MATI ....

,
".1~.6. N

ni~earCh
},ngineering
/, ~

page 6 THE TECHNICAL INFORMATION PROBLEM

'I,I
I
,I
THIS

TELEMETER

MAGNETICS

SUPPORTS

THE WEIGHT

OF ONE

YOUNG

ELEPHANT

has a jaster, more reliable memory, too!

SILLY TEST
but we hope it attracts your
attention to the thorough
three-stage inspection and testing
given every TELEMETER MAGNETICS
memory product - from core to
array to buffers and memory systems.

economical· reliable· fast· Telemeter Magnetics Modular Memory Systems are designed and manufactured for absolute dependability

Because iELEMEiER M,\GNETICS manufactures the ll1emory system complete from core pro-
duction through array wiring to finished units ... and because TELEMETER MAGNETICS tests
each phase of production thoroughly .. ,you can be that much surer when you specify memory
systems designed and engineered by TM!.
Modular design permits produr.tion of custom memory systems to satisfy practically any
desired configuration. Units of from 100 to 1,000,000 bits are common ... memories can be
supplied incorporating several million bits. In addition, TMI offers you a selection of memory
units with cycle times of 24 microseconds, 6 to 8 microseconds, and 3 microseconds.
Eler.tronic circuits ill TELEMETER MAGNETICS memory systems employ solid state elements
20-1Il/('r()S('I'OIU/ srstelll
(J throughout .. ,transistors, diodes, and ferrite cores. Amplifiers, registers, drivers, and logic
20J8 1{,(III/S oj 22 liit8 ('(1ch are 011 plugin cards for compactness and maintenance ease.

."'/ I '?lj //i/'i l "'i" :',

\ !' (I ;' , II I (/ ~ .t /J ,... i,'" I ,I i / ( .f /".' / f ( '( ,. /' J ',"l' ~" I' ( ) If' ( !! '" ({ \ ,~ " ' /I/"/li Ii 111',\

IMPORTANT JOB OPPORTUNITIES

TELEMETER MAGNETICS Inc.

2245 pontius avenue, los angeles 64, california • 306 "h" street, n.w. washington 13, d.c.
 DRTR
EDITORIAL
5 Is There a Technical Information Crisis?

f
•I,' FEATURES
6 Facets of the Technical Information Problem
16 Review of In-Line Indicators
31 Elementary Principle of Parametron and Its Application
to Digital Computers
42 Digital Computer Aids in Optical Systems Design

the automatic handling of ARTICLES

information 13 International Conference on Information Processing
36 Computing in the USSR-Work Started on SESM
volume 4, number 36 WJCC Committee Named
38 IBM Announces Transistorized 7070
40 EJCC Site-Philadelphia
41 Computer Symposium-Armour Sponsors for Fifth Year

DEPARTMENTS
15 Datamation Abroad
39 Datamation Book Capsules
3 Datamation in Business and Science
35 Important Dates in Datamation
45 New Datamation Literature
21 New Products in Datamation
29 People Moving Up in Datamation

OVER 24,000 CIRCULATION PUBLISHER & EDITOR Frank D: Thompson

Research and Engineering is circulated without
charge by name and title to the manufacturers and VICE-PRESIDENT Gardner F. Landon
users of automatic, information-handling equip-
ment in all branches of business, industry, govern-
ment and military installations. Qualified indivi- ASSOCIATE EDITOR Santo A. Lanzarotta
duals in the United States and Canada are invited
to request this publication on their company letter-
head, stating position and the firm's business. PRODUCTION MANAGER Bette L. Howe
Available to others by subscription at the rate of
$5.00 annually; single Issue, $1.00 when available.
No subscription agency is authorized by us to so- CIRCULATION MANAGER Martha Knowles
licit or take orders for subscriptions in the United
States, Canada or overseas. SALES MANAGER Richard W. Nielson
Published bi-monthly by the Relyea Publishing
Corp., Frank D. Thompson, president. Executive and
Circulation office: 103 Park Ave., New York 17, NEW ENGLAND DISTRICT MANAGER Warren A. Tibbetts
N. Y., LExington 2-0541; Editorial and Advertising
·office: 10373 W. Pico Blvd., Los Angeles 64, Calif., 434 Webster St., Manchester 4, N. H., NAtional 5-9498
BRadshaw 2-5954. Published and accepted as a
controlled circulation publication at Indianapolis,
Indiana. Copyright 1958, The Relyea Publishing EASTERN DISTRICT MANAGER Richard W. Nielson
Corp. The trademarks R/E and Research & Engi-
neering are the property of the Relyea Publishing 103 Park Ave., New York 17, N.Y., LExington 2-054}
Corp., registered with the U. S. Patent Office.
R/E, The Magazine of DATAMATION, wel-
comes manuscripts of three general types: applica- CLEVELAND DISTRICT MANAGER Albert Pudvan
tion articles dealing with small, medium and large
electronic data processing systems; articles cover- 3537 Lee Road, Shaker Heights 20, Ohio, WAshington 1-9370
ing the design of components, sub-systems and sys-
tems; and other general articles of interest to the MIDWEST DISTRICT MANAGER Gilbert Thayer
data processing industry. Material submitted
should be accompanied by pictures and illustrations 201 N. Wells St., Chicago 6, III., Financial 6-1026
when possible. Unsolicited manuscripts to be re-
turned to writer should Include return postage but
. editor assumes no responsibility for their safety WESTERN DISTRICT MANAGER Hamilton S. Styron
or return (although all reasonable care will be
taken). 10373 West Pico Blvd., Los Angeles 64, Calif., BRadshaw 2-2643

Research & Engineering, September/October 1958 1

 . .. . will quickly and automatically plot
your punched. card and tape stored re-
search, test, and statistical data.
Data Input
Punched Cards: Read with commercially avail-
able equipment.
Punched paper tape: Read with available
accessory equipment.
Keyboard: Manually operated (supplied with
plotter) .
Data Output Available
Single points, 12 distinctive symbols, or con-
tinuous line.
Output Speed .
Points and symbols: Up to 50 points per minute.
Continuous line: Up to 25 points per minute.
Output Accuracy
Points or symbols: 0.05 % of full scale.
Continuous line: 0.1% of full scale.
•

ELECTRONIC
MANUFACTURERS- DEVELOPERS: ANALOG COM-
ASSOCIATES
PUTERS - PLOTTERS - DATA-REDUCTION EQUIPMENT
9~
LONG BRANCH. NEW JERSEY. TEL. CAPITOL 9-1100 D-9
Circle 2 on Reader Service Card.

2 The Magazine of CRTRM.MTi

 DR T R M R TI O:N in business and science

UNIVAC 1105 Construction of a Univac Scientific 1105 computer

SET FOR BUREAU OF CENSUS has been completed by Remington Rand and the first
.unit is being delivered to the Bureau of Census in
Washington for 1960 census use. The Bureau will
eventually use two 1105's in its Washington
headquarters and several others are to be placed in
U. S. universities for collaborative census use.
•• • ,,:Philco Corp;, and the Uptime Corp. of Rawlins,
Wyo., will develop. and market a new high speed
punched card reader, the Speedreader 2000. Equipment
. was inve~ted by Raymond B. Larsen, Uptime president.

HAPPENINGS,C<;>mputer Equipment Corp., is the new name for the

RECENT AND NOTABLE .... 'company formerly known as Digi tron, Inc. The name
'change was effected "in order to clarify the nature
of the business in which our growing electronic
firm is engaged," according to A. C•. Bellanca,
presiden.t of the Los Angeles firm. • • • Reese
Engineering, Inc., Philadelphia, has formed a
Digital Systems Engineering department. It will
provide a complete service from'consultation to
systems assembly with particular emphasis on
special data handling equipment. Lowell S. Bensky
heads ~h~ department • • • Servomechanisms Inc.,
Hawthorne, Calif., has received an order. amounting
'. to nearly a half-million dollars from Lockheed
Aircraft for production of true .airsp~ed computers.

SUBSTITUTE The area of datamation now referred.' to as "data

'BANKING' FOR 'RETRIEVAL?' retrieval" is predicted to grow to major proportions
during the next decade, according to. Bernard S. .
Benson, Benson-Lehner Corp., pr~sidenti Ina recent
talk to leading members in the data processing
field, Benson pointed out.that a~misrio~er.is being
created in the use of this phrase b~cause .the .
. retrieval of information is only'one part of an
overall process. He suggested that this particular
operation bB renamed "data banking" which covers
investment, internal organization and withdrawal.
U. S. NAVY'S Navy Management REVIEW has released a table showing
OVERALL' COMPUTER PICTURE the growth of ele.ctronic digital computer. .
installations in the United States Navy. Figures for
1958 and>1959 are·estimated. Dollar amounts exclude
cost of supporting punched card installations. This
table is reprinted by courtesy of the REVIEW.

END Number of Systems Number of Dollars

. FISCAL Installed People (Thousands)
YEAR Total Bus. Sci. Total Bus. Sci. Total Bus. Sci.
1954 5 1 4 102 49 53 $ 849 $ 234 $ 615
1955 10 2 8 166 64 102 4720 605 4115
1956 20 8 12 262 94 168 3891 1160 2731
1957 29 13 16 : 586 368 218 9184 4234 4950
1958 48 26 22 988 718 270 12299 7272 5027
1959 72 46 26 1242 930 312 23057 15961 7096

Research & Engineering, September/October 1958 3

 computer input-output

"

electroplotters

If you are gathering data in the form of trace or pictorial

records for later processing by a digital computer; or if you

require a graphic presentation of data as part of your computer

output, you will be interested in Benson-Lehner's complete
line of record reading and graphic display equipment. Write to:

~ bensc»n-Iehner
corporation
11930 Olympic Boulevard, Los Angeles 64, Calif.

Offices: Los Angeles; Washington, D. C.; Dayton, Ohio; Service Centers in 28 cities throughout the world.

Circle 3 on Reader Service Card.

4 The Magazine 0/ DRTRM.RTIC N

 editorial

IS THERE A
TECHNICAL INFORMATION CRISIS?
•
The problem of organizing technical information has received more stu"dy, particu-
larly at high levels of the governmental and scientific communities, in the past year than
in the entire previous era of modem science. Does this attention imply that the nation is
suddenly faced with a new situation arising from the Russian satellite program-some
new pattern of circumstances that requires an immediate countering action in order to
prevent losing the technological race?

Obviously not. Specialists concerned with the organization and dissemination of

technical information have been pointing out for many years that the rapid increase in
level of technological activities has far exceeded the concept and rate of growth of in-
formation processing facilities. Those who have responsibility for organizing and budg-
eting research and development efforts have been increasingly aware of the rising cost
and frustrating inadequacies of the available mechanisms for finding recorded know-
ledge. The rapidly increasing accumulation of technical information and the lack of
adequate organization for its utilization imposes an increasing economic burden on our
society and threatens to drown, our scientists and engineers in a flood of paper.
:. ,~

International events have emphasized the time dimension of present-day technology,

and particularly the military significance of time. In technical development the major
time interval is between the initial scientific discovery and the design of a prototype de-
vice. More time is lost-or more time is to be gained-between the laboratory and the
Merritt L. Kastens is an assistant drawing board than there is between the drawing board and the production line. In this
director of Stanford Research Insti-
tute, Menlo Park, California. He period, where ideas rather than physical materials are involved, speedup is most feasi-
has responsibility for certain spe- ble. Thus, from a military standpoint, the information system itself has become a "weap-
cial program developments as well on system." It. is the weapon system on which all military devices, as well as our peace-
as the Institute's research service
departments. time progress, depend.
A native of Chicago, Kastens
received a B.S. degree in chemistry
in 1944 from Roosevelt College in The problem is an old one. What is new is the general appreciation of its seriousness.
Chicago. Later he took graduate Out of this arises hope that major steps toward solution may at last be· u'ndertaken. We
studies on inorganic physical chem-
istry at the University of Chicago. can no longer afford the piecemeal efforts toward fragmentary solutions, which have
During 1944-45 Kastens was a been the only kind of efforts this problem has enjoyed in this country until now.
research chemist with the Armour
Research Foundation in Chicago.
His wartime service as an ensign The acute awareness of the need arrives at a time when there is prospect for major
in t!1e U.S. Navy included duty as help for solution from technology itself. Recent developments in machine systems for
engineering officer on board an at-
tack transport in the Pacific. information storage and retrieval, although still far from adequate for the job at hand,
From 1946 to 1952 he was as- are definitely encouraging. The technique of "organized invention" for the solution of
sociate editor of Chemical & Engi- practical problems has long since proved its efficacy in the applied-research laboratories of
neering News and Industrial and
Engineering Chemistry, working the world. The newer techniques of operations research, systems analysis, and applied
in San Francisco, Chicago and behavioral sciences, are providing increasing evidence that an organized, systematic ap-
New York. He is the author of
several articles on chemical prod- proach to very complex problems is effective, economical, and, indeed, may be the only
uction techniques,' the chemical feasible approach.
industry, and research manage-
ment. He joined SRI in 1952.
The cost of inadequate technical information processing facilities in duplication, in
delay, in the failure to solve both military and civilian problems in the shortest time is
real; it is large; it is constantly growing. The tools for the solution of the problem are
available. Failure to use them now can only delay and make more difficult the effort
ultimately required to prevent chaos in technical communications.
M. L. KASTENS

Reprinted from STANFORD RESEARCH INSTITUTE JOURNAL, Number One, 1958.

Research & Engineering, September/October 1958 5

 FACETS OF THE

TECHNICAL INFORMATION PROBLEM

Technology, so adept in solving problems of man and his environment, must be directed to
solving a gargantuan problem of its own creation. A mass of technical information has been
accumulated and at a rate that has far outstripped means for making it available to those
working in science and engineering. But first, the many· concepts that must be considered in
fashioning such a system and the needs to be served by it must be appraised. The complexities
in any approach to an integrated information system are suggested by the following questions.

by CHARLES p~ BOURNE have recommended a piecemeal attack on components of

and DOUGLAS C. ENGELBART the problem.
Stanford Research Institute believes that the techniques
Recent world events have catapulted the problem of the of systems analysis coupled with an understanding of the
presently unmanageable mass of technical information potentials of machines permit a powerful approach to the
from one that should be solved to one that must be solved. solution of this many-faceted problem. In fact, it may very
The question is receiving serious and thoughtful considera- well be that only by grappling with the problem as a sin-
tion in many places in government, industry, and in the gle, integrated system can a realistic and lasting solution
scientific and technical community. be attained.
One of the most obvious characteristics of the situation However, to deal with the informatiori system as a
is its complexity. A solution to the problem must serve ~ whole, it is necessary first to define its complexities with
diversity bf users ranging from academic scientists en- as great detail as possible. As an aid to the preliminary
gaged in fundamental investigations to industrial and mapping of the system, a study group at SRI polled a por-
governmental executives faced with management decisions tion of the Institute's own professional staff of engineers
that must be based on technical considerations. The solu- and scientists for questions they believe_musLb~nswered----c:--­
--~tion-must--accommodate-nn-almost overwhelmmg quantity before an effective system can be designed. A representa-
of technical and scientific information publicly available in tive list of the questions raised in this fashion is given in
many forms through many kinds of media and in many this article.
languages. The list is impressive, but obviously not exhaustive. It
Some students of the problem, including men with many does confirm the multiplicity of points of view that must
years' experience in various aspects of information hand- be appreciated before this problem can be attacked.
ling, have viewed this complexity and concluded that the Many of the questions require simple factual answers
problem cannot be solved in its entirety. These authorities (see "Data Needed About Information Sources and Serv-
ices," p. 9) . They can be answered by straightforward
Reprinted from STANFORD RESEARCH INSTITUTE JOURNAL, Number techniques of counting, surveying, sampling, and estimat-
One, 1958. .

6 The Magazine of:C RTRE'\r't

 ing. A few of the answers are already available, but the
fact that most questions of this type cannot be answered
• from available sources emphasizes the pressing need for a ;a proposal for a national
much better quantitative assessment of the size and nature technical information service
of the information problem before a rational attempt to
solve it can be undertaken.
Another group of questions involves essentially matters Members of Stanford Research Institute have long
of national and scientific policy that ultimately must be given thought to the increasing disparity between the
answered arbitrarily. Data and analysis can give guidance accumulation of new knowledge and the means for or-
to the answers but the ultimate decision will be based on ganizing it for widespread utility. With this problem
judgment of relative needs and relative values. brought into sharp focus by recent events on the inter-
national scene, the Institute believed it appropriate to
formalize its views on the magnitude of the problem
questions relating to policy and to suggest a possible solution. In January, a draft
program for a National Technical Information Service
What are the specific aims of the program? was prepared and copies distributed to members of the
President's staff, to selected members of Congress, to
Will the system start with only new information? Or will various agencies within the federal establishment, and
it process back literature, and, if so, how far back? to industrial leaders and technical societies, all known
to be concerned over the state of technical informa-
Will the Service process requests from allied countries? To tion affairs.
what extent? Will it coordinate with the Soviet Union? This document describes a program to solve the na-
tion's technical information problem through the es-
Can part of the operations be done abroad? What about tablishment of a national service for the collection,
translations? processing, storing, retrieval, and dissemination of
Will an international classification, indexing, or retrieval scientific and technical information from both foreign
and domestic sources. The program comprises five
system be adopted or promoted? phases, interrelated and partially concurrent:
Will the system be designed to serve the brilliant, the 1-Establish a central organizing and administer-
sophisticated, as well as the more unsophisticated? ing, federally constituted Agency.
2-Determine the gross dimensions of the problem.
Will the Service be financially self-supporting?
3-Establish an interim information center using
Will big business have any better access than small busi- existing services and techniques.
nesses or individuals? 4-Analyze the factors that determine the design
Could a private citizen or scholar afford to use the Serv- and operation of an ultimate National Technical
Information Service.
ice?
5-Encourage present and initiate additional re-
How will prices be established for the Service? search and engineering development programs
leading to systems and equipment necessary to
What is the range of subject matter to be included? implement the ultimate National Technical In-
Will classified information be included? formation Service.
This proposal, and others, for solution of the prob-
Will safeguards be established to insure that classified in- lem are currently under study by the interested bodies
formation is kept under proper control? of the nation. Meanwhile, at the Institute, study of
various phases of the technical information problem,
What type of information should be included? Books
both in the gross, and of specialized aspects of data
(texts, tables)? Technical and trade journals? Confer- handling, storage, and retrieval, is continuing.
ence proceedings and papers presented but not pub-
lished? Industrial and government interim and final
project reports, etc.? Operation and instruction man-
uals? Patents? Manufacturers' catalogs? Newspapers
and general magazines?
Who will control the policy in the matter of designing,
Who will be responsible for selecting the material to be establishing, and/ or operating the Service? An ap-
included? pointed committee, such as for the NACA? A civil
What protection will be provided users who want their servant? A political appointee? A committee elected by
queries to remain confidential? scientific organizations?
Should service be provided outside the technical commu- Would it be feasible to establish legal authority to speed
nity? To congressmen? Executives? Businessmen? up the standardization and coordination of existing
High-school students? facilities (such as the F.C.C.)?

Research & Engineering, September/October 1958 7

TECHNICAL INFORMATION PROBLEM

Should the Service act as a publisher for collections of

•
Who is competent to design, establish, and/or operate the
System? Would this be a civil-service organization? papers (reprints) in very new and special fields?
Could the objectives of the Service be 'achieved by ex- How will the priority schedules be fixed for the Service? •
panding existing government agencie.s (e.g. Bureau of How soon could the Service be initiated? With an immedi-
Standards, the Library of Congress, Armed Services ate manual system? With, an ultimate mechanized
Technical Information Agency)? system?
If the Service were not directed by some existing govern- What factors will determine the location? Can strategic
ment agency, ~ould it not be best handled by some dispersal considerations influence the location without
university? adversely affecting efficiency?
Would it be economically feasible for any sort of commer- Is the proposed Service simply an attempt to copy Russia?
cialenterprise or non-profit corporation organized by
Might not an interim solution be to translate and distribute
the professional community, or by private industry, to
the exhaustive Russian abstracts, thus leaving our in-
establish and run a Service which would assure con-
terim energies free for other uses?
tinued social and technical progress?
Might it not be better to reduce the amount of literature
If we must look to the federal government' for support,
produced rather than go to the tremendous expense of
what residual responsibilities remain with the profes-
providing super-service for all of it? Can a quality filter
sional societies? Should private groups continue to
be applied to this output?
sponsor special collections?
Why not allocate federal money to support more' direct
What economic and political limiting factors, exist with
interchange between working scientists? Perhaps more
respect to the freedom one would have in utilizing or
meetings, special conventions, seminars, etc., would be
changing those organizations already actiye, in the
more economical than better literature processing?
documentation field, and whose existence could be
Couldn't the money be better spent on education to
over-shadowed by a national Service?
achieve a given increase in scientific effectiveness?
What about copyrights? Would royalties be forthcoming
Could a substantial portion of the information problem be
to the owner of the copyright if the Service distributes
solv~d by teaching the users more about present-day
the material? What will be the impact on the technical
documentation techniques?
publishing industry?
.-
questions requiring research

Some of the questions' posed to the study group will re-

quire considerable study and research to produce valid
answers. The research will be in many fields-in the social
as well as in the natural sciences. Some of the study must
be quite profound-even theoretical. Some will be more
straightforward. Many of these questions must be an-
swered before the policy decisions implied in the previous
group can be made with confidenbe.
Can we separate apparent need, influenced by present con-
cepts and experience, frorri real need? Lack of aware-
ness of the potentialities of recehtli developed melhods
(or methods not yet developed) canea!;'ily result in an
,~~~~unimaginative~formulation~of~the-p6ssibilities~and-{)p~~~~­
portunities for advantageously using recorded informa-
tion.
How will users' habits and needs evolve as a good System
becomes available?
How are the information needs of a user affected by his
age, educational level, profession, type of position
held, etc.?
What are the characteristic information needs of the basic
(academic) scientist? The applied researcher? The en-

8 The Magazine of IJ RTRl'"IlPlTI 0 ~J

• data needed about information sources and services

Before the designers of an overall How much has been produced to date in the various tech-
information center can sketch in the nical subject categories in journal, book, and report
outlines of the System problem, a form? What is the physical mass of each? Are back
large amount of data about the in- copies available?
formation input and the existing in- What libraries with technical collections, abstracting
formation services must be collect- services, indexing services, and translating services
ed. Some of the kinds of essential data are suggested by are in existence? Where is each located? What is its
the following. organization? How is it financed?
What subject fields are covered by the various journals,; What is the size and training of the staff of the various
books, and reports? And in each case, in what depth? technical-information handling
What are the physical sizes of journals, books, and re- or processing organizations? In
ports? Page size and number of pages? Frequency of
publication? Kind and size of distribution? Cost or
subscription price? . . :~~~p;:~~o
terlal ?
~a;~l: c~~;~~~z:~:~ ..~."..,.,;-=-.-.'~~e~
~ ,. .,;.·.~l:·
~'. ~ l-

In what language(s) do the journals, books. and reports In what field(s) does each informa-:'~'
appear? tion handling or processing unit
Does each have an index? Are abstracts' published, and operate? .
where? Where is the information indexed? What classification and indexing systems are in use?
Who, principally, are the contributors to the technical What is the normal time between publication of a docu-
journals? Who selects or reviews ment and its appearance in the libraries? When is it
papers for publication? How abstracted? Indexed? Translated?
long, generally, between prepar-
ation and pUblication? What are the types and numbers of scientific and techni-
cal people using libraries, and the abstracting, index-
Are microfilm copies of books, jour-
ing, and translating services? In what ways does the
nals, and reports available?
technical community feel it is being adequately or
Who are the publishers of technical journals, books, and inadequately aided by these services?
reports? Where is each located? And how long in
operation? Would the various libraries and services be amenable to
negotiation of changes or increase in area of coverage,
How is each publishing operation financed? or other changes of service, to fit
What are the policies and objectives of the respective pub- ~
lishers in each field? .
What fields of science and technology does each publisher
,
~11 ,I '] r;l;:~;:~ ;
a reasonable, overall system, if
;i~~:~~~ent controlled and sub-
operate in? In what fields does each concentrate or ....."'~ , ~::.;,":i•.,""~;,/l: ;;,:~
What are the charges for service by
specialize? libraries? Abstractors? In-
In what language(s) does each publisher produce his dexes ? Translators? Which of these services are self-
journal(s), books, or-reports? supporting?
Could publishers of journals, books, and reports provide Are special compilations of abstracts, bibliographies, or
paper tape or other machine-readable copies of their translations available? And for what fees? How long
works? At what cost? required to provide such special services?

gineer? The decision maker? Are they all equally. criti- How many scientists and engineers have a definite pro-
cal, or is the "applier" of knowledge the one with the gram of "keeping up with the literature"? How much
biggest problem? time would they "like to spend?" What keeps them
What is the role of information retrieval, storage·, etc. in from spending more time?
the decision-making process of the research worker, How much of the literature that would, with reasonably
engineer, scholar, administrator, etc? high probability, be useful to a scientist or engineer, is
How much use does the scientist and engineer make of caught by him now by his own regular surveillance of
the facilities that are presently availa.ble? the literature? How far out of his way will the average
By what processes does the scientist and engineer keep user go to be sure that he hasn't missed some possible
abreast of the advances in the art now? What are the information ... considering the usual distracting pres-
relative importances of each of these proees,ses? sures on him, his familiarity with the sources, etc?

Research & Engineering, September/October 1958 9

TECHNICAL INFORMATION PROBLEM

How many pages of literature in various categories rela-

tive to the level and interest-area of the user can we
the' soviet approach to the •
expect him to scan or search' for his different infor-
mation needs? information problem
What are the relative merits of the different types of refer-
ence information services with regard to the user and The Soviet Union has a comprehensive technical in-
his needs, desires, habits, and limitations? formation system in operation. In 1952 the Soviet All
Union Institute of Scientific and Technical Informa-
What are the relative importances of the users' various tion was established in Moscow. By 1957 the Institute
informational needs? On one hand, he needs to know had a permanent staff of 2300 translators, abstractors,
the newsy items such as who is working on what, what and publishers. This staff is supplemented by more
his current attack is, who disagrees with whom and than 20,000 cooperating professional scientists and
engineers throughout the U.S.S.R. who offer their
basically why, etc.; and on the other hand, he also services as part time translators and abstractors in
needs to be able to study in detail the carefully written their specialized fields.
treatises that may have bearing on his work Can these
The Institute publishes 13 "abstract journals"
different kinds of needs be met by a single system? which annually contain over 400,000 abstracts of te'ch-
What are the special information requirements for different nical articles from more than 10,000 journals originat-
specialty fields? ing in about 80 countries. It systematically translates,
indexes, and abstracts about 1400 of the 1800 scientific
Does the user, when he goes outside his special field for journals published in the United States.
supporting information, want information in different To reduce the time between the initial appearance of
form or different levels than which he seeks in his own the more important information in any of the world's
field? For instance, would he be looking more for journals and its reaching the hands of Soviet scientists
"cook-book" techniques or fo~ survey-type information? and engineers through the normal route of the ab-
stract journals, "Express Information Journals" are
How valuable would broad, multi-disciplinary searches be also printed. These carry summary information on
if they could be conducted effectively? How great is foreign technological developments within two or
the problem of differences in nomenclature between three weeks after their receipt. Reports made on the
work done indicate that it is not only comprehensive
fields? but also of high quality.
What type of questions now go unanswered at the li- The Institute provides numerous other technical in-
braries? formation services,such as provision of bibliographies,
micro and full size copies of original printed material,
Isn't the main problem of information retrieval one of technical dictionaries, foreign-language dictionaries,
identification-since people so seldom express satis- and other variea source material.
factorily their needs to the documentalist? The Institute maintains an extensive program aimed
What are the major limitations in the various methods to introduce machine methods to information handling.
This includes translating machines, and mechanisms
presef{tly used in classifying and inaexing scientific for codifying, storing, ,and retrieving technical infor-
literature? mation. Significant progress by the Soviet All Union
Institute towards information mechanization methods
Is the problem that the information now is just not avail- and systems is reported.
able at all, or is it that it is just hard to find?
Why aren't the existing services that process technical'
information -satisfactory?
How many places does a user of each discipline have to
look for index listings of, a given special interest?
..
What increase ,in efficiency of the scientist or engineer
How can the processing of recorded information be would result from improving the accessibility of re-
planned so that it can be effective in spite of humah corded information?
limitations, or of limitations in numbers of human be- What are the probable net benefits, short 'and long'range,
ings? of an effective information Service to military, indus-
How much is missed by technical people leaning too heav- trial, commercial, scholarly, government groups?
ily on librarians? Can dollar costs be derived for reasonably well-proven
What relative gain in efficiency could be achieved by delays and duplications, and can the total national loss
integration, merging, or better managing of existing rate due to this problem be realistically estimated? Can
documentation services? it be determined that the expense of delay and duplica-

10 The Magazine oj CRTRE\kHT!

t

tion now is .greater than that of establishing and oper- niques, that a highly mechanized system will be possible.
ating an information Service? How soon can an interim system be functioning?
•
What is the lack of an information Service costing govern- How much can be done just by concentrating on abstract
ment agencies? distribution and better dissemination techniques?
Can the savings in Federal money now spent on other in- Would it be feasible for the abstracting publications to use
formation programs be diverted to a national informa- a standard format and type font, such that mats (or
tion Service? something similar) could easily be distributed to other
What are the relative costs and characteristics of different interested publishers, thus saving printing expenses?
reproduction techniques that might be applicable to
What technical societies could cooperate to publish a
some of the dissemination and rriassive processing
single journal instead of numerous splinter journals?
problems of an information service?
What about the scale of the Service? Does it have to. be a
What are the techniques and costs involved in keeping up
big system or nothing?
and in using .large mailing lists in taking care of distri-
bution of journals, etc.? Does "having a large information Service" necessarily
What are relative costs of providing the information in mean the physical collection of all activities at one
micro· form as against making original-size photo central location?
copies? Would a group of smaller centers, for specific fields, be of
Of the currently-operating abstracting services, how many greater utility and more tractable?
are operating merely to satisfy an obligation of a pro- Would a collection of special libraries be more useful?
fessional society that would rather have somebody else What can a national service provide that is different than
do the abstracting? what is now available? Is this to be an entirely new
What services does the Russian All-Union Institute really type of service, a real advance in the state of the art,
provide? What is the reaction of a Russian scientist to or is it to be just more and better of the same thing?
this information center?
Will the System have a finite capacity? One system might
How important is it to know what the rest of the world is work well with a few million entries, but be hopeless
doing? with a hundred million.
Are any projects or areas of work reported almost exclu- As the System grows in size, will it . be possible to make
sively in foreign literature? changes easily in the classification scheme and bring
What is the expected rate of growth of the system? the old coding into the new scheme?

What are the potential information processing capabilities If a private consultant, with "need to know" established,
of existing mechanical devices? were to work on a government project, how would he
locate and procure pertinent classified material?
What are the theoretical capabilities of existing or antici-
pated machine components which might be applied to Will financial filtering of requests by a uniform fee struc-
the information processing problem? ture be desirable or effective, or would it be necessary
to make non-uniform fee structure so that there is
How often will the system presumably be searched? How essentially some "priority" given?
definitive will the search have to be? What volume of
information should a search produce? How fast should What means can be used to pry loose useful information
the system respond? that customarily doesn't get into the published tech-
nical information channels?
characteristics of the information service Will the service include a positive program. to· declassify
material under security restrictions?
As increasing data becomes available it will become pos- Whatis an acceptable delay in getting· information entered
sible to consider some of the last group of questions'-':'those into this system?
dealing with the desired or necessary operating character-
istics of a comprehensive technical-information processing Will all material in the subject fields be included or will
system. Certainly, the first system implemented would be there be an editor or a censor?
of an interim nature using existing resources, which un- Will an attempt be made to standardize the form of the
fortunately employ largely manual techniques. However, material before it gets into the center? Does the ma-
ultimately it is inevitable, in view of the impressive ad- terial have to be on standard-size ~heets or forms?
vances made almost daily in information processing tech- What happens when the system becomes overloaded?

Research & Engineering, September/October 1958 11

TECHNICAL INFORMATION PROBLEM

Should service to users just be late, or should the serv- Will there be a «special communication network" in which
ice just be less complete? workers in the various specialized fields can easily
How can we protect against freezing the specifications circulate working papers or "think pieces?" A central •
until enough systems work has been done to make clear agency could maintain printing, listing, (in appropriate
what would be optimal? subject-interest categories), and mailing facilities for
this sort of service.
Will the policy makers make sure that the final methods
chosen for a retrieval system are not influenced too Will the information Service be able to retain a file of
heavily by the requirement of compatibility with past questions to be asked of all new input material, thus
systems? providing up-to-the-minute data for standing questions?

Will abstractions be done? What kind? Descriptive? Criti- Will it be possible to stimulate more writing of «review-
cal? Informative? How can we get good-quality ab- the-literature" papers by qualified people in the various
stracts? Should the Service use volunteer abstractors fields, in order to provide guides for other workers?
directly or a staff of full-time abstractors? Or should it Can a partial search be made? (For example, can 1/10 of
allow the' various technical societies to organize their the file be searched and the results checked to deter-
own volunteer abstracting services? mine if further searching is justified?)
Will any effort be made to review old documents, and to Could the information Service operate on a «just search
remove or recode when necessary? Y2 the file for me; I don't need a comprehensive search"
Is a standard (or artificial) vocabulary necessary? How basis?
much work will be required to design and institute What kind of communications network will be needed for
such a vocabulary? the operation of the interim information Service? Will
What techniques and devices can reasonably be developed it be accessible to anyone by telephone or other direct
and applied for facilitating such immediate require- device, such that the searcher can interrogate the file
ments as printing, reproducing, storing, microfilming, directly and at will?
billing, communicating, etc.? Would the Service be available for browsing?
What kind of data-processing system will the Service
What technical-manpower drain would the proposed infor-
need just to keep track of its operation?
mation Service program have on other high-priority
Would the information Service keep a collection of the scientific programs?
original documents?
What professional and educational b~ckground is needed
What special precautions must be taken to store primary for the personnel to operate the Service?
records? Would a duplicate file and collection be main-
tained to prevent disruption of service due to fires, or Could university science students be used part time and
other catastrophes? How much would this cost? during summers to help with the various processing
tasks, as a means of alleviating the shortage of people
What is the useful life of various forms of records? In use? with adequate technical backgrounds?
In storage?
Will there be special training for abstractors and trans-
What will the information Service physically provide in
lators or for documentation and information specialists,
response to information requests?
etc.?
Will the output be in English, or a code that must be
How much research is needed? What research budget is
translated?
reasonable?
Will microform copies be acceptable to the users? If not,
what improvements need be made in order to gain If an information Service were established, how soon
user acceptance? could present partial services by government agencies •
be terminated and funds diverted' to the Service?
Will the information Service output be in a form that the Could some special activities in industrial libraries be
researcher can determine which of the documents are eliminated?
in a locally accessible collection?
These questions, by the very nature of their origin, are
Will the system give answers (e.g., «yes," «no," «5;000 tons random and fragme'utary. Even the full list from' which
in 1945," etc.) as well as references? they have been selected is far from comprehensive How-
Why not periodically publish inventories of research in ever, we have found them a helpful stimulus as well as a
progress, to indicate what research projects are cur- disciplinary aid in viewing the technical-information prob-
rently being undertaken in each specialty field, thus lem in its broadest dimensions. We hope thClt others inter-
helping to eliminate duplication? ested in this problem will be similarly served.

12 The Magazine of DRTRl'v1.MTI C N

 INTERNATIONAL
CONFERENCE
ON

INFORMATION
PROCESSING
Representatives of at least twelve countries will be at-
" tending the International Conference on Scientific Infor-
mation, November 16 - 21, 1958, to be held at The May-
flower Hotel, Washington, D. C. The opening session
address, Sunday, November 16, at 8 p.m., will be made
by Sir Lindor Brown of England, Secretary for Biological
Sciences (The Royal Society) followed by a reception.
The five day conference will be divided into seven
areas, morning and afternoon sessions, during which the
papers of scientists of· various countries and of this coun-
try will be presented and discussed.
Sponsors of the conference are, National Academy of
area five
Sciences, National Research Council, National Science
Area five will review the organization of knowledge for
Foundation, and the American Documentation Institute.
storage and retrospective search: I~tellectual problems
This scientific information conference is expected to
and equipment consideration in the design of new systems.
attract more than usual interest because of the participa-
Dr. Gilbert W. King of the 1.B.M: Research Center will
tion of the U.S.S.R. and Czechoslovakia.
be the discussion panel leader. Twenty papers are to be
area one reviewed and countries represented in this area are-the
Beginning the morning of November 17, area one pro- U.S.S.R., Netherlands, France, Great Britain and the U S.
poses the knowledge now available and the methods of One paper is entitled, "On the Coding of Geometrical
ascertaining scientist's requirements for scientific literature shapes and Other Representations, with Reference to
and reference services. Discussion panel· leader is Dr. Archaeological Documents." This has been submitted by
Philip Morse, Department of Physics, Massachusetts In- Jean-Claude Gardin of Centre Nationale de la Recherche
stitute of Technology. Among the thirteen papers to be Scientifique in Paris.
presented will be one from Czechoslovakia entitled, «Sys-
area six
tematically Ascertaining Requirements of Scientists for
The topic for area six is the organization of information
Information," by Jiri Spirit and Ladislav Kofnovec of the
for storage and retrospective search: possibility for a
Prague Research Institute for Materials and Technology.
general theory of storage and search. Leading the discus-
area two sion will be Dr. John Tukey, Department of Mathematics,
Leading the discussion on area two's topic is Dr. Elmer Princeton University. Among the six papers to be pre-
Hutchisson, American Institute of Physics. Subject matter sented will be, «The Structure of Information Retrieval
- function and effectiveness of abstracting and indexing Systems," by B. C. Vickery, Imperial Chemical Industries,
services for storage and retrieval of scientific information. Ltd., Welwyn, England.
The fifteen papers to be presented include a contribution
area seven
from Russia: «On the Functioning of the All-Union Insti-
On Friday morning, November 21, the concluding area
tute of Scientific and Technical Information of the Aca-
seven is devoted to a subject that is surely of universal
demy of Sciences of the U.S.S.R.," by A. 1. Mikhailov of
interest. To be discussed - the responsibilities of govern-
the Moscow academy of which he writes.
mental bodies, professional societies, and other groups,
area three to provide improved information services and promote
Subject for area three is the effectiveness of scientific research in documentation. Mr. Verner Clapp, Council
monographs, compendia, and specialized 'information cen- on Library Resources, Inc., will be heading the panel.
ters in meeting the needs of scientists. Present trends and Nine papers will be reviewed and include contributions
new and proposed techniques and types services will be from ·France, England, the United States and one from
elaborated. The panel leader is Dr. Alexander King, Euro- the Union of South Africa entitled, the «Responsibility
pean Productivity Agency. Five papers are scheduled and for the Development of Scientific Information as a National
include one entitled, "Recent Trends in Scientific Docu- Resource, by Hazel Mews, Department of Librarianship,
mentation in South Asia: Problems of Speed and Cover- University of Witwatersrand, Johannesburg.
age," by P. Sheel of Insdoc National Physical Laboratory, There will be exhibitors at this conference from manu-
New Delhi, India. facturers of equipment utilized in the field of data storage
area four and data retrieval.
Organization of information for storage and search . For conference contact, see Important Dates in Datama-
comparative characteristics of existing systems, will be tion, page 35.

Research & Engineering, September/October 1958 13

 •

D·EFE·ND
Today, creative engineering at Hughes is on the move
to DEFEND ... to counter the threat of aggressive ac-
tion with electronic speed and precision.
Hughes in Fullerton, California is now creating these
systems for total defense: '

ELECTRONIC SCANNING RADARS. with beams that

provide three-dimensional protection.

DATE PROCESSORS. which monitor the action of hun-

dreds of aircraft and store the shifting tactical situations
for high-speed assignment of defense weapons.

ELECTRONIC DISPLAY SYSTEMS which present tactical

information in symbolic or language form.

Also under development are new three-dimensional

radar systems for' inst~llation 011 surface and subsurface
na~al ~e~sels. Study programs have been' initiated in
,radar, ~omputers, displays and integrated defense systems.
To further these studies, creative engineers 'and physi-
,
cists .are now needed. These' positions pr~mise unusual
r----------------~-------~
reward and opportunity for growth. Inquire further ... I I
'for-information write to Mr.T. D. Wike at the address I I

at right.
i
I
HUGH,ES i
I
IL ________________________ I
~

GROUND SYSTEMS
PERSONNEL SELECTION AND PLACEMENT
Hughes Aircraft Company
, Fullerton, Orange County, California
Circle 4 on Reader Service Card. ':

14 The Magazine o/:cRTRiv1,~TiCN

.I!I.
·~I~.c=J··­
~~-.-

···1-=
•••
DRTRMRTION abroad
·~.u
IN LONDON:. Based on developm~nt work carried out at the
ANALOG ANALYZER DEVELOPED Imperial College of Science, London, on a
transformer analog system devised by C. L.
Blackburn, an advanced;transformer analog analyzer
has been recently built and installed at the Witton
Works of General Electric Co., England. The'new
analyzer is called the WINA (Witton Network
Analyzer) and operates on 50 cycles supply. It has
a livery highll inherent' accuracy ofO.l%,according
to the manufacturer. The equipment is arranged in 4
sections,· each consisting of a central plugb6ard
with 52 standard.analyzer units arranged in racks.

COMPUTER ElectroData division of Burroughs Corp. has shipped

INSTALLATIONS ABROAD a 205 electronic computing system to .the South
African National Life Assurance Co., Capetown. The
insurance firm· became the first in South Africa to
employ datamation for policy handling • ~ ~ Italy
and Germany have displayed keen interest in
computing equipment that can provide shortcuts to
problems in.aircraft and guided missile'programs,
according to Paul Dennis of Bendix Computer. Dennis
said his firm will install a unit at the University
of Rome for the Italian Air Ministry to be used in
an Italian missile program. He said major West
German industrial plants such as Dornier, Zeiss and
Zahnradfrabrik are vitally interested in this work
• • • an IBM 650 and auxiliary equipmen~ has been
installed by the National University of Mexico in
its. electronic computing center.

IDP GROUP Data processing equipment users in Stockholm, .

ORGANIZED IN SWEDEN Sweden, have formed an IDP Group, composed of ab'out
40 members. Secretary is Peter Hansen of the Swedish
Commerce Bank. Exchange of ideas, viSits to
installations, keeping up with latest changes, and
explor~ng technical data, are,some of the targets
which this group has set for itself. The IDP group
idea is making huge strides ~nd certainly its' future
is very promiSing, according to our Swedish
correspondent •
.' ' IBM SUISSE At the end of May, IBM, Extension Suisse~' opened a.
OPENS, CENTER· IN 'ZURICH 0'aata processing center in Zurich. Officiating at the
"opening ceremony was H. R. Luthy, director of IBM's
Swiss subsidiary. Leading Swis,? personalities and
the president of IBM World Trade~Corpciration, A. K.
Watson, were in attendance. The center has a 650 and
a staff of 12scientists:and engineers under the
supervision of Dr. Jakob.Haller.
- "
JAPANESE'. Word comes from Japan that at least three companies
MARKET COMPUTERS IN ... ' 59 are expected to be producing computers commercially
next year. The Musasino One (see page 31) will be on
exh~'bi t by its commercia:L ma~er at a data processing
~onterence in Europe next year.

Research & Engineering, September/October 1958 15

 I IEJwl
11~[J[o~EJ[cI[[llBJ~IDI •

In two previous issues of DATAMATION we have surveyed two forms of

data display devices. In the January-February issue we presented graphic dis-
play devices, and in the July-August issue - high-speed printers.
This month we offer a survey of in-line visual display devices which are
1;Ised when the human observation of data is required. These in-line displays
. can be used in applications varying from single decades for indicating a par-
ticular channel number, etc., to complex tote board type displays used by such
groups as the North American Defense Command and the Strategic Air Com-
mand in their operations control centers.

TALLY REGISTER
in-line indicator
This in-line indicator is intended to serve as an input link
- primarily for conversion from serial input to parallel
output with simultaneous translation from one code to
another (BCD to decimal, etc.) Model 274 data control
unit is a tape-fed, shift register with parallel readout. Infor- length from two to seven characters. Additional controls
mation is serially read at 60 characters per second and is are provided to permit one-character-at-a-time or one-
shifted through the control unit. When the nnal position word-at-a-time operational modes. (Tally Register Corp.,
has been loaded, a signal is generated to control the out- 5707-37th S. W., Seattle 6, Washington.)
put program. Controls are provided to preset the word Circle 101 on Reader Service Card.

KIN TEL, COHU

digital readout
IIIJ
quirements are % amp at 63 volts AC for each symbol or
digit that is lit. Vertical space of 3% in. is required in the
Model 471 digital readout presents numbers on a single standard 19 in. rack mount. Special models are available
plane, with no overlapping characters. It employs a pro-
with up to nine individual readouts consisting of any
jection system ... providing 7,000 to 8,000 hours of lamp
combination of digits or symbols required. These digital
life. Display provided by the 471 consists of four digits
readouts are designed for use with all types of digital
(decimal and 0 through 9) plus a symbol readout with
converters and encoders. (Kin Tel, a division of Cohu Elec-
symbols +, -, AC, NN, ohm, kilohm, megohm, A/B, tronics, Inc., 5725 Kearny Villa Road, San Diego 12,
AC/B, AC/AC, X. Individual digits are 1% in. high. The
California. )
total display area is 2 in. high by 7% in. long. Power re- Circle 102 on Reader Service Card.

16 The Magazine of CRTRLvr

PATWIN
mag indicator
Numbers or symbols are displayed on this electro-magnetic
indicator at a rate of two per second. Only '.028 amperes
are required for the duration of signal pulse and the dis-
play holds position, without power, until the next pulse.
One moving part, weighing % oz, and mounted on a mini-
ature precision bearing, insures long life and high relia-
bility. Digits or symbols are large enough to be read at
distances up to 25 feet and are free from parallax and
glare, the manufacturer contends. Interlocking construc-
volts. Average response time varies between .5 and .6
tion facilitates stacking. An external dc pulse of approxi-
seconds for maximum rotation. (Patwin Division, Patent
mately one watt is required to actuate the indicator. Indi-
Button Company, Waterbury 20, Connecticut.)
cators are available with nominal voltages of 12, 24 or 36 Circle 103 on Reader Service Card.

MILMAN
digital readout
Presented on this digital readout are the lighted digits
through 9 and decimal point, plus other information such
°
as polarity signs or special symbols. Modular design of
the units allows side-by-side mounting for in-line presenta-
tion of information. Message readout will display, separate-
ly in one panel area, up to three different color-coded
printed messages, greatly increasing the accuracy of visual
observations, states the manufacturer. Message and color
combinations are made up to suit individual requirements. engraving. The units are designed for operation on 6, 14,
The superpositioning of messages in the readout reduces or 28 volts. (Milman Engineering Co., 1831 Pontius Ave-
the amount of panel area required to present a given nue, Los Angeles 25, California.)
amount of information and eliminates panel art work and Circle 104 on Reader Sen'ice Card.

GLOBE
binary decoder
A completely self-contained unit operating from pulsed
input signals on four wires, this binary decoder is a high
speed relay controlled decoding matrix operating into a
single plane illuminated digital readout. Operating speeds
are compatible with modern tape punch and printing
equipment. Various input circuits are possible. These in-
clude a parallel input containing signal and no signal
conditions to indicate coded information, or a pulsed sys-
use in low cost remote readout or telemetering systems
tem in which power or ground pulses are fed to the input
which require serial transmission of information, accord-
terminals. The inputs may be scanned serially from a me-
ing to the manufacturer. Mechanical synchronization of
chanical or electronic multiplexing device. It has been
scanning switches on a start-stop system has proven suc-
found that a high speed relay at the input to a multiplexer
cessful in many installations. (Globe Industries, Inc., .525
will provide power and ground pulses to operate a bank
Main Street, Belleville 9, New Jersey).
of these decoders. This unit is particularly suitable for Circle 105 0/1 Reader Sen'ice Card.

Research & Engineering, September/October 1958 17

I. E. E.
in-line display
This is a new unit which will display any letter A through
Z, any number 0 through 9 and + or - symbols. Designed
primarily for use with electronic data handling and pro-
cessing systems the alphanumeric display unit features ..
one-plane presentation. Characters are of uniform size and
intensity and readability is insured from any angle of view-
ing, it is claimed. The unit utilizes a 12-b.ar matrix, each •
bar having a miniature projection lamp behind it. By pro-
jecting each bar through its own lens system onto a sin-
gle-plane screen, and by the selection of the proper com-
bination of any of the 12 bars, any desired letter, num- manufactures a 90 degree inline display and a standard
ber, or symbol may be formed. The size of the character in-line display line. (Industrial Electronic Engineers, 3973
displayed is approximately % in. high x % in. wide. Size Lankershim Boulevard, North Hollywood, California.)
of the viewing screen is 1-15/16 in. wide. This firm also Circle 106 on Reader Service Card.

BECKMAN
in-line indicator·
Model 5916 in-line indicator displays up to six red digits
1 % in. high. Each digit is formed by a pattern of seg- erates in response to a 1-2-2-4 binary code supplied by
ments, each segment evenly illuminated by a grain-of- the manufacturer's counting instrument or some other
wheat lamp operated at reduced voltage for long life. source. The in-line indication does not change while the
This display differs from indicators using stacked lucite counter is actively scaling, but changes only at the end
plates or stacked figure wires in that all digits are formed of a counting interval, at which time it displays the new
on the same surface plane. The surface image was em- total. Maximum display rate is 15 presentations per sec-
ployed in order to create an unobscured display which ond. (Beckman/Berkeley Division, Beckman Instruments,
could be read from wide angles. Red color is used to Inc., 2200 Wright Avenue, Richmond 3, Califor~~lia.)
permit readings in high ambient light. The indicator op- Circle 107 on Reader Service Card.

UNION SWITCH
digital indicator
A motor driven digital indiCator operates on a direct· drive
basis, does not rely on the use of intermittent drive me-
chanisms to position the characters. Numbers o through ..
9 and two blanks are displayed in sequence,in response
to four bit binary coded decimal input. Openiting on an
open circuit principle, complete code agreement of both
binary ones and binary zeros is checked to assure positive
and correct positioning of the indicator. The one inch
character indicator utilizes the major portion of its frontal
to exclude dust and moisture, thus providing operation
area for display. Its ability to operate as a nonvolatile bi-
under widely varying environmental conditions. (Union
nary readout, in conjunction with its inherent capabilities
Switch and Signal, Division of Westinghouse Air Brake
of storing binary data, is claimed to be its most desirable
Company, Swissvale, Penna.)
feature. The indicator is mounted in a gasket sealed case Circle 108 on Reader Service Card.

18 The Magazine of :CRTR!\:i.J~Tg

 BURROUGHS
indicator tubes
Nixie Indicator Tubes are completely electronic devices
which display numerals or alphabetical characters in a
common viewing surface. Due to this fact, a number of
• Nixies, when placed side by side, form an "in-line" presen-
tation which reduces operator fatigue and makes for an
easily read display. The Tubes consist of 10 metal cathodes
formed in the shape of numerals or other characters and a
common anode which provides a uniform electrical field.
These elements are enclosed in a glass envelope which is
filled with neon gas. By applying a potential between one
at a 50 KC rate. There are four basic types available: BD-
of the cathodes and the common anode, a bright neon
200-S miniature, 6844-A standard, 7153 super, and BD-
glow is caused to form around the selected element, thus
307 jumbo. (Electronic Tube Division, Burroughs Corpora-
causing it to appear clearly and distinctly at the single
tion,Plainfield, New Jersey.)
viewing surface. The presentation can be caused to change Circle 109 on Reader Service Card.

FISCHER/PORTER
digital indicator
Featured in this direct-reading digital indicator, designed
primarily for use with this company's terbine Rowmeters,
is a completely digital system, operating from the fre-
quency output of the turbine meter, which takes full ad-
vantage of the linearity and repeatability inherent in the
turbine meter primary. Some features: in-line, illuminated,
digital indication. Inch-high numbers providing digital
readout up to 25 feet away. Direct reading of Row, with
manual correction for specific gravity as an optional fea-
ture. True integrated Row rate - integration is performed
range switching is virtually instantaneous even at maxi-
during the sampling interval and the indication is correct-
mum accelerations. Customized time base is also featured
ed immediately following each sampling period. An octu-
along with 'manual specific gravity adjustment. (Fischer
pIer provides information at many times the output fre-
and Porter Company, 758 Jacksonville Rd., Hatboro, Pa.)
quency of the primary. Automatic, frequency-controlled Circle 110 on Reader Service Card.

Non-Linear Systems
+"°1 9 5 9
.0
RATIO

digital ohmmeter
Series 20 digital volt-ohmmeter has been designed speci-
fiically for missile and weapons system check-out. It auto-
matically and continuously measures and displays dc volt-
age, voltage ratio and resistance. The unit is available in
four and five digit models. It makes three readings per
second with accuracy to .01%. It is the first combination
volt-ohm ratio meter having functions selected by a panel using precision stabilized wired-wound resistors and con-
switch or remotely by means of electrical signals. Among trolled by mercury relays. Series 20 has high input im-
exclusive features claimed by the manufacturer are inter- pedence, 10 megohms to 1000 megohms, operates card
changeable plug-in circuit boards, illuminated numerical punches, electric typewriters and other printing devices.
read-out that snaps out for bulb inspection (with remov- (Non-Linear Systems, Inc., Del Mar, California.)
able non-glare hood), a Wolff-Poggendorf potentiometer Circle IlIon Reader Service Card.

Research & Engineering, September/October 1958 19

 •

Significant data from airborne vehicles ... missile and aircraft structures ...

power plants ... complex weapons and control systems ... are being gathered

by Consolidated Systems designed to produce useful information in a

minimum of flights or firings. These are systems of proven design, based on

unmatched experience in recording, converting, and reproducing selected

in-flight data. Read the complete story in BULLETIN CEC 3014-X5. I_~
systems division Co nso Ii da ted E I eet rodyn am ics

300 N. Sierra Madre Villa, Pasadena, California / Offices in Principal Cities Throughout the World e
Telemetered and recorded data
are made available in the forfrt
of analog or digital magnetic
tapes, punched paper cards or . f

tapes, visual displays, and

direct-readout oscillograms.

Circle .5 on Reader Service Card.

20 The Magazine of CFlTFlMHT~ i'J

 I> new products in 0 R T R

all the indicating lights, switches, in-

• transistors
ternal wiring, power and logic con-
A new line of 30-volt, one-half ampere trols essential to the operation of an

,,-'-
PNP germanium transistors for use in analog computer system. A writing
data processing equipment has been shelf, at desk height, extends across
the console unit. Occupying the cen-
ter of the shelf is a raised turret sup-
porting the control panel which places
all the computer co~trol and moni-
"""" toring switches within reach of the

"<~~
sory permits standard 80 column
operator. All indicators are centralized cards, punched in conventional num-
and readily visible. For information

;~~.
eric or alphanumeric codes, to be pro-
write COLORADO RESEARCH cessed at high speed. It also handles
CORP., Broomfield Heights, Colo- high speed tabulation of numeric or
rado or use reader service card.
'I
designed for medium power amplifier
t Circle 151 on Reader Service Card.
alphanumeric printed copy. Informa-
tion may be read or punched at the
rate of 100 cards per minute by sum-
and low frequency, high current mary punches such as the IBM 514 or
switching applications. The four 523 (pictured at right). Output infor-
models available: 2N524, 2N525, voltage-to-digital converters mation may be tabulated at the rate
2N526 and 2N527. They have a tri- of 100 lines per minute by the IBM
angular lead arrangement and are This four-decimal digit unit, model 402 or 403. Three IBM units - one
housed in JETEC TO-5 package. V16-AD, utilizes 17 printed circuit for input, one for output, and a third
With a collector current of 20-milli- cards and comes in a standard hous- for either input or output - may be
amperes and a voltage of I-volt in a simultaneously connected to the CA-2.
common emitter circuit, the 2N524 Cards may be read or punched in
has a typical forward current gain of standard code and alphanumeric and
35, the 2N525-52, the 2N526-73 numeric characters may be mixed.
and the 2N527 -91. Beta holdup on Special characters, indicated by mul-
all models is typically 75% of the 20- tiple holes in card columns may be
milliampere value at 200-milliam- ing. Eight printed circuit card posi- read or punched. For information
peres. For information write GEN- tions are left vacant for auxiliary write BENDIX COMPUTER, 5630
ERAL ELECTRIC, Semiconductor equipment. For additional flexibility Arbor Vitae Street, Los Angeles 45,
Products Dept., Syracuse, N. Y. auxiliary equipment may be integrated California or use reader service card.
Circle 150 on Reader Service Card. Circle 153 on Reader Service Card.
within the housing for precision po-
tentiometer testing or analog compu-
ter readout. Inputs are full scale, 1,
analog computer 10 and 100 volts. Input impedances digital module
are lK, 10K, and lOOK respectively.
Model SM-10 is a delay unit with
Model 200 is enclosed in a console A high impedance input amplifier can
amplifiers which does not incorporate
unit equivalent in size to three stand~ be incorporated for 100 megohms in-
any logic. It has three basic sections
ard racks. Included in the console are put impedance. The unit occupies
5 1,i" of panel space. It is capable of
up to 1000 independent conversions
per second. For information write
ADAGE, INC., Dept. P, 292 Main
St., Cambridge 42, Mass. or use card.
Circle 152 on Reader Service Card.

accessory ca-2

the patchbayassembly, a .01% refer- Versatility and rapidity of input-out-

ence voltage divider, a precision panel put operation for the G15 computer
mounted vacuum tube voltmeter, and is increased by this unit. The acces-

Research & Engineering, September/October 1958 21

NEW PRODUCTS

. . . a driver circuit, a magneto-stric- actions performed by a series of sub-

tive delay line, and an amplifier cir- systems. The selection of symbol,
cuit. Its input is designed to be driven plotting value, paper position, etce-
by a standard LE-10 logical T-PAC tera, all are performed along the
or other logical circuitry with right-in principles of a digital servo. Values
and erase control exercised at this and instructions are entered into the
point. The output of the SM-10 is secondary buffer. Upon receipt of an
similar in low driving capability to an instruction signal the mechanical sys-
LE-10 package. Both assertion and tems are motivated until their posi-
negation outputs are provided. Delays tions is in agreement with the con-
up to 560 microseconds are available. tents of the secondary buffer. After
For information write COMPUTER this, the act of printing occurs and
CONTROL CO., INC., 92 Broad St., the secondary buffer is reset. For in-
Welle sly 57, Mass. or use reader card. formation write TALLY REGISTER
Circle 154 on Reader Service Card. tection, according to the manufac- CORP., 5300 - 14th Avenue, N. W.,
turer. The rigid structure of the cap, Seattle 7, Washington or use card.
which will not distort with tempera- Circle 157 on Reader Service Card.
tape punch ture changes, allows complete freedom
of handling of the assembly line with-
This punch features .046 in. diam- out the necessity of tweezers or special
eter feedhole, .072 in. diameter code- tools. Another claim-the permanent
hole and is spaced .100 in. in both shaft encoder
sealed protection of the cap lowers
production costs for users. The fitted Special logic used in these airborne
cap, which covers the bobbin Ranges, binary shaft encoders automatically
adds nothing to the bobbin width or performs brush selection within the
window area. Less wire is required for
windings. The caps are unbreakable
in normal use and handling. For in-
formation write MAGNETICS, INC.,
Box 391, Butler, Penna. or use card.
directions. The standard punch is sup- Circle 156 on Reader Service Card.

plied with the feedholes in line with

the codeholes. The standard location encoder eliminating electronics other-
of the feedhole is .394 in. from the digital plotter
wise required for double brush V-
guide edge of the tape. The punch scan systems. Digit and complement
operates on a single cycle basis by Model 201 plots visibly at speeds up
to eight points per second with four are simultaneously available in natural
demand and can be operated at any binary. Design provides noise-free life
speed up to twenty seven cycles per symbols or up to twenty per second
by preventing commutation of load
second, depending upon the external currents on the disc, yet allowing con-
requirements of the equipment feed- tinuous or pulse reading at up to 200
ing it. The punch requires one elec- RPM. An error detecting scheme pro-
trical input for each codehole con- vides a means of detecting and reject-
nection. Two cam-actuated auxiliary ing erroneous readings. For informa-
contacts are provided on the drive tion write NORDEN DIVISION,
shaft. For information write PRECI- United Aircraft Corp., Wiley St., Mil-
SION SPECIALTIES, INC., 1342 ~I
ford, Conn. or use reader service card.
East 58th Street, Kansas City, Circle 158 on Reader Service Card.
Missouri or use reader service card.
Circle 155 on Reader Service Card. f

with random symbols. Results of com-

bobbin cores puters can be plotted immediately. wroc 452
The plotted result is a series of me-
New «Poly Cap" tape wound bobbin chanically generated printed impres- This unit is designed to increase the
cores are capped with a glass pol- sions on paper or vellum. The act of co selector, pilot selector and digit
yester which offers complete core pro- plotting is a complex of mechanical selector capacity of the IBM 407

22 The Magazine of DRTR~4HTa ON

 NEW PRODUCTS

nounced. The system consists of three tape reader

basic seven-pin miniature plug-in
modules, which are energized by a Model C301 is a photoelectric per-
100 KC R. F. power supply. The first forated tape reader. It is available in
of the three modules, type NZ100, is versions to handle anyone of the
a "not" plug-in. It is comprised of a
series-type pulse magnetic amplifier.
The second module in the system is
the OZ100. It consists of three silicon
double anode, zener diodes with ap-
propriate zener breakdown voltages.
The third module-AZ100-contains
three zener diodes of different break-
down voltage from the OZ100. For
information write HOFFMAN ELEC-
TRONICS CORP., 'Semiconductor
Div., 930 Pitner Ave., Evanston, Ill. standard punched tape widths. Read-
Circle 160 on Reader Service Card. ing speeds range from 100 characters
per second to 750 characters per sec-
ond. The start-stop feature of this
tape transport permits intermittent
reading of tapes at slower ra~es. The
tape converter reading head has been successfully
printing tabulator. The WROC 452, operated at temperatures to 600 C.
as pictured, consists of a flexible type The Kinetape Converter, 768H-1 (left This model uses short strips requiring
control panel containing 32 selectors, below), when used with the TE-206 six inch leaders or loops of tape. Reel
each one of which may, by control Kineplex Data System (right below) feed units are also available. After a
panel wiring, be made to operate in stop command, 1.8 milliseconds is re-
the manner of an IBM type pilot or quired for tape to come to rest. For
co selector. It is also equipped with information write DIGITRONICS
four multitransfer selectors consisting CORP., Albertson Ave., Albertson,
of 10 transfer rows and one common L. 1., N. Y. or use reader service card.
row, each row having 10 positions. Circle 162 on Reader Service Card.
Other features include an alpha nu-
meric emitter, 100 position - two
column digit selector and converter stepping motors
units. For information write MAN-
AGEMENT ASSISTANCE, INC., 40 Translating pulses into incremental
Exchange Place, New York, N. Y. shaft positions, these stepping motors
Circle 159 on Reader Service Card. may be used to rotate control me-
provides high speed transmission of
digital data from magnetic tape over
miniature modules voice quality telephone circuits. Data
rate is 300 characters per second and
Development of a new approach to is adaptable to either IBM or Univac
the design of digital logic called the tapes. Both the 768H-1 and TE-206
"Manalog System" has been an- are completely transistorized and em-
1- ploy etched circuit cards to provide

l high reliability. New signaling tech-

niques allow signal-to-noiseperform-
ance and spectrum utilization. Error
detection and correction is automati-
chanisms, potentiometers, counters,
cally· accomplished by error coding
and rotary switches on computers
and data transmission techniques. For
information write COLLINS RADIO and other electronic devices. With an
angular increment of 36 degrees per
CO., 2700 W. Olive, Burbank, Calif.
Circle 161, on Reader Service Card. pulse it gives 10 indexing positions

Research & Engineering, September/October 1958 23

MINNESOTA MINING AND

•• • WHERE RESEARCH IS THE KEY TO

The term "SCOTCH" is a registered trademark of 3M Company, St. Paul 6, Minn. Export: 99 Park Avenue,New York. Canada: london, Ontario.

24 The Magazine of CRTRE",fJ..HT! ~J

 TRUST
Instrumentation Tapes assure absolute dependability
because only SCOTCH BRAND

inch after inch ... reel after reel

You can't afford to compromise with accuracy, reliability These are precision tapes - engineered in the world's
and uniformity in critical recording work - instrumen- leading tape laboratories to meet your specific needs.
tation, computers, machine tool control and other tech- You can trust "SCOTCH" Brand Instrumentation Tapes
nical applications. You need a magnetic tape of proven because they're made of only flaw-free materials and·
instrumentation quality, "SCOTCH" Brand Magnetic Tape. every reel put to more than 100 rigid quality control tests.

PHYSICAL AND MAGNETIC PROPERTIES OF

"SCOTCH" BRAND MAGNETIC TAPES-INSTRUMENTATION QUALITY

Tape Number
Description
II II II II
108
Std. Instrumentation
109
Std. Instrumentation
128 Hi-Output
Instrumentation
~~~-':'~

159 Extra Play

Instrumentation
Physical Properties
Backing Material Polyester Acetate Polyester Polyester
Thickness in mils
Backing 1.45 1.42 1.45 .92
Coating .55 .55 .65 .35
Ultimate Tensile strength
1/4" Wide-
Room Condition 9# 5.8# 9# 7#
Yield Strength 5%
Stretch in 1/4" Width 5.4# 4.5# 5.4# 3.8#
Elongation at Break 100% 25% 100% 100%
Coefficient of Friction 0.33 0.33 0.30 0.33
Residual Elongation 0.5% 1.5% 0.5% 0.5%
Slitting Tolerances +.000 ins. +0.0% +.000 ins. +.000 ins.
-.004 ins. -0.8% -.004 ins. -.004 ins.
Toughness
Tear -grams 26 3 26 12
Impact - Kc - cms 100 20 100 70
Coefficient of Expansion*
Humidity (units per
% RH change) 1.1 x 10-5 15 X 10-5 1.1 X 10-5 1.1 X 10-5
Temperature (units
per OF.) 2 x 10-5 3 X 10-5 2 X 10-5 2 X 10-5
Temperature Limits for
Safe Use
Low -40°F. -40°F. -40°F. -40°F.
High +140°F. +140°F. +185°F. +140°F.
tRelative Wear Ability 100% 100% 250% 100%
Magnetic Properties
Intrinsic Coercivity (Hci) 250 250 240 240
Oersteds Retentivity (Brs)
Gauss 700 700 1100 1100
Remanence (Flux lines/
1/4" tape) 0.6 0.6 1.2 0.6
Relative Output in db
at 1% distortion**
15 mil Wave Length 0 0 +6 0
Relative Sensitivity
in db**
15 Mil Wave Length 0 0 +3.5 +1.5
1 Mil Wave Length 0 0 0 +3.5
Erasing Field 1000 1000 900 800
Uniformity at 15 Mil Wave Length
Within a Roll ±3% ±3% ±3% ±3%
Roll to Roll ±10% ±10% ±1O% ±1O%
Dropout Count**
Errors/1 Roll
*These coefficients are unitless and represent the change per % RH or degree Fahrenheit over ***Measured by recording 200 non-return pulses per inch on a 0.035" track. A reduction to less
the following ranges: than 50% normal signal amplitude constitutes a signal error. Zero errors are measured by
Humidity: 20% RH to 80% HR saturating the tape undlrectionally. Each spurious signal greater than 10% of normal signal
Temperature: -30°F. t01-"130oF. amplitude constitutes a zero error. Errors per roll based on recording 7 tracks on rolis X"
**At optimum bias for each tape type. x 2500'.
tRelative wear ability Is considered as 100% for 109 Tape. Relative output Is established by 109
which Is designated as zero. Ali other tapes are expressed as gradations from this reference point.

~---------------------------------~
I Minnesota Mining & Mfg. Co., Instrumentation Tape Div. I
900 Bush Avenue, St. Paul 6, Minnesota
FREE BOOKLET! Get all the facts about
I Please send me a free copy of your instrumentation booklet. I
America's most complete line of instru-
I NAME I
mentation quality tapes. Mail this cou- I I
pon for your free specification catalogue. I POSITION I
I COMPANY I
I I
I ADDRESS I
IL _________________________________
CITY 70NE--STATE ~ I
Circle 6 on Reader Service Card.

Research & Engineering, September/October 1958 25

 NEW PRODUCTS

with ± %0 detent accuracy at a maxi- 250 pounds. Servo motors operating

mum stepping rate of 15 per second. at 400 cps provide increased speed. ~
. Motors need all· applicable environ;,:~ . -~
It will'· operate' iuan)' 'position from (
mental tests of MIL-E-5272A. Mount- horizontal to vertical. For information 'i
ing is either servo or stud and models write ELECTRONIC ASSOCIATES, •
are available for wide range of op- INC., Long Branch, New Jer-
erating voltages. For information write sey or use reader service card.
G. H. LELAND, INC., 123 Webster Circle 166 on Reader Service Card.
Street, Dayton 2, Ohio or use card.
Cirel, 163 fJn Reader Service Card.

in the range 2-200 micro amps, the annunciator

control system STI026 features drift of only .05 mil-
li-microamps per degree C and .5 This annunciator has Rashing se-
LIBRATROL - 500 system is devel- milli-microamps per day. This low quence alarm, no drain circuit and
oped around a rapid-response digital drift makes the transistor useful in has been designed for monitoring
computer. The unusual aspect of this circuits with high impedence sources.
Many new low current applications
are opened up by the high beta (typi-
cally 25 at 5 ua to 70 at 100 ua). For
information write TRANSITRON
ELECTRONIC CORPORATION,
Wakefield, Mass. or use reader card.
Circle 165 on Reader Service Card.

x-y plotter
system is the provision for a full range
New design feature of this transistor-
of process control - from accurately
ized variplotter models 205S and T
processing data that provides under-
includes complete transistorization of complex automatic equipment from
standable information to a human op-
24 to 96 points in the utility and con-
erator, to complete automatic control
tinuous process industries. Series 61
of the entire process. Such a building
annunciator is a completely integrated
block concept offers the user a single
unit with all plug-in relays hermeti-
system with the capacity to accom-
cally sealed. No power is used by
modate expanding functions. A grad-
series 61 and all signals are normal.
ual change-over from manual process
Instant operator attention is directed
control to a completely automatic
to off normal conditions by Rashing
computer-controlled plant may be
sequence and audible alarm. For
made without replacing this system.
monitoring more than 96 points, an-
It has been designed for connection
nunciator systems may be connected
to existing equipment in any process-
to operate in parallel.' For informa-
ing plant for a wide variety of indus-
tion write PANNELIT, INC., 7401
tries. For information write LIBRA-
North Hamlin Ave., Skokie, Illi-
SCOPE, INC., 808 Western Avenue,
nois or use reader service card.
Glendale, California or use card. Circle 167 on Reader Service Card.
Circle 164 on Reader Service Card.
all amplifier and power supplies. In
this way vacuum tube failures from
burnout or shock damage is no longer
input transistor a problem, the manufacturer claims. printing machine
Other announced features - instant
A new silicon transistor operates at warmup, greater speed, large 30 in. Providing a printed copy or a printed
low current and d.c. amplifier input by 30 in. plotting surface and high copy and a punched tape, this ma-
stages. Recommended for operation reliability. The .announced weight is chine is equipped with its own con-

26 The Magazine of CRTRM~T! 0 N

 NEW PRODUCTS

tape which has been punched with mentations and data reduction sys-
any type code - 5, 6, 7 or 8-channel. tems features collector rings for each
The· converter operates in conjunction of the three poles, provided in a sep-
with any standard IBM 024 or 026 arate hermetic connector. The switch
card punch, but in no way hinders is comprised of three poles with
normal manual operation of the card thirty BBM contacts per pole operat-
punch. If the card punch is equipped ing at 5 rps. Two of the poles scan
with the IBM self checking number differential thermo couple and strain
device then certain selective indica- gauge output signals while the third
tion data is automatically verified. pole provides the timing function.
trolling circuits and power supply. For information write SYSTEMA T- Special metal brushes are driven
The user needs to provide external Ies' INC., 60 East 42nd Street, New through a suitable gear reduction sys-
contacts for data entry and start, in York 17, N. Y. or use reader card. tem by a 115 volt a.c. 400 cps single
Circle 169 on Reader Service Card. phase 15 watt motor. Noise levels in
order to operate the unit. Dependent
upon the external control circuitry and the order of 20 to 30 microvolts are
functions required, the maximum maintained in this switch through-
speed of complete machine cycling, potentiometer out its 1,000 hour operating life. For
10 digits per cycle, is from 180 per information write INSTRUMENT
minute printing only, to 64 per minute A new sign/ cosign potentiometer for DEVELOPMENT LABORATORIES,
printing, accumulating and punching. use in computer assembly has a stand- INC., 67 Mechanic Street, Attleboro,
The entry of a number of digits ard conformity plus/minus 1% peak- Mass. or use reader service card.
Circle 171 on Reader Service Card.
simultaneously, either printing or
printing and punching in groups of
digits, reduces readout control circuit
complexity and greatly adds to the
reliability of the system by reducing
the number of operations required of mag tape dataplotter
the system control components. For
information write MONROE CAL- This unit is capable of accepting tape
CULATING MACHINE CO., INC., codes written on IBM, Remington
Electronics Components Div., 60 Main Rand, Electrodata, and other digital
St., San Francisco, Calif. or use card.
Circle 168 on Reader Service Card.

to-peak, or a special conformity of

plus/minus ¥z% peak-to-peak. The
tape-to-card converter unit employs oil-impregnated bronze
bearings for a recommended maxi-
Featured in this unit is a removable mum speed of 30 rpm. Precision wind-
plugboard which can be programmed ings, plus a low wear wiper design,
to handle many applications, control produces a guaranteed life of 500,000
cycles, the manufacturer claims. The
unit has a flange mount with a diam-
eter of 2.050 in. and 1.925 in. diam-
eter beyond. Shaft torque is 1.0 oz.
in. For information write CLARO-
STAT MANUFACTURING CO.
INC., Dover, New Hampshire computing equipment. X-Y graphs of
f or use reader service card. data recordered in digital form on
Circle 170 on Reader Service Card. magnetic tape is accomplished at high
speeds. Used with any of the manu-
facturer's plotting boards, data can be
automatic card punching and provide sampling switch presented in points, symbols, or con-
the new converter, model C749, with tinuous lines. Statistical charts of all
flexibility. A second plugboard or cod- A new three-pole sampling switch types and many forms of mechanical
ing matrix equips the unit to convert for use in high speed aircraft instru- drawings are drawn. For applications

Research & Engineering, September/October 1958 27

 NEW PRODUCTS

other than plotting, analog voltages gram. In computers, the RCA-7224 terminals. Holes are provided for con-
are readily available. For information can be employed to translate informa- venient stacking or right angle and
---w-I--:·i..,...te---.;E".........L..."E...,C""'T"""'"'R~O"'""N 'E.....,S.....,--t,....,1-on---rf-ro-m-p-u-n-chr-e-d~c-a-rd-rs-a-n-.dr-p-u-n-c-.hr-e-d..----p-e-rp-e-n-d-rl~c-u..,...Ia-r-m-o-u-n""":"t.....m-g-.-.,....
...........IC,...--rA""7's..,..S....O""""C,. .,. . IA"T"'I"'r1.... I'-ap-er-re-c-e-p--------;o
INC., Long Branch, N. J. or use card. tape. In actual operation a beam of tacles are brass, gold plated over
Circle 172 on Reader Service Card. light passes through the whole of the silver for low contact resistance and
card or tape and activates the tiny freedom from corrosion. A variety of
cell which will trigger the computer molding materials are available in-
memory. The cell employs a germani- cluding Melamine, Diallyl Phthalate,
photoreader
um pn alloy junction and features fast Alkyd and Phenolic Mica. For infor-
rise and fall characteristics. mation write DEJUR-AMSCO COR-
This unit, according to the manufac-
For information write RADIO COR- PORATION, Electronic Sales Div.,
turer; reads 1000 characters per
PORATION OF AMERICA, Electron 45-01 Northern Boulevard, Long Is-
second. It is adaptable to standard
Tube Division, 415 South 5th Street, land City, N. Y. Or use reader card.
Harrison, New Jersey or use card. Circle 176 on Reader Service Card.
Circle 174 on Reader Service Card.

relay
miniature switch
Model 219 relays are designed for
A new mercury switch that weighs
performance on the order of twenty
1.8 grams with leads attached is de-
million operations. They are mechani-
signed for use in computers and other
electronic devices. The extremely low
shift of mass involved in actuation
facilitates gang-mounted assemblies.
The switch, designated AS419A1, may
be mounted in any position through
360 degrees around its longitudinal
axis. It may be actuated by slow, snap
or fast-tilting action. For information
write MICROSWITCH, a division of
Minneapolis - Honeywell Regulator
Co., Freeport, Ill. or use reader card.
Circle 175 on Reader Service Card.

width tape, from five to eight level

terminal blocks
code. Plastic reels are available in two
sizes, for tapes of 350 or 700 feet
These terminal blocks have been de-
(40,000 or 80,000 characters). Fea- cally protected by plastic covers and
signed for various computer applica-
tures include automatic rewind and are designed with plug-in construc-
tions and printed circuitry. They ac-
end-of-tape sensing, true straight-line tion for servicing. Accepted standards
loading and drift-free design. The of insulation include spacings of 1/8 in.
photoreader is available as a com- through air, % in. over surface and
ponent for mounting in any standard a minimum of 1500 volts AC dielec-
19" cabinetry. For information write tric test. Contacts have 10 ampere
ELECTRODATA DIVISION, Bur- current carrying capacity. Plug and .~
roughs Corporation, 450 Sierra-Madre socket combinations are limiting fac-
Villa, Pasadena, Calif. Or use card. tors on ratings. Stock contact arrange-
Circle 173 on Reader Service Card. ments are DPDT on octal plugs and
cept standard "AMP 53" solderless
DPDT plus two normally open con-
taper pins and are available in any
tacts on 12-pin octal style plugs. Op-
combination of feedthru individual or
erating coils may be AC or DC. For
photoiunction cell shorting terminals. External wiring
information write STRUTHERS-
has been eliminated by completely
DUNN, INC., Pitman, New Jersey.
A new light sensitive cell for use in protected, mold-in internal buss con-
or use reader service card.
computers weighs approximately 1 nections between any combination of Circle 177 on Reader Service Card.

28 The Magazine of D RTF:U\,J2&=1Ti CJ N

 people moving up in :c 1=1 T R

Clair C. Lasher is the new general manager of General Samuel Ochlis is the new sales manager of Instrument and
Electric's Computer Department in Phoenix. Joining the Equipment Division, Epsco, Inc., Boston: The division de-
company in 1939, he became manager of marketing at the signs, manufactures, building blocks in large data handling
inception of the department in 1956. He supersedes H. R. systems. Wallace E. Rianda is vice-president and general
Oldfield, Jr., who has been appointed general manager of manager of Epsco-West, newly established· west coast di-
a new. component division, unnamed as yet ... Philco has vision. He was formerly marketing manager of Beckman
announced promotion of Sol Zechter to manager of Transi- Instruments, Inc. William F. Gunning, technical director,
storized Devices Laboratory in Philadelphia. Since joining Ralph McCurdy, in charge of production-were also as-
the company he has handled development of equipment sociated with Beckman.
under government contracts in transistorized communica-
tions, telemetering-holds two patents on the latter, five Taft B. Russell is now manager of systems sales and re-
pending on electronic devices ... Jack Cudahy will head search and development contracts of General Devices,
the new west coast office of Technitrol Engineering Co. Inc., Princeton, N. J . . . . . Consolidated Electrodynamics
Philadelphia. This office will provide manufacturers with Corp. has established an International Department within
technical assistance on problems of complex design. the Marketing Division and appointed Rodney W. Meyer,
as director.
Newly appointed member of Control Data Corporation's
mechanical engineering group, Dean M. Roush, will have Computer Services Division of ,the Corporation for Eco-
specific responsibilities associated with design of Min- nomic and Industrial Research has announced appointment
neapolis company's new CDC 1604 scientific computer. of Robert L. Patrick deputy director. . .'. appointed man-
The initial order on a $600,000 Navy contract was an- ager of Sylvania's newly formed Needham, Mass., data
nounced recently. Roush was formerly mechanical engi- processing facility is Frank M. Thomas. Richard R. Fidler
neer with Remington Rand Univac's Military Division. will head advanced development-data conversion depart-
Other Control Data appointments: to computer engineer- ment. ... John B. Olson is named chief engineer of Com-
ing staff, James D. Harris as senior administrative as- puter Measurements Corp . . . . Stromberg-Carlson has
sistant, Carl E. Koehler as systems logical designer, appointed William G. Alexander assistant general man-
James E. Thornton as senior electrical engineer. Also, Dr . ager.
Robert E. Smith named senior mathmatician on profes-
sional staff. Homer M. Sarasohn is named director of engineering
planning on corporate staff of IBM. He will exercise staff
Joseph A. Resca is now the New York district sales man- supervision of all product development, engineering ac-
ager for Burroughs' ElectroData Division. He will be in tivities and provide liaison among IBM's operating .di-
charge of sales and service of company's E101, 205 and visions. Dr. Morton M. Astrahan is appointed functional
220 computers, and EDP systems in that area. Lately man- manager, responsible for research on data processing needs
ager of the division's Dallas district he was formerly with of small business. Richard W. Porter is appointed program
Telecomputing Corporation. ElectroData has also es- manager for a large government contract.
tablished two new district sales offices. Claggett A. Jones
will head the Atlanta, Ga., office and Charles V. Hoge is
in charge of Denver, Colo., branch ... William E. Brug-
man has been appointed to the newly created position of
components sales manager at Telemeter Magnetics, Inc.

Expansion of DA T Amatic, data-processing division of Min-

neapolis-Honeywell Regulator Co., resulted in the opening
of a Washington sales office. Robert F. Anderson has been
appointed manager, Fillmore Dobbs is assistant. . . .
f'
Gerard Q. Decker was elected vice president of Servo-
mechanisms, Inc., by board of directors. He was promoted
from division manager of the company's Subsystems Di-
vision . . . Computer Engineering Associates, Inc., ap-
pointed Dr. Richard H. MacNeal as manager of their
Engineering Service Division .... Philip Balaban has been CLAIR C. LASHER DEAN M. ROUSH JOSEPH A. R
named director of research for Mid-Century Instrumatic GE's Control Burrough's
Computer Data ElectroData
Corp., in New York. Department Corporation Division

Research & Engineering, Septer.nber/October 1958 29

 OF ALL COMPLETE COMPUTERS,
ONLY THE COMPACT, POWERFUL
ROYAL PRECISION

GIVES YOU THE SPEED, CAPACITY

AND MEMORY YOU WANT
AT LOWEST DOLLAR COST! •
Desk-sized, mobile. Plugs into ordinary
wall outlet. Self-cooled. Installed
without charge • Lowest cost ever for a
co m pi ete co m pute r syste m • Su b-
routines and programs available
• Customer train.ing • Large staff of ap-
plications analysts. Sales and service
Remarkably simplified command structure and
facilities available coast-to-coast. controls require only minimum computer experience.
Front panel oscilloscope for concise display.

High-speed photoelectric reader-with or without Full flexibility of stored-

mechanical punch-available for system expansion. program operation
Loads entire drum in approximately 5 minutes. with unusually large
non-volatile memory
of 4096 words.

Complete format control.

Alpha-numeric input-out-
put via keyboard or
punched paper tape. No
additional equipment
needed to operate.

For further information and speciftcations,

write Royal MCBee Corporation, Data
Processing Division, Port Chester, N. Y.

ROYAL
data processing division

Circle 7 on Reader Service Card.

30 T he Magazine of :c Fl T Fll\ll.t=i T i 0 i'-3

 ELEMENTARY PRINCIPLE OF

AND ITS APPLICATION TO DIGITAL COMPUTERS

by SABURO MUROGA amplitude. In the terminology of radio engineers, a small

Senior Researcher, Electrical Communication Laboratory signal which takes one of two possible phases has been
Nippon Telegraph and Telephone Public Corporation amplified into one with much larger amplitude which re-
tains the original phase. And it is interesting to note that
The idea of the parametron was born in Tokyo and is so these two possible· phases are different from each other
named because its working principle makes use of "a by pi radian; that is, the voltage of one oscillation is equal
parametrically excited oscillation." Its principle is quite to the minus of the other.
common in our daily lives. Examples range from a swing Parametron is the very realization of this by electronic
and a yoyo which children like, to a parasitic oscillation means, that is, simply a pair of magnetic cores, a resistor,
of a speaker cone which is an engineer's problem. and a condenser, as shown in Fig. 2. A resonant circuit
Take the first example in Fig. 1 (page 32). As a child with a variable inductance is constructed with these com-
moves his body up and down repeatedly, the swing starts ponents. Then an alternating current with a' frequency 2f
to oscillate. Suppose in this case that the vertical change and a direct current superposed on that are applied on
of the gravity center of his body has a frequency "2f". the primary windings of the inductance. But nevertheless
Then the movement of the swing has a frequency "f". An the secondary windings are made to cancel an induced
important fact in this case is the· relation between an ini- output to the· primary input, and an alternating current
tial state of the swing and a phase of its final oscillation. with a frequency f starts to oscillate as the swing did
The final oscillation of the swing can take one of two from a vertical movement of the human body. This in-
possible phases, even though their amplitudes and fre- duced oscillation can take one of the binary phases, and
quencies are the same, and which one of them is to be if a signal with a small amplitude and a frequency f is
taken actually depends on which side of the rest point given initially in the circuit, it will be amplified. These
the swing was placed initially, as shown in Fig. l. two are essential points of parametrons. Here the resonant
In principle, however small this initial displacement of frequency of the circuit, in other words, the proper fre-
the swing, left or right of the point of perfect rest, it quency of the swing, should be nearly equal to half of
. determines a phase of the final oscillation with a large the excitation frequency 2f.
The idea of applying this principle to switching cir-
SABURO MUROGA was cuits was discovered by E. Goto of Professor H. Taka-
born in Shizuoko-ken, south hashi's laboratory at Tokyo University in 1954. He made
of, Tokyo, in 1925. In 1947 its application possible by devising a three-beat excita-
he _ graduated from Tokyo tion and a majority· decision logical operation.
University with a (legree in Any number of parametrons can be assembled to rea-
Electrical Engineering, and lize a switching circuit with application of three-beat ex-
last year received a Ph.D. citation as- follows: Connect parametrons serially, bridging
for his thesis, "Information their terminals with resistances and transformers as shown
Theory." Before coming to in Fig. 3, and apply the excitation currents on their pri-
Nippon Telegraph and T ele- mary windings sequentially from the left~ Then the os-
.. phone Laboratory in 1950, cillation of the parametron PI will be coupled into the
he ,was with the laboratory second parametron P2 as a~ initial small signal input to
of the National Railways find it, so that a phase of the former will be conveyed to the
then the Radio Regulatory second. Similarly, P2 will supply the initial signal to the
Commission. In 1953 he was a participant in the Foreign third parametron P3. Repeating this process over any
Student Summer Project at Massachusetts Institute of number of panimetrons, a phase. of the first parametron
Technology. He stayed on at M.l.T. for another six will be correctly conveyed to the last parametron. This
months, and then went to the University of Illinois for a is a kind of shift register which is important in switching
half year. He holds ten patents and has fifteen papers pub- circuits. For its physical realization, it is economically
lished on information theory and digital computers, in- impossible to prepare a source of excitation current for
cludin material in the IRE Proceedin s. each parametron. The whole set of parametrons is divided

Research & Engineering, September/October 1958 31

ELEMENTARY PRINCIPLE OF PARAMETRON

A
Movement of
a human body

f t ~ \~\'~~t!
99 Q ~ Q,. q q 5> $>

B
Oscillation

If with a phase
of 0 radian

Figure 2. Parametron
C
Oscillation
with a phase
of " radian

_time

Figure 1. Parametron principle in a swing.

into three groups so that every adjacent three parametrons parametrons are coupled to inputs of the next parametrons
belong to the groups I, II, and III, respectively. with equal resistances. Here in Fig. 5 where a circle in-
For each of these groups, only three sources feed cur- dicates a parametron, all the amplitudes of oscillated vol-
rents il, ill, and illl, the tails of whose envelopes overlap tages .in three parametrons are assumed to be equal. Then
each other, respectively, as shown in Fig. 4. For exam- a phase of the parametron which is coupled from these
ple, the tail of il overlaps the head of ill, but no part of th-:ee parametrons is obviously determined from an alge-
illl. During this overlap, phases of the oscillations in a set braic sum of input voltages from these parametrons of
of parametrons belonging to group I are conveyed to those the majority decision.
of group II, but not to those of group III. Thus, with N ow assume the phase of the first parametron to keep
three-beat excitation the transmission of binary signals in o radian and the other two to be variables. If the first
a single direction could be realized in a whole network variable takes a phase of pi radian and the second 0 ra-
of parametrons. dian, the two voltages of the constant and the first varia-
maiority decision principle ble cancel, leaving simply the voltage of the second
Another important principle is the application of ma- variable with _a phase of 0 radian. Thus the coupled para-
jority decision to logical operations of parametrons, which metron _will oscillate a voltage of 0 radian for this input.
are indispensable for switching circuits. As widely kn~wn Similarly, if the first and second both take 0 radian, a
from Boolean Algebra, the logical operations "and", "or", voltage of 0 radian with triple amplitude will become
and "not" are sufficient for switching circuits to perform the input to the coupled parametron. The output phases
any digital functions. For this purpose outputs of all the for other combinations of the variables may be seen from

PI P2 P4

~II
i311~~ III _
-Group I II IV

Figure 3. A shilt register with parametrons

32 The Afagazine oj:C RTRi\i:i.~T! w f>J

 11------1

Table I which shows this circuit of parametrons to be

"and" circuit, if the phases df o and pi, I<!dians are re-
garded as binary digits "0" and "1", respectively. Also it
will be clearly understood that a circuit· of parametrons
• Figure 4. Excitation currents in three beats
will work as "or" if the first parametron takes a constant
phase pi radian and the other two are regarded as varia-
bles. See Table II. The remaining "not" can be. realized
quite easily by reversing the polarity of the coupling trans-
former between the parametrons.
It is interesting to see that the majority decision prin-
ciple of logical operations of parametrons is quite similar
to that of neurons in living organisms. In general, logical constant
elements standing on the majority decision principle which
have been studied by the author have some interesting,
unique properties as performers of thc Boolean Algebraic
operations and will be discussed elsewhere.
Fig. 6 shows some exjmples which might be encount- first
ered in switching circuits like digital computers, where a variable
circle indicates a parametron, a coupling line with a cross-
ing "not" and + or - inside a circle the constant 1 or O.
In these circuits the number of inputs into a parametron
is limited to a maximum of three, but could be increased
to five, seven, or more. second
The original type of parametron uses a pair of ferrite variable
cores with an outer diameter of about 4 mm. and a single
turn primary winding for excitation and a ten turn sec- Figure 5. Majority decision principle
ondary winding for the resonant circuit. Its resonant fre-
quency was chosen at 12 mc. The condenser and resistor
are of normal type. As easily seen from this structure, low
cost and almost limitless life are expected. This type of
parametron was used for our computer "M-l". Parametron
consumes less than 80 mw on average. This power con-
(a) binary adder
sumption characteristic has been much improved by efforts
of Z. Kiyasuand others in our laboratory which backed x

y Sumz

Table 1- "and" circuit for constant with 0 radian

w
1st variable 2nd variable Output
0 0 0
0 pi 0
(b) binary counter
pi 0 0
pi pi pi

Table II - "or" circuit for constant with pi radian

1st variable 2nd variable Output
0 0 0
o ¢ ¢
¢ 0 ¢
pi pi pi

Fi,ure 6. Example& 01 the parametron circuits

Research & Engineering, September/October 1958 3

ELEMENTARY PRINCIPLE OF PARAMETRON

the development of an electronic telephone exchange and

construction of a digital computer with parametrons. It
was improved by three to five times.

musasino 1

Our large scale digita1 computer which has been con-

structed by the laboratory people including the author
primarily for scientific problems was named "Musasino 1"
after its location. It was completed in March 1957 with
a small magnetic core· memory capacity and had been
used to calculate tables of elementary functions. In March
of this year its memory was enlarged to 256 words, mak-
ing its debut in public as ready for general use in our
laboratory. This is .the first large scale digital computer
with parametrons. It is shown in Fig. 7.
The panels on the right are the source of excitation
currents of 2.4 mc. The sixteen square panels in the center
contain the parametrons, six upper panels being the con-
trol unit and the lower ten the arithmetic unit. Above Figure 7. The M-l
these parametron panels are neon indicators for the ac-
cumulator, a quotient register, an order pair register, and
various alarms, while manual control switches are to the wires of excitation. Ferrite cores with non-rectangular hy-
left. At the rear of the room is the electrical power source. steresis curve are satisfactory for this memory principle.
The M-1 is a program-stored binary computer with A paper tape with six holes for a Japahese standard
fixed point in a parallel system. The word structure and teletype is used for input and output. One of these holes
instructions are modeled after the Illiac at the University is for odd parity check and when the tape is read by a
of Illinois. Consequently, a single word represents forty photoelectric reader, the computer stops for failures of
binary digits as a number or a pair of orders. About 130 the check. A teletype punch is used for the output.
different instructions are available, including ones for A number of features for speedup and easy use are
handling magnetic tapes, for performing three different included in the machine's design. A carry detector an:d a
Boolean .Algebraic operations, for controlling brightness high speed multiplier, for example, are incorpo~ated:· The
of the beam spot of a' cathode ray tube display and others. former is to sense an end of carries in addition' or s~b-
The M-1 uses 5356 parametrons, of which 2800 are used . traction and consequently average times for addition arid
for the arithmetic unit, 1600 for the control unit and 1456 division are greatly reduced. Many alarms for incorrect
for the magnetic memory, and 519 vacuum tubes for the orders, misoperations of the input, changes of source
source of high frequency excitation currents, high' power .voltages and others are provided.
drivers, the neon indicators and others. The M-1 consumes The life ofa parametron is almost infinite because of its
a stabilized -dc power of 5· kc for the primary ac 9 kva. simple structure. There has been no replacement of para-
The magnetic core memory, where a number of ferrite metrons during the 4000 hour run of the M -1. The only
cores of 2 mm diameter are used, has a unique feature. troubles have been bad soldering connections, but once
Its· working principle, which was also Goto's idea, depends these are fixed, that is the end of the trouble. The cost of
entirely on alternating currents. A current of a frequency the magnetic core memory is low because the required
£/2 with a constant phase for selecting a desired word characteristics of the core are not rigid. The M-1 has
among the whole memory is superposed on information worked very stably for these four months without any
currents of a frequency f, each of whose phases is speci- failures and with almost no maintenance. Every day, ten
fied by each of the forty binary digits of a. number to be minutes after electric:ity. is applied, it is available for
stored and then it is flowed into the magnetic cores of the computations.
desired word, writing the number on these. For reading Only relatively low speed of operation· might be con-
out a stored number from the magnetic cores, a current of sidered a defect of parametrons, but the M -1' s speed is
the frequency £/2 with a constant phase' flows into the almost comparable with vacuum tube computers in a
desired cores from a selection matrix to induce second serial system. Its speed is linearly proportional to the
harmonic currents, each of whose phases depends on a repetition frequency and orders for addition _and multi-
polarity of the residual magnetism of the core. The selec- plication take 2 and 10 milliseconds on average, respec-
tion matrix, too,consists of parametrons, bf which only tively, for 10 kc repetition which may be at least doubled
one is placed at a crosspoint of the vertical and horizontal by further adjustment in the near future.

34
 Important dates iri; DRTR I

, Oct. 9-1 1: Information Storage .. and Retrieval Systems Nov. 16-21: International Conference on Scientific In-
Conference, Graduate School of Library Science, U niyer- formation, Mayflower Hotel, Washington, D. C. Spon-
sity of Texas. C(}ntact Dr: R. H. Douglass, Graduate School sored by NAS, NRC, NSF and AD!. Contact Secretariat,
. of Library Science, University of Texas, Austin, Texas .. International Conference on Scientific Information, Na-
• Oct. 13-15: International Systems Meeting, Systems and tional Academy of Sciences, 2101 Constitution Avenue,
Procedures Association, Hotel Penn-Sheraton, Pittsburgh, 'N. W., Washington 25, D. C. (See page 13)
Penna. Contact A. M. Motter, Jones and Laughlin. Steel Nov. 17-18: Federal Govt. Accountants Association's 8th
Corp., #3 Gateway Center, Pittsburgh 30, Penna: Annual Symposium. Theme: "Management and Electronic
Oct. 16-18: The Institute of Management Sciences An- Data Processing." Contact Martin C. Powers, 1523 L St.,
nual Meeting, Philadelphia, Pennsylvania. N. W., Washington 5, D. C.
Oct. 20-21: Remington Rand Univac Users Conference~ Nov. 17-20: Fourth Annual Conference on Magnetism
John Hancock Mutual Life Insurance Company, Boston, and Magnetic Materials, Sheraton· Hotel, Philadelphia,
Mass. Contact R. M. Petersen, Secretary, Univac Users Penna. Sponsored by AlEE. Contact John Leslie Whitlock
Conference, General Electric Company, Appliance. Park" Associates, Exhibition Managers, 6044 Ninth St., North,
AP 1-109, Louisville, Kentucky. Arlington 5, Virginia.
Oct. 20-24: National Business Show, Coliseum, N. Y. C. Nov. 19-20: Northeast Electronics Research and Engi-
Contact Rudolph Lang, Managing Director, 530 5th Ave.,: ~eering Meeting, Mechanics 'Hall, Boston, Mass. Sponsored
New York ·36, N. Y. by IRE. Contact J. J. Faran, General Radio Company, 22
Oct. 22-25: The National Businessmen's Exposition, Baker Avenue, . West Concord, Mass.
Great Western Exhibit Center, Los Angeles, Calif. Spon- Nov. 20-21: Conference on Electronic Computation,
sored by NMA. Contact Robert W. Caldwell, NMA Show Kansas City, Missouri. Sponsored by the Kansas City
Chairman, National Businessmen's Exposition, 2807 Sunset Section 'and the Committee on ElectroniC Comput~tion
Boulevard, Los Angeles 26, Calif. of the Structural Division, ASCE. Contact'Secretary, Ste-
Oct. 23-24: Operations Research Society of America ven J. Fenves, 203 Civil Engineering Hall, University of
National Meeting, Statler Hotel, St. Louis, Missouri. Illinois, Urbana, Illinois.
OCt.23-25: The National Society of Professional Engi- Nov. 20-21: American Mathematical Society Meetings;
neers - fall meeting, St. Francis Hotel, San Francisco, Pomona, Calif.; and Nov. 28~29: Northwestern Univer-
Calif. Contact Kenneth E. Trombley, National Society of sity, Evanston, Illinois; and Durham, North Carolina.
Professional Engineers, 2029 K St., N. W., Washington
Nov. 28-Dec. 4: National Physical Laboratory Sympo-
6, D. C.·
sium and Electronic Computer Exhibition, London, Eng-
Oct.23-25: 1958 National Simulation Conference, Stat- land. Contact C. V. Wattenbach, Deputy Managing Di-
ler-Hilton Hotel, Dallas, Texas. Sponsored by IRE-PGEC. rector, Dictograph Telephones, Ltd., London, England.
Contact J. E. Howard, 2100 Menefee Dr~, Arlington, Tex.
Dec. 3-5: Eastern Joint Computer Conference, Bellevue-
Oct. 25: American Mathematical Society Meeting, Prince-
. Stratford Hotel, Philadelphia, Penna. Contact John M.
ton University, Princeton, New Jersey.
Broomal, Burroughs Corp., Paoli, Pa. (publicity informa-
Oct. 27-28: Fifth Annual East Coast Conference on tion) or Dr. F. M. Verzuh, MIT Computation Center,
Aeronautical and Navigational Electronics, Lord Balti- Cambridge 39, Mass. (program information). (See p. 40.)
more Hotel, Baltimore. Sponsored by IRE. Contact Harry
Rutstein, Publicity Chairman, Lord Baltimore Hotel, Balti- Dec. 9-10: Mid-America Electronics Convention, Muni-
more, Maryland. cipal Auditorium, Kansas City, Missouri. Contact Wilbert
O'Neal, The Vendo Co., 7400 E. 12th, Kansas City. Mo.
Oct. 29-30:. Fifth Annual Computer Applications Sym-
posium, Morrison Hotel, Chicago. Sponsored by ·the 'Ar- Jan. 20-22, 1959: American Mathematical Society -
mour Research Foundation, Illinois Institute of Technol- 65th Annual Meeting, U. of Penn., Philadelphia, Pa. '
ogy. Contact the Foundation at 35 W. 33rd St., Tech- Feb. 12-13: Transistor and Solid State Circuits Con-
nology Center, Chicago 16, Illinois. (See page 41.) ference, University ,~f Pennsylvania, Philadephia, Pa. Spon-
Oct. 30-31: Fourth Electronic Business SystemsC~~­ so·red by the PGCT, the AlEE and the University of Penn-
ference, Olympic Hotel, Seattle. Sponsored by the ~estern sylvania. Contact Arthur B. S·tern, General Electric Co.,
division of the NMAA. Contact E. B. S. Conference, Building 3, Syracuse, N. Y. . .
NMAA, . P. O. Box 1~4, Seattle 11, Washington. . Mqr. 2-6: Western Joint Comput~r Conference, Fair-
Nov. 3-7: Fifth.)nstitute on Electronics in Management, mont Hotel, San Francisco, ,Calif. Sponsored by PGEC;
The American University, Washington, D. C. Contact AlEE; and ACM. Contact M. L. Lesser, IBM Research
Lowell H. Hattery, Fifth Institute on Electronics in Man- Laboratory, San Jose, Calif.
agement, The American University, 1901 F Street, ·N~\W., June 15-20: International Conference on Information
Was~ington .0; ~D. C. Processing, Paris,. France.

Research & Engineering, September/October 1958 35

 •

BORG
and the field of
Engineer L. Maiboroda at the controls of SESM, Russian calculator.
ELECTRONICS
COM'PUTING IN THE USSR Borg is well-known and highly
respected for its sound,
trial work started on sesm creative engineering. The pre-
cision qualities of Borg com-
Trial work on SESM, a Russian electronic calculator, has been initiated ponents for systems are widely
at Kiev, in the calculating center of the Ukrainian Academy of Sciences. recognized in both the com-
This specialized computing machine is capable of solving linear algebraic mercial and military fields.
equations having up to 400 unknown factors and is the first of its kind to be
produced in the USSR or anywhere in Europe, according to Soviet scien- • AIRCRAFT INSTRUMENTS
tists. It is being used to .reckon complex hydro-technical building and ma- Aircraft components, instruments
and electronic sub-assemblies.
chine building designs and to solve the problems in geodesy and mathe-
matical physics. • FREQUENCY STANDARDS
Extensive use of both vacuum tubes and solid state devices are featured Crystal controlled oscillator type
in SESM's construction. The machine occupies 86 sq. ft. frequency standards.
Systems engineering was accomplished under the supervision of S. A. • MICROPOTS
Lebedev and Z. L. Rabinovich at the Ukrainian Academy of Sciences. Precision potentiometers in a wide
range of single-turn, multi-turn
and trimming models.
W JCC COMMITTEEMEN NAMED
• MICRODIALS
Committee heads have been named for the 1959 Western Joint Compu- Precision MICRODIALS for
single and multi-turn devices.
ter Conference to be held in San Francisco, March 3-5, 1959. Indexed accuracy of up to one part
Joint sponsors are the Institute of Radio Engineers, the American Insti- in 1,000.
tute of Electrical Engineers and the Association for Computing Machinery.
Headquarters and meeting place will be the Fairmont Hotel. • INSTRUMENT MOTORS
Robert R. Johnson of the General Electric Computer Laboratory, Palo Precision motors, synchronous and
induction types. Gear trains,
Alto, Calif., is general chairman and has announced the composition of the
steering committee for the conference, all Californians, as follows: LET BORG HELP YOU
Richard W. Melville of Stanford Research Institute, Menlo Park, vice- Borg can assist you in the design
chairman and chairman of the technical program; Charles Asmus of General and construction of prototypes,
Electric Computer Laboratory, Palo Alto, conference secretary-treasurer; Complete facilities for pilot runs
Byron J. Bennett of IBM Product Development Laboratories, San Jose,
and quantity production. Write for
Catalog BED-ASO or call us today.
..
publications; George A. Barnard, III of Ampex Corporation, Redwood City,
publicity; Harry K. Farrar of Pacific Telephone & Telegraph Co., San
Francisco, exhibits. MICROPOTS
Also, Kenneth F. Tiede of Univ~rsity of California Radiation Laboratory, MICRODlAlS
Livermore, field trips; Robert M, Bennett, Jr., of IBM Research Laboratory, MOTORS
San Jose, registration; L. D Krider of University of California Radiation
Laboratory, Livermore, printing; Mrs. Joanne Teasdale of General Electric
Computer Laboratory, Palo Alto, women's activities; Earl T. Lincoln of BORG EQUIPMENT DIVISION
Stanford Research Institute, Menlo Park, mailing; and Robert C. Douthitt The George W. Borg Corporation
120 South Main Street, Janesville, Wis.
of Remington Rand, EI Cerrito, local arrangements.
Cirrii' 8 on Reader Serf1ice Card,.

36 The Magazine of CRTRlvkh··TI

•

THERE'S A
BORG MICROPOT®
TO MEET YOUR EXACT
SPECIFICATIONS!

BORG MICROPOTS •••

- ,,;

the Ultimate in Multi-Turn Precision Potentiometers

Borg offers a complete line of high-precision, linear potentiometers
called MICROPOTS. Precision is Borg's business . . . that's why
MICROPOTS offer so many exclusive advantages. Superior design and
production methods ma~e Borg MICRO POTS available in any quantity~
Write today for the name of your nearest Borg Jobber or "Tech-Rep."

..
A precision MICROPOT that offers
your products a price advantage in
today's competitive markets. Lug or
lead type terminals. Accurate
dependable ••. long lived.
900 Series
MICROPOTS 205 Series
MICROPOTS
Standard tea-turn and three-turn' A quality MICROPOT. Designed for
models to fit most special design needs. both military and commercial appli-
, Extremely accurate and dependable
under adverse environmental con-
cations. Proven in many different
mobile and stationary types of elec-
ditions including severe vibration tronic circuitry.
Small in size, lightweight, rugged and and shock.
dependable. Three types of terminals
•.. printed circuit, solder lugs or in-
sulated wire leads. WRITE FOR COMPLETE ENGINEERING DATA. CATALOG BED-A90

Circte 9 fm Reader Service Card.

Research. & Engineering, September/October 1958 37

IBM ANNOUNCES TRANSISTORIZED 7070
Details have been released on IBM's newest computer, the Company officials pointed out that the 7070 will also
7070, billed as "a completely transistorized computer in be able to compile Fortran language, the same as used by
the intermediate price field." IBM 650 and 700-series systems. The term Fortran stands
By using transistors (manufactured by Texas Instru- for formula translation and is a means of expressing mathe-
ments, Inc.) one third is cut from installation and environ- matical processes in a manner that the machine can read
mental control costs, IBM claims. Another claim: computer and translate into a group of program steps. The pro- •
time is saved by overlapping machine operations. By using grammer writes the formula in a manner quite similar to
two tape channels, the 7070, while computing, reads or normal mathematical notation, and the machine creates its
writes magnetic tapes at the rate of up to 125,000 charac- own program to carry olit the operation. For example, the
ters per second. IBM says the system can handle simultane- formula (a+b)~ c/d means: add a to b, multiply the sum
ously "several" 400-card-a-minute readers and 250-card-a- by c, and divide this product by d.
minute punches. A tape sorting and merging routine will be available
In order to permit in-line data processing, four files of for the 7070. In addition, there will be utility routines for
50 magnetic disks can be incl~ded in the system, providing clearing core storage between limits, moving disk storage
for: the random access storage of 24 million digits. Also to output, load routines, tracing routines, and so on.
available-an immediate access memory of 50,000 to 100,- Another feature of the 7070 is Automatic Priority Pro~
000 digits held magnetically in miniature ferrite cores. cessing which makes it possible to combine two programs
and virtually eliminate any lost" time waiting for an opera-
As an installation aid, IBM has developed a series of
tion to be completed. Often programs are said to be "in-
general-purpose, automatic programs for the 7070.
put-bound," "seek-bound," etc., referring to parts of the
One of the assembly routines available for the 7070 is application that other phases must wait for at some point
similar to the Autocoder program, which serves as a sim- or points in the program. With automatic priority process-
plifield system of program writing for the 705. ing, there is no delay.
Although the language used is not Autocoder language, One of the programs, called the main routine, has a
it resembles it somewhat. Operation codes are in mnemonic comparatively large number of program steps. The other,
notation (a memory-aiding device) and the programmer called the priority routine, has relatively few instructions
can refer to the field to be processed in a manner that has but involves almost continuous use of a card reader, card
a definite meaning to him such as "Grosspay," "FICA" and punch, printer, tape unit, or disk file.
other familiar terms. Comments such as "Compute Exemp- The main routine functions normally, while the tape,
tion Amount" on the first instruction of a, sub-routine can -disk-storage, or input/output unit of the priority routine is
be written to assist the programmer in keeping track of operating. This may include reading a card, punching a
blocks of instructions. Not only is the task of writing the card, reading tape, writing tape, seeking a disk file-record,
program made easier, but additions, deletions, and correc- reading a file record, or writing a file record. When that
tions can be made without changing addresses written. operation is completed, the main routine is signalled auto-
Instructions are punched into cards and fed into the matically. It is possible to have more than one tape, disk
7070. The program converts· each instruction into a ma- storage, or input/output unit operating on a priority basis
chine-language program step, assigning locations for the during the main routine. However, only the main routine
instructions and for the data used in the program. An out- can be signalled for. priority; it is not possible to do this
put card is punched for each instruction, and these cards to a priority routine. If a second priority is ready while a
comprise the program deck. Assembling of the program first one is in progress, it will wait until the first is com-
and the record-keeping phase of writing the program are pleted. The main routine is resumed when there are no
performed by the 7070 itself. priority routines waiting.
Circle113 on Reader Service Card.
 DRTR book capsules

CENTRALIZED INFORMATION SERVICES:-OPPOR- AUTOMATION EXPRESS, International Physical Index,

TUNITIES AND PROBLEMS by Allen Kent and James Inc., 1909 Park Avenue, New York 35, N. Y., $57.50 for
W. Perry, 1958, Interscience Publishers, Inc., 250 Fifth one year.
. Avenue, New York, N. Y., 156 pp., $5.00.
This book presents an introductory analysis and sum-
Volume one, number one, published in May of this
year, is the initial 40-page offering of a comprehen-
mary of studies directed to the investigation of advantages sive digest of current Russian literature dealing with auto-
and feasibility of centralized, cooperative information mation topics. Sifting timely articles from the current issues
services. of many Russian technical journals each month, the pub-
The results of eight studies and surveys carried out lishers of the new digest present such articles as Digital
over the past three years by the Center for Documenta- and Analog Computer Systems, Servomechanisms and
tion and Communicatio'n Research at Western Reserve Components, and Magnetic Amplifiers and Circuits, with
University have been analyzed; systematized and sum- liberal use of figures and diagrams, accurately translated
marized. The purpose of the authors is to clarify the goals from the Russian into easy flowing English.
and indicate the practibility of future developments. Some
contents of this study: Questionnaire on Special Classifica-
tions and Information Systems; Information Processing by
Professional Societies; High Speed Telecommunication in AUTOMATIC DATA REDUCTION SYSTEM-AMPLI-
Centralized Information Services. TUDE-DISTRIBUTION AND CORRELATION ANAL-
YSES by A. Shapiro, Naval Research Laboratory, Dec.
1957, OTS, U. S. Department of Commerce, Washington
SCIENTIFIC PROGRAMMING IN BUSINESS AND 25, D.C., 11 pp., .50 cents.
INDUSTRY by Andrew Vazsonyi, 1958, John Wiley & Described and discussed are many forms of radio and
Sons, Inc., 440 Fourth Avenue, New York 16, N. Y., 474 radar data and a data reduction system which performs
pp., $13.50. these analyses automatically, using data recorded on film
An attempt to develop a mathematical language under- as input. A film reader converts data to digital voltages
standable to businessmen using scientific techniques to which are totaled in a 30-level amplitude distribution.
solve managerial problems in terms of business, rather
than in terms of mathematics, is presented. These tech- GRIN AND BEAR .IT
niques are clarified and mathematical programming and
its applications are defined.
The book contains descriptions of the use of linear pro-
gramming in transportation allocation; the use of statisti-
cal methods in production and inventory control. This
approach has been applied with successful results, claims
author Vazsonyi, in programs discussed with business
people· and case histories have been included.

NONDESTRUCTIVE READOUT OF MULTILEVEL

MAGNETIC MEMORY by R. L. Van Allen and C . B.
House, 1958, Office of Technical Services, U. S. Depart-
ment of Commerce, Washington 25, D. C., OrderPB
131475, 26 pp., .75 cents.
, This is a report on a: new method developed for nonde-
structive readout of memory cores as a result of Armed
Forces-sponsored· research and development.in electronics.
The method, an infinite-resolution of reading the flux level
in' a· ~agnetic core without destroying this flux level, uses
solid-s~ate devices; requires less than ten milliwatts supply
'duririg nondestructive interrogation, while standby power
drain:is in th'e micro watt range.
. Output information is in the form of an alternating
waveform whose ·frequency is a function of the flux level
of the storage core. Frequency ratios of 30: 1 were ob-
tained. Also developed - a circuit for clearing and reset- "This eliminates guess work in medicine, doctor! Feed
it symptoms and it diagnoses the case, writes a pre-
ting a core in preparation for further information storage. scription and indicates the size bill the patient can pay!"

Research & Engineering, September/October 1958 39

 EJCC SITE-
PHILADELPHIA
Over 3,000 delegates are expected to
attend the .Eighth Annual Eastern
J oint Computer .Conference which
will be held' on December 3, 4 and 5
in Philadelphia-at the Bellevue Strat-
ford Hotel.

Included in the conference pro-

gram are technical sessions, exhibits,
a party - reception and a luncheon.

"Modern Computers - Objectives,

Designs and Applications" is the
theme of this year's EJCC which is
sponsored by the Institute of Radio
Engineers, the American Institute of
. . Electrical Engineers and the Associa-
tion for Computing Machinery. Con-
"TRANSISTORIZED DIGITAL ference Chairman is Peter E. Raffa
of Technitrol Engr. Co., Philadelphia.
'M~GNETIC TAPE' HANDLER "This industry is in a transition
period-going from huge expenditures
rv1~~~~ ~~~ for research and development to a
marketing program which will induce
commercial sales and applications," an
EJCC spokesman noted, adding, "this
Optimum performance trend will be reflected both in the ses-
in virtually all tape handling applications sions and the exhibits."

The advanced design of the completely transistorized Potter Model 906 Tape Technical papers, dealing with a
Handler provides improved performance in virtually any tape handling application. wide range of data processing sub-
Replaceable Capstan Panel permits use as Perforated Tape ReorJer with a . jects, will be presented by individuals
remarkable new brake capable of stopping on the stop character at speeds up
prominent in the industry. Exhibits
to 1000 characters per second. Using a small vacuum loop buffer, Model 906
features: will cover equipment, components and
o Complete front accessibility-singl~ • Capable of continuous cycling at any services related to all phases of com-
panel construction frequency from 0 to 200 cps with- ,puter and data processing systems.
o Pinch rollers capable of 100 million out flutter
start-stop operations Rewind or search at 300 ips Installation of exhibits will begin
• In-line threading,· end of tape sens- • Better than 3 ms starts on Dec. 2. They will be open from
ing and tape break protection • Better than 1.5 ms stops
9:30 a.m. to 9 p.m. on Dec. 3, and
• Speeds up to 150 ips • Tape' widths to 1-1/4"
• As many as 4 speeds forward and • 'Up to 47 channels
. from 9:30 a.m. to 6 p.m. on ,Dec. 4
reverse • All functions remotely controllable and 5. Handling all exhibit arrange-
The 906 may be supplied with a transistorized Record-Playback Amplifier ments is John Leslie Whitlock and As-
featuring a separate module for each channel. Electronic switching from record soc.~ 6044· Ninth St., Arlington 5, Va.
to playback function is available as an optional feature.
Vice Chairman for Registration is
Potter also manufactures a complete line of Magnetic Tape Handlers, Perforated
Tape Readers, High Speed Printers, Record-Playback Amplifiers and Record-Play- William E. Bradley, Philco Corp.,
.back Heads. G and I Division, Philadelphia .
Contact your Potter representative or call In the November/December issue,
or write direct for further information. DATAMATION will feature complete

~
and detailed coverage of the Eastern
POTTER INSTRUMENT COMPANY, INC. J oint Computer Conference with pic-
Sunnyside Boulevard, Plainview, N. Y. tures, complete program details and
Tbe mark of OVerbrook 1-3200
E' Q r exhibitor information. This is in keep-
ing wIth our policy of providing full
Potter has career opportunities for qualified engineers who
coverage for every Eastern and West-
like a challenge, and the freedom to meet it.
ern Joint Computer' Conference.
Circle 10 on Reader Service Card

40 The Magazine oj DRTRi\4£.=i-r'! CJ i'-J

 COMPUTER SYMPOSIUM
armour sponsors for fifth year
With the purpose of providing an effective medium of communication for
persons and organizations concerned with the broad field of computers,
Armour Research Foundation is again sponsoring-for the fifth year in suc-'
cession-a Computer Applications Symposium. The 1958 meeting will be
• held at the Morrison Hotel, Chicago, on October 29 and 30 .
Symposium committee chairmen, F. C. Bock, R. B.Wise and M. J. Jans,
have stated that the sessions will stress the use of new computers and ac-
cessories, new techniques of computer programming, organization and
operation of computer installations . . . and new applications. (This year
Armour Research Foundation of Illinois Institute of Technology, have an-
nounced the expansion of their computing facility.)
Reservations can be made for one or both days by writing to Mr. M. J.
Jans, Conference Secretary, Armour Research Foundation, 10 W. 35th St.,
Chicago 16, Ill. Fee is $20 for one, $30 for two days, luncheon included ..
Monday morning, October 29: 8:00-Registration. 9:00-Welcome ad-
dress by Mr. H. A. Leedy, Director, Armour Research .Foundation. 9:10-
"Operations Research and the Automation of Banking Procedures," R. A.
Byerly, Director of Research,. National As.sociation of Bank Auditors and
Comptrollers. 9:50-"Information Systems Modernization in the Air Material
Command (Univac 1105; IBM ,709)," D. E. Ellett, Colonel, USAF, Chief
of Data Development Division, Air Materiel Command. 11:00-"Utilization
of Computers for Information Retrieval," Ascher Opler, Consultant with
Computer Usage Company. INFORMATION
Monday afternoon: 12:20-"Problems and Prospects of Data Processing
for Defense," C. A. Phillips, Director of Data Systems Research Staff, Of-
SEARCHING
fice of the Assistant Secretary of Defense (Comptroller). 1:50-"An Inte-
grated Data-Processing System with Remote Input and Output (NCR 304),"
and RETRI EVAL
R. D. Whisler, Systems and Audit Manager, S. C. Johnson & Sons, Inc: A new and major program at
2:30-"The Role of Character Recognition Devices in Data-Processing Sys- The Ramo-Wooldridge Corpo-
tems," R. L. Harrell, Director of Electronic's Processing, The Reader's Digest ration is devoted to the design
Association. 3:40-:-"Inlmt-Output, Key or Bottleneck?" R. D. Elboum, Chief and development of a large-
of Components and Techniques Section, Data' Ptocessing Systems Divi~ion, scale system for the automatic
U. S. Department of Commerce. 4:20-Pariel discussion between session handling of reconnaissance
chairmen arid speakers. t
information. The basic systems
problems include the handling
Tuesday morning, October 30: 8:00-Registration. 9:00-Welcome ad- of ordinary language on com-
dress by V. H. Disney, Manager of Electrical Engineering Research Depart- puters and the design of auto-
ment, Armour Research Foundation. 9:10-"Scientific Uses of a Medium- matic searching and retrieval
Scale Computer with Extensive Accessory Features (IBM 650)." Richard A. techniques.
Haertle, Supervisor, Engineering Mathematics Group, AC Spark Plug Di- Inquiries are invited from elec-
vision, General Motors Corporation. 9:50-"Optimizing Designs with Com- trical en gineers, mathematicians
puters," D. D. McCracken, Associate Research Scientist, Institute of Mathe- and physicists whose back-
matical Sciences, New York University. 11:00-"Computer Applications in grounds include operations re-
search analysis and systems
the Numerical Control of Machine Tools," R. B. Clegg, Engineer, Servo
analysis of digital computing
Machine Tool Division, Kearney and Trecker Corporation. equipment.
Tuesday afternoon: 12:20-"Frontiers in Computer Technology," R. W.
For additional information,
Hamming, Member of the Technical Staff, Bell Telephone Laboratories,
write to Mr. Leslie Levin.
1:50-"Computer Sharing by a Group of Consulting Engineering Firms,
(Bendix G-15D)," E. M. Chastain, president, and J. McCall, general man-
ager, Midwest Computer Service, Inc. 2:30-"Cllrrent Developments in The Ramo-Wooldridge
Computer Programming Techniques (IBM 650, Univac 1)," Frederick Way,
III, Assistant Director, Computing Center Case Institute of Technology.
Corporation
3:40-"The Future of Automatic Programming (Univac 1103A, IBM 704)," P. O. BOX 45215, AIRPORT STATION
Walter F. Bauer, Director, Computation and Data Reduction Center, Space LOS ANGELES 45, CALIFORNIA

Technology Laboratories. 4:20-Panel discussion.

Circle 11 Ofl Reader Service Card.

Research & Engineering, September/October 1958 41

DIGITAL COMPUTER AIDS
IN OPTICAL SYSTEMS DESIGN
era of "profit-squeezing." Design complexity, with its
associated volume of mathematical problem solving and
performance prediction, is especially prevalent in the field
of advanced optics.
In optical system development, the mathematics em- •
ployed actually predict the performance of the lenses and
r lens systems in terms of the deviations of light rays pass-
I
ing through the system from those optical paths that would
U'
\ give the desired object-image relationship.
\._---
These actual deviations from the theoretically required
paths are imposed by the physical limitations of optical
materials. Also, there exists always some finite difference
between the nature of the point-source world and physical
reality.
This relationship frequently imposes severe problems
and involved calculations in optical design because the
product of these deviations is the difference between the
object as considered by the optical system and the image
actually produced by the system.
The difference between the image and the object are
grouped under such nomenclature as focus, resolution,
by EUGENE THORBURN depth of field, and other representative optical perform-
Optical Engineer
Pacific Optical Corporation
ance parameters.
No one has ever constructed a perfect lens system. This
Common to the successful solving of practically all design is a physical impossibility. Fortunately though, it is possi-
problems in the development of today's highly complex ble to accurately predict what degree of imperfection any
physical systems is the mass of computations that must be given system will have.
continually processed as the work progresses. But this process of predicting lens system performance
How well this load of computation is managed may well is highly complex, and its solutions require not only long
be the determining factor in the profit or loss aspects of a but also tedious mathematical calculations. The extent
new development; at the least, it strongly influences the of the mathematical labor borders on the fantastic and
degree of precision attained in the final product. can frequently price a system right out 6f the realm of
The formulation of the basic approach, the selection practicality, certainly remove it from a competitive cost
of the appropriate theories and concepts, the garnering range.
of the necessary data, and the establishment of the correct The length and complexity of the mathematics required
design procedures, these are all matters within the tech- can rapidly be appreciated by considering that, prior to
nical control of the design engineer. the advent of electronic computing methods, a competent
But the rapid and accurate evaluation of mathematics lens designer often spent two or more years developing
that represent the system's performance is not. For this, and perfecting a lens system of average complexity. Now,
the design engineer today is dependent on the extent and with the rapid advances in the field of optics, manual
appropriateness of the computing and calculating equip- calculations by a designer could take a lifetime.
ment that is available to service him at the right time The majority of the designer's calculating time is not
and at the right place for maximum speed in the hand- necessarily spent in formulating the basic system design.
ling of this mathematical load. In most instances, he has to crank through the arithmetic
Mathematical evaluations on what any system will involved in determining the effects of various adjustm~nts
accomplish when it is built is called "proving out" the in the component characteristics of the overall system. This
design philosophy. And unless continuation of the design means that, whenever he substitutes one component for
approach can be maintained by frequent mathematical, a more suitable lens part, he has to recalculate these
evaluations accomplished quickly, it will either delay final changes and how they effect the overall system.
production of the system or waste a great deal of the
design group's time - a serious cost penalty in today's optical firms and computers

Diagram (above) illustrates parameters involved in mathematical Before electro-mechanical desk calculators first came
representation 0/ path o/light ray through optical systems. System along, the designer would spend years 'calculating any
performance is measured in terms 0/ aberrations which are devia J given project. The new calculators reduced the figure
tions 0/ actual image from image derived assuming perfect lenses. to months of computing time. But even this considerable

42 The Magazine of:C RTRl\,4HTI 0 fL'\3

 reduction left tHe overall cost in time too heavily balanced trends of the system as shown by the computations. On
in favor of essential computing time. the basis of the calculations, complemented by his ex-
Although the later developed electronic computers are perience and design judgement, the designer must make
now frequently applied in the more complex segments of those interacting changes and adjustments in lens configu-
general indu~try, their use by optical firms has been rela- ration, material, and system concept that will eventually
• result in satisfaotory system performance.
tively rare. .
This lack of utilization might possibly be explained by This cut-and-try procedure is as old as optical system
the fact that, although optical manufacturers have had design, but at least relief from the drudgery of the com-
very definite need for their own computers, the initial cost putations has been provided by the digital computer.
of· computing systems in relation to the potential sales
volume of optical systems was difficult to justify by cost- optical mathematics
conscious management.
On the other hand, a management team in the optical The figure on page 42 is a functional diagram of the
industry, aware of all contributing phases of their indus- relationship between light rays and optical surfaces that.
try's problems, needs to balance a considerable first fi- form the basis of optical system design. This interaction
nancial outlay against direct savings in design labor costs and its associated mathematics must be investigated at
and indirect saving resulting from more rapid completion each optical surface of the system and for a large enough
of final deliveries of the systems. assortment of light rays from various portions of the ob-
In a great many cases it has been proven that the in- ject to give a proper evaluation of the system performance.
troduction of fully electronic computers greatly reduced Opening and surface equations would express the relation-
the amount of time spent in optical system development ships at the first surface. Transfer equations would relate
and changed the status of optical designers from arith- the results of the first surface effects of the second surface
metical monitors to creative engineers. where the surface equations must be applied again. Clos-
Under the regime of the electronic computer, calcula- ing equations would establish the coordinates of the light
tions as well as evaluations of the complex mathematical ray as if reaches its focal point at the end of the system in
representations of optical system performance requires terms of height above the optical axis and distance from
not months, but minutes, and more often seconds. the theoretical focal plane.
To the casual observer the mathematics predicting the The final coordinates of the ray and the intermediate
performance of a single lens or the composite of lenses orientations are exact traces. of the ray's path through the
in a multi-lens optical system might seem rather elemen- system. The deviations of this path from the path pre-
tary in comparison to the highly complex forms in use dicted by assuming perfect lens performance are a meas-
in today's technology. And, no optical designer would take ure of the optical system's performance.
issue with this observation. However, it is not the degree The final design of a lens system requires the thorough
of mathematical sophistication involved, but the sheer investigation of system performance by means of this
weight of the computational burden that has turned the
o1?tical designer to the use of digital· computers. Royal McBee's LGP-30 desk-sized digital computer is now operating
in the. design department of Pacific Optical Corporation. The com-
This mass of mathematical labor is the result of optical
puter's flexibility and memory capacity make it well suited for the
design being· more of an art than science. While the re- field of optical systems design, according to Pacific officials.
lationships between the behavior of light rays and the
characteristics of various media are exactly bound by un-
equivocal equations, the utilization of these relationships
to produce high performance optical systems depends to
a major degree on the judgement, experience and patience
of the optical designer.
The various optical aberrations that constitute the de-
viation of the actual image produced from that produced
by a theoretically perfect lens system cannot be singled
out one by ·one and corrected without certain penalties
in other aspects of system performance. This interaction
between the several forms of aberrations require that op-
tical designers operate in a constant state of compromise
to arrive at the 'optimum' design.
This nece~sity for compromising advantages and weigh-
ing disadvantages sets the requirements for the ability of
the designer to follow very carefully the performance

Research & Engineering, September/October 1958

OPTICAL SYSTEMS DESIGN

.exact ray trace method. As many as thirty rays testing the tem modifications, since the amount of computation was
performance of various portions of the lens surfaces must drastically reduced in comparison to the exact ray trace
be carried surface-by-surface through the system. The lens procedure. However, the utilization of the LGP-30 has
designer observes the system p_erformance in terms of the permitted more frequent application of the ray trace tech-
ray orientation and distribution, and makes adjustments nique. The entire ray trace procedure has been pro- .,
and modifications required to optimize the system. grammed on the LGP-30.
Fortunately, the nature of optical aberrations is such As in the programming of the third order aberrations,
that they may be pitted against each other to achieve forty optical surfaces may be 'considered, and the inputs
overall improvements. That is, carefully chosen use of of curvature, thickness, and index of refraction associated
certain aberrations in some lenses may result in a counter- with the several surfaces are all that are required. Com-
ing of the aberrations resulting from other lenses to the puter output consists of Y, sin I, and sin U, at each sur-
end that the final overall system aberration is much smal- face plus values of Y, U, hs, and Xs, at the focal plane.
ler than the individual lens contributions. Consideration of the capabilities of digital computers in
This portion of the· design procedure places the greater optical design problems have led to the concept of utiliz-
portion of responsibility for success on the designer's ing the computer as a means of accelerating the optimiza-
mastery of the "art." His ability to recognize the nature of tion process. Under the proposed system the basic optical
the aberration, his knowledge of the most efficient correc- system would be established and the corresponding sur-
tive action, and his appreciation of the effect of the cor- face data fed to the computer. A suitable criterion for
rective action on various other system parameters, marks optimal system performance would be established as the
the difference between success and failure of the design. It computer objectives. A program would be proposed per-
is in this area of the design effort that the digital com- mitting the computer to make adjustments in the charac-
puter, by furnishing the designer with rapid evaluations of teristics of the surfaces on the basis of systematic trial and
the effects of his design judgement, proves most valuable. error operations.
Pacific Optical Corporation is devoting considerable
two design aids effort in the development of such a computerized design
program. In fact, the anticipation of the long range. neces-
To permit rough estimates of system performance' dur- sity for and advantages of such a program had considera-
ing the preliminary design stages, optical designers em- ble weight in making the choice of computers be pur-
ploy approximations to the ray trace equations which pro- chased. The flexibility and storage capabilities of the LGP-
vide reasonable evaluation of the third order aberrations 30 make it suited for application to these computing con-
and overall system performance. The usual procedure is cepts, according to Pacific Optical officials.
for the designer to prepare, on the basis of past experience In considering any segment of our rapidly advancing
and theoretical performance calculations, the complex of technology, no part can be isolated from the whole. Every
lenses and optical surfaces he deems necessary to perform science today is being buffeted and shaped by the needs
the required optical task. and demands of other sciences.
Once the basic system has been established the third Our recent leap into space with missiles and satellites
order aberrations are computed. At Pacific Optical, the has loosed a flood of demands for more precise and elab-
LGP-30, purchased from the Royal McBee Corp, has been orate optical systems for visual tracki~g, astranavigation.
programmed to perform this series of computations. The The streamlining of industrial manufacturing is open-
capability of the program is such that systems consisting ing a broad market for optical measuring techniques yield-
of as many as forty optical surfaces may be analyzed. ing increased resolution in process control systems.
To use the computer, the designer feeds in the curva.: Television is impatiently awaiting improved camera
ture, thickness, and index of refraction associated with lenses, motion pictures are desparately searching for bet-
each surface of the system. The output of the computer ter depth dimension effects, and the progress of aerial
consists of the following aberrations: spherical, coma, as- photo reconnaissance and mapping in three dimensions is
tigmatism, distortion, transverse longitudinal color, trans- hungry for improved equipment.
verse oblique color, and Petzval curvature. Nor is the matter entirely one of merely broadening
These values are printed out in terms of the contribu- and refining the product. Along with expanding applica-
tions of each surface, and the total value of each form of tions has come a compacting of the time with which these
aberration is also printed. Plotting these values permits new demands for optical equipment must be satisfied.
the designer to re-evaluate the performance of the system On both of these counts, broadened application and
and begin the series of modifications that will lead to the sharply constricted delivery schedules, the in-plant, read-
final design. ily available computer has become inevitable if optical
Previously, the majority of design work was done using systems manufacturers are to meet their responsibilities
the third order' aberrations, except for the very final sys- in the years ahead.
Circle 114 on Reader Service Card.

44 The illagazine of DR T R 1\.1'1

 neWCRTR I literature
TRANSISTOR MANUAL: This is Memory model SM-10; T-Pac Static Study of the Stability of a Molten
the third edition of this transistor Flip-Flop model FS-10; T-Pac Thyra- Zone Refining Process used in the
manual first introduced in 1957. Ful- tron Driver model TO-10; Indicator Production of Transistors," it describes
ly covering circuits of various types, Panel model TI-10; Thirty Unit Delay and illustrates by diagrams technique
a completely revised section on appli- Chassis model DU-10; Plugboard of simulating on the analog computer,
cations and giving specifications, it Panel model PB-10. For copy write conditions experienced by a germani-
tells in its 168 pages how to build COMPUTER CONTROL COMPA- um rod under various conditions of
almost everything using transistors. NY, INC., 92 Broad Street, Wellesly temperature, rod diameter, and length
Describing basic semiconductor the- 57, Massachussetts or use reader card. of molten zone. Explained is how the
ory, the meaning of transistor para- Circle 202 on Reader Service Card. analog computer can quickly, accur-
meter symbols, how to read a tran- ately solve non-liner problems for
sistor specification sheet, the company conditions impossible to observe ac-
contends it is of interest to expert curately in their true state. For copy
FLIGHT DATA SYSTEM: Twenty-
and student. Copies are priced at write ELECTRONIC ASSOCIATES,
three page brochure on this manu-
$1.00. For copy write GENERAL INC., Long Branch, New Jersey
facturer's Victor System 272 has fif-
ELECTRIC COMPANY, Semiconduc- or use reader service card.
teen diagrams and photographs illus- Circle 205 on Reader Service Card.
tor Products Department, Syracuse, trating the FM Flight Test Data Sys-
N. Y. or use reader service card. tem designed to record and process
Circle 200 on Reader Service Card ..
data aboard flight test aircraft with
high accuracy. Automatic correction TAPE RECORