point, because of the resultant flattening of the modula-
tion envelope on the one side.
THE 6E5 AS A NEGATIVE-PEAK OVER -MODULATION INDICATOR
Another application of the 6E5, not involving the vacuum-tube volt-
meter, is shown in Figure 8. Here the electron-ray tube is used in con-
junction with a high -voltage, half -wave, vacuum -tube rectifier (the 879
is shown, as an example) . This arrangement operates as a negative -
peak over-modulation indicator in the following manner: When the
120 RCA REVIEW
a -c modulating voltage at point "x" swings positive, the 879 does not
pass current because its filament is at a positive potential with respect
to its plate. When the modulating voltage swings negative at point
"x ", the rectifier still fails to pass current until the negative a -f peak
exceeds the d -c plate voltage applied to the Class C modulated r -f am-
plifier. When this does occur, causing carrier cut off, the instantaneous
voltage at point "x" is negative with respect to the plate of the 879
and the latter then passes a rectified d -c current through the load
resistor (R1). This current, causing a d -c voltage drop across R1,
biases the grid of the 6E5 negatively. The pattern on the 6E5, there-
fore, flips shut whenever the negative a -f peaks are great enough to
cause carrier cut off ; this happens irrespective of possible carrier shift.
MODULATED
CLASS C AMP. I MEG.
TYPE
879
50000011
(-)
I
(-)
R
TWISTED
X I PAIR TYPE
B
6E5
HIGH -VOLTAGE
INSULATION +B +200V.
TO 250 V.
A -C
MODULATOR LINE
tB
Fig. 8 -A negative -peak over-modulation indicator using the 6E5.The value
of "C" controls the time lag, and may be 0.05 to 0.5 microfarad. Larger
values provide slower action.
It is apparent that this device is exceedingly sensitive to the slightest
negative -peak over -modulation, inasmuch as only 7 volts are required
to close the "eye" completely.
The sensitivity can be controlled by potentiometer R1 (Figure
8),
which regulates the amount of excess modulating voltage applied
as
bias to the 6E5. The size of condenser C controls the speed with which
the "eye" reopens after an excessive modulation peak has passed. That
is, although the pattern closes rapidly, it can be made to reopen
slowly,
to assist the operator in making the observation. This type of over -
modulation indicator is much to be preferred to one indicating
only
positive peaks, for reasons already discussed.
THE 6E5 AS AN INDICATOR IN NON -AVC RECEIVERS
It is, of course, well known that the 6E5 is ordinarily used in
re-
ceivers employing automatic volume control with a diode detector.
The electron-ray tube can also be used in a receiver having neither

avc nor a diode detector. The alternative arrangement is suitable for
t -r -f or superheterodyne receivers employing a cathode -resistor- biased
detector, as illustrated in Figure 9. Potentiometer R2 is set at the
detector cathode potential. This places a positive voltage, equal to the
no- signal bias of the detector, on the grid of the 6E5, opening the
"eye ". Cathode resistor (R1) in the 6E5 circuit is next adjusted just
to close the "eye" to a narrow line. When an r -f signal is tuned in,
the plate current of the detector rises, the voltage drop across R2 and
R3 increases, and the 6E5 pattern opens slightly. At the point of
best
tuning or resonance, the pattern will have opened to a maximum; de-
tuning the receiver causes it to close again. The 6E5 thus acts as a
visual tuning indicator, but operates exactly backward with respect to
CATHODE- BIASED
DETECTOR
+250V.
TYPE
6E5
250000 IL
6.3V. R
0.5 0.5 fif

MEG.
2500011 50000 n
0.1
F-
}if
Fig. 9-Circuit using the 6E5 as a visual tuning indicator in receivers not
having AVC or a diode detector. The "eye" opens to a maximum when
a signal is properly tuned in.
its usual pattern movement. The backward operation of the "eye ",
however, is not especially objectionable
better the tuning.
-
the larger the shadow, the
If a very strong signal develops so much voltage across R3 that the

pattern opens fully and loses its indicating value thereby, the arm of
R2 should be moved toward ground. R1 must then also be readjusted
for a closed pattern, under no- signal conditions.
NULL INDICATOR USING THE 6E5
A more unusual application of the 6E5 is that of a null or voltage

minima indicator for bridge circuits. Figure 10 illustrates such a
device, including a 6J7 signal amplifier. The input terminals of the
null indicator should be connected across the balance points in the
122 RCA REVIEW
bridge circuit. Either a-c or d -c voltage may be used across the bridge
circuit, as far as the null indicator is concerned.
The 6J7 is operated as a d -c amplifier, with an adjustable grid -bias
supply provided by the 400 -ohm potentiometer in the bleeder circuit.
The grid -bias and screen -voltage controls are adjusted, under no-signal
conditions, so that the plate current is almost cut off. The screen volt-
age is important as regards the sensitivity of the 6J7 on small a -c
signals. A small amount of no- signal current flowing through the plate
resistor of the 6J7 will produce about 3 or 4 volts of negative bias on
the grid of the 6E5, through the 0.5-megohm filter resistor. This bias
voltage will cause the 6E5 pattern to close about half way, or 50 degrees.
If a small d -c or a -c voltage appears across the input terminals,
as is the case when the bridge circuit is unbalanced, the plate current
of the 6J7 increases, thus placing more bias on the 6E5 grid and closing
1.0 MEG.
ABC
OR
D -C
INPUT
VOLTAGE
400.n. 5000 n 12000 -n-

D C SUPPLY VOLTAGE
Fig. 10 -This arrangement can be used in conjunction with bridge circuits
where a sensitive indicator is required.
the pattern to a narrow line. As the balance control on the bridge is
adjusted to a point on either side of the correct nodal point, the "eye"
will flip entirely shut. At the nodal point, it will open to the same
maximum angle at which it rests under no- signal conditions. The nodal
point can be accurately determined by noting the position of the balance
control on the bridge at the two points, one on each side of the nodal
point, at which the "eye" just closes to a narrow line. The nodal point
can then be found on the balance control half way between the two
points mentioned. The sensitivity of the indicator is estimated to be
about 10 millivolts.
The 10000 -ohm protective resistor in the 6J7 grid circuit is essen-
tial. In case a relatively large voltage appears across the input ter-
minals, this resistor will prevent excessive d -c grid current, due to the
extra d -c bias automatically applied to the grid of the 6J7.
SUMMARY - 6E5 APPLICATIONS

A number of applications for the 6E5 have been described, other
than its normal use as a visual tuning indicator. Some of these involve
quantitative measurements, others only the indication of maxima or
minima. Obviously, such applications might be continued indefinitely.
It is surprisingly easy to fit the 6E5 into various types of circuits, just
as in the examples shown.
THE 913 -A ONE -INCH CATHODE -RAY TUBE
A new type of visual- indicator tube, designed on the cathode -ray
principle, has recently been made available. This tube, the RCA 913,
has two sets of electrostatic deflecting plates and a fluorescent viewing
screen about one inch in diameter. An outstanding design feature is
the fact that, although it is a high-vacuum type, the 913 can be oper-
ated with anode voltages as low as 250 volts. The metal shell of the
913 serves not only to maintain the other electrodes in a high vacuum,
but also to act as the second anode, in which role it assists in focusing
the electron beam.
The design of the 913 involves much more than meets the eye upon
casual inspection. For example, a similar low- voltage tube in a plain
glass bulb would not function properly, due to electrostatic charges
building up on the glass wall. These charges would interfere with the
focusing of the electron beam. The metal-shell design substantially
eliminates this problem, and also provides an excellent electrostatic
shield.
Figure 11 illustrates a typical cathode -ray oscillograph circuit using
the 913. Features of this circuit are the linear time -sweep oscillator,
employing an 885 gas triode, and the two a -f amplifiers, using 57's con-
nected as pentodes. One 57 amplifies the output of the saw-tooth oscil-
lator and provides, through switch S3, the horizontal time -sweep volt-
age for deflecting plates D3 and D4. By means of switch S2, the same
57 may be used to amplify an external sweep voltage. An unamplified
external voltage can also be applied to the horizontal deflecting plates
by means of switch S3. The other 57, referred to as the "signal ampli-
fier" or "vertical- deflecting -voltage" amplifier, applies through switch
S4 an amplified signal for deflecting plates Dl and D2. This voltage
produces a vertical deflection on the fluorescent viewing screen. The
switches S3 and S4 should be of a low- capacity type, because r-f voltages
may be applied to either set of deflecting plates, in some applications.
The 57 was chosen for this circuit in preference to the 6J7 because the
particular power transformer available did not have an extra 6.3 -volt
4. ÿo r
yU
U >
GIS P:
o
In
cl
a3
4'0
ÿ 3 3
ai
,. In
ÿ ir
Y
aò
cC
S'34'°' m^vî vï y
ó
.3 p.
11'J
ÿ
°
c'3
>r!
ÿ
3 3
ó000 `n 0
Q`ní0nóí0nóóo+-n
>
oo"' , n°
>0 "o`n
óg p0 4.
óónI~°0
ó °tw ao
Eó
Üva-<+-ï`nníóooó,- ó
v-^ Î~
4-I"'
o ` ó óUNDó °0°ó
ó ó
C^..cóo E
000,-innw `nÓ ó2Zó

o
°o
.,,,-7
w
°Ó6 oo óóó°ó2óó2°0
2ci000knooó
óóó
II II
© ©
ió
r
II
dóddUC>cicì
II vU
II UcU
11
ó°onc°ó°.
II II 4444 4
II II4 4
ti
11 11
I I I I II
°
II II 11
II 11 11
II II
II II 11
II .ó II
11 11
11 11 II
00 0 00 0 0 0 0 C>c>3cSLscSa4 4 4r 4 44444444 49
II
I-
1111
o
á ^
I
Úi
vv\---IU
_r AA/V-0 a)
(f) TS
N
"- I
I
Ó
W
ár -U U") II O)
.''
~ ÚQ
N N
Y
N
N
U
O
o o
u)
,90,.4994940,
o
heater winding for the amplifier tubes. The amplifiers should not be
operated from the same heater winding as the 913.
The double power supply used in this oscillograph is of special
interest. The 80, in a full -wave rectifier circuit, provides about 450
volts above ground for the operation of the amplifiers and the saw -
tooth oscillator. The l -v, in a half-wave circuit, provides about 450
volts below ground for the electrode supply of the 913. Because the
positive terminal of this supply is at ground potential, the shell and
second anode of the 913 can be directly grounded. This obviates the
necessity of insulating the shell of the cathode-ray tube from ground,
and eliminates the need for condensers in series with the free deflecting
plate leads.
Provision for synchronizing the saw-tooth oscillator is made by
means of the double-throw switch, S1. In one position, synchronization
with the frequency of the a -c line voltage is provided through the con-
nection to the heater of the 885; in the other position, the oscillator
can be synchronized with an external voltage source.
The linearity of the time -sweep voltage is quite good at frequencies
from 30 to 7500 c.p.s. Potentiometer R5 provides the vernier control
for the sweep oscillator. The amplifiers have a flat response curve from
about 30 to 20000 c.p.s., their voltage outputs dropping to about 50 per
cent at 70000 c.p.s. These characteristics of the amplifiers and of the
sweep oscillator are satisfactory for most audio- frequency applications.
Although the viewing screen of the 913 is relatively small, an oscil-
lograph using this tube is suitable for many applications where larger
cathode -ray tubes have previously, of necessity, been employed. The
913 lends itself to a compact, light- weight, economical oscillograph
design. The oscillograph shown in the circuit of Figure 11 is rather
elaborate, inasmuch as it contains most of the features included in
commercial equipment using larger cathode -ray tubes. The individual
builder can, in many cases, greatly reduce the number of component
parts, to meet his particular requirements. In some cases, only one
amplifier may be necessary. In others, a linear sweep circuit is not
required. Stripped to its bare essentials, an oscillograph employing
the 913 might consist only of the cathode-ray tube, a heater trans-
former, and the bleeder circuit, with provision for connection to an
external d -c voltage supply.
In conclusion, the writer wishes to acknowledge the assistance given
by Messrs. P. A. Richards, J. F. Dreyer, and F. H. Shepard, Jr., of the
Research and Development Laboratory, RCA Radiotron Division, RCA
Manufacturing Co., Inc., Harrison, N. J., in the design of many of the
circuits shown in this discussion.
OUR CONTRIBUTORS
Max C. BATSEL, upon receiving his degree of B.M.E.
in 1915 from the University of Kentucky, joined the
Western Electric Company. In 1916 he went with the
Bureau of Standards. In 1918 he was commissioned a
lieutenant in the Signal Corps. In 1920 Mr. Batsel be-
came engineer in charge of radio receiver development
and design for the Westinghouse Company and in 1929
joined RCA Photophone, Inc., as Chief Engineer. When
the Photophone Company was taken over by RCA Victor
in 1932, Mr. Batsel became section engineer, Photophone
Development and Design.
RALPH R. BEAL is an alumnus of Leland

versity and has been continuously identifiedStanford Uni-
in radio engi-
neering since 1912. His career began with the early
development of the arc converter as a generator of con-
tinuous waves for long distance radio communication.
During the war he was resident engineer in Washington,
D. C., for the Federal Telegraph Company. He later
served for five years as Chief Engineer and was in charge
of the installation of high power radio transmitting
equipments in stations of the United States Navy. In
1926 he joined RCA as Pacific Division Engineer and
four years later became Manager, Pacific Division, R.C.A.
Communications, Inc. In 1934 he was appointed to his present position of
Research Supervisor of RCA.
HAROLD H. BEVERAGE -See page 3.
CHARLES M. BURRILL studied electrical engineering at

the University of Minnesota, graduating in 1923. He
then went with the General Electric Company. Follow-
ing three years of general training in the G. E. Advanced
Course in Engineering, he joined the Radio Engineering
Department, and in 1927 was placed in charge of Tuned
Radio Frequency Receiver Development. Since 1930 he
has been with RCA at Camden, N. J., with the exception
of a year and a half in 1931-32 spent with the Rogers -
Majestic Corporation of Toronto, Canada, in charge of
research. Since returning to Camden he has been engaged
in general research, first in sound recording, and more
recently in interference and noise suppression and in ultra short
propagation. wave
IRVING F. BYRNES entered the General Electric Test

Department in 1918 and later engaged in radio develop-
ment in their Engineering Laboratory. From 1920 on he
was occupied in the development of radio equipment for
commercial and military vessels, submarines and air-
craft. He participated in the design and tests of the
early ship -to -shore duplex radio telephone equipment
used on the SS. America in 1922. Mr. Byrnes joined the
Engineering Department of RCA Manufacturing in 1930,
later in that year transferring to the Radiomarine Cor-
poration of America of which he has since been in charge
of engineering activities.
126
OUR CONTRIBUTORS 127
E. W. ENGSTROM is the man responsible for the de-

velopment and design of apparatus used in the present
RCA television field test. He is Director of General Re-
search for the Victor Division of the RCA Manufacturing
Company, Inc. Since joining that company in 1930 he
has been associated with engineering on Photophone
apparatus, broadcast receivers and research. Prior to
that he devoted seven years to radio transmitters and re-
ceivers while in the engineering organization of the Gen-
eral Electric Company. Mr. Engstrom graduated from
the University of Minnesota in 1923.
D. W. EPSTEIN received his B.S. degree in engineering

physics from Lehigh University in 1930, and an M.S.
degree in electrical engineering from the University of
Pennsylvania in 1934. He joined RCA in 1930. For five
years he was in the Research Division, Engineering De-
partment, of the RCA Victor Company and since then
with the RCA Victor Division, RCA Manufacturing Com-
pany. Mr. Epstein is a member of the American Physical
Society and an associate member of the Institute of Radio
Engineers.
H. B. MARTIN attended the University of Illinois from

1924 to 1928 receiving a B.S. degree in Electrical Engi-
neering. For a short period in 1928 he was employed by
the Westinghouse Electric and Manufacturing Company
at East Pittsburgh, Pa., as a student engineer. Later in
1928 he went with the Signal Corps Procurement Office,
New York, as Purchase Engineer and Electrical Inspec-
tor of U. S. Army radio and electrical equipment. In 1929
he was transferred to the Radio Research Section of the
Signal Corps laboratories at Fort Monmouth, N. J.,
where he engaged in the design and field testing of
portable radio equipment. In 1930 Mr. Martin entered
the Engineering Department of Radiomarine Corporation of America as
assistant in the development and flight testing of aircraft radio apparatus
at the Roosevelt Field Laboratory of that company, and later was engaged in
the design of marine radio equipment at the Radiomarine Laboratory in
New York. In 1932 he joined the instruction staff of R.C.A. Institutes
where he conducted courses in transmitter engineering. Three years later
he became Assistant Chief Engineer of Radiomarine Corporation, where he
is at present engaged in development and application engineering in various
phases of marine radio, including facsimile weather map, harbor radio
telephone, auto alarm and direction finding equipment, as well as trans-
mitters and receivers for regular telegraphic communication. Mr. Martin
is an Associate, Institute of Radio Engineers, a member of Theta Tau, and
a member of Phi Pi Phi.
GEORGE McEI.RATH, Operating Engineer for the Na-

tional Broadcasting Company has been associated with
radio since 1918. He attended the U. S. Naval radio
school and after the war was employed in the plant de-
partment of the American Telephone and Telegraph
Company. In 1922 he was transferred to the radio broad-
casting department of that Company and assisted in the
installation of station WEAF. He worked in the studio
and field groups and in 1924 was appointed Engineer in
Charge of NBC operations at their Washington office.
He was assigned to his present position in 1928. Mr.
McElrath is a member of the Institute of Radio Engineers.
128 RCA REVIEW
GEORGE O. MILNE, Eastern Division Engineer for the
National Broadcasting Company, attended the Western
Electric Company Installation and Machine Switching
School in 1922 and assisted in the installation of the first
full mechanical telephone exchange in the East at Pater-
son, N. J. In 1924 he was transferred to the radio broad-
casting department of the American Telephone and Tele-
graph Company and worked in the studio, field and trans-
mitter groups. Two years after the formation of NBC
in 1926 he was appointed Operation Supervisor. He was
placed in charge of NBC's Eastern Engineering Division
when it was organized in 1930.
WALDEMAR J. POCH was born in London, England. He

received his B.S. degree from the University of Michigan
in 1928, and his M.S. degree from the University of Penn-
sylvania in 1933. He was with the General Electric Com-
pany at Schenectady in 1928 and 1929 and joined the
RCA Manufacturing Company in 1930. Since 1931 he
has been engaged in television development.
C. N. REIFSTECK completed his course in Electrical

Engineering at the Iowa State College in 1925, following
which he joined the Radio Engineering Department of
the Westinghouse Electric and Manufacturing Company
where he was engaged on radio phonograph combination
equipments. In 1930 he became a member of the Photo -
phone Engineering Department of the RCA Manufac-
turing Company in Camden in charge of Product Design,
for which work he is still responsible. Mr. Reifsteck is
a Fellow in the Society of Motion Picture Engineers.
DAVID SARNOFF, President of the Radio Corporation of

America, has been continuously identified with radio
since 1906. He received his early education in New York
public schools and later was graduated from Pratt Insti-
tute, where he took the electrical engineering course. He
is a Fellow, Institute of Radio Engineers and served as
Secretary and Director of I.R.E. for three years. Mr.
Sarnoff is a Member, Council of New York University;
Member, Academy of Political Science and Member,
American Institute of Electrical Engineers. He holds
the honorary degrees of Doctor of Science from St. Law-
rence University, Doctor of Science from
lege, and Doctor of Literature from Norwich University. He isMarietta Col-
an honorary
Member of Beta Gamma Sigma and an honorary of Tau Delta Phi. He is a
Colonel SC -Res., U. S. Army.
LELAND C. WALLER received his degree of

B.S. in elec-
trical engineering from the Iowa State College in 1929,
and upon graduation joined the Operating Division of
R.C.A. Communications, Inc., at Rocky Point. Early in
1930 he was transferred to the Foreign Sales Division of
the RCA Victor Company in New York City. Since the
latter part of 1930 he has been a member of the Com-
mercial Engineering Section, Research and Development
Laboratory, RCA Radiotron Division of the RCA Manu-
facturing Company.

RCA Review 1937 Jan

Uploaded by

Copyright:

Available Formats

RCA Review 1937 Jan

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RCA Review 1937 Jan

Uploaded by

Copyright:

Available Formats

RCA REVIEW

A Quarterly Journal of Radio Progress

VOLUME I January, 1937 NUM BER 3

Partial Suppression of One Side Band in Television Reception 19

Equipment Used in the Current RCA Television Field Tests 36

Reproducing Equipment for Motion Picture Theatres 65

Some Notes on Ultra High Frequency Propagation 76

Frequency Assignments for Television 88

Behind the Scenes of Two Notable Broadcasts 94

Applications of Visual- Indicator Type Tubes 111

Our Contributors 126

SUBSCRIPTION: $1.50 per Year. Foreign, $1.85. Single Copies 50 cents.

Copyright, 1937, by RCA Institutes, Inc.

Previously unpublished papers appearing in this book may be reprinted,

At the banquet of the Society of Motion Picture Engineers, held at

HISTORY in the making is always out of focus. But time and

THE GROWTH OF RADIO

THE MARINE DECADE

vessel carrying more than fifty persons -

THE WAR DECADE

make radio an instrument of democracy. It took it out of the hands

radio telephone, however, presented one serious difficulty -

THE BROADCASTING DECADE

dreamed of such a thing - -

technicians, but by the everyday business requirements of business

This is a significant statement, for it cannot be made concerning

BROADCASTING IN TUNE WITH ITS TIMES

point message service to a wide popular usage

made in Washington to nationalize the industry

radio facilities in this country; second, to perpetuate the benefits to

This brings me to the second part of this discussion -the future of

When it comes to transmitting and receiving the instantaneous,

series of preliminary field tests to bring television from the laboratory

The National Broadcasting Company is making an intensive study

glamorous of industry's children - -

RADIO CONTRIBUTES TO OTHER FIELDS

If science has taught us anything it is to be humble; to realize that

EARLY television development followed the precedents estab-

APPARATUS USED IN EXPERIMENTAL WORK

MODULATION CSCILLATOR OSCILLATOR LE AMPLIFIER 2ND

Fig. 1 -Block diagram of the transmitter and receiver equipment.

had a sharp selectivity characteristic compared with that of the picture

Figs. 2 and 3 show the selectivity characteristics of the transmitter

Fig. 2-IF selectivity of side band suppressor system.

MEASUREMENT OF FIDELITY AND PHASE CHARACTERISTICS

Measurements of fidelity and phase characteristics were made in

Fig. 3- Receiver IF selectivity.

Fig. 4-Measured fidelity. 1. Over-all fidelity double side-band operation.

ticipated. Since at low modulation frequencies both side bands are

downward in the response curve. It is possible, of course, to carry

the ellipse method, was converted to phase delay by the following

Fidelity and delay characteristics for conditions similar to those in

7- Selectivity and phase characteristics of the intermediate -

intermediate frequency, it is worth reviewing, for the sake of clarity,

e = E cos wot -}-

where w = 24, E is the amplitude of the unmodulated carrier, and m

where A, Au, and AL are the amplitude ratios of output to input of

u = Ve cos (wot -}- (1,)

-E- mAe Au cos (wit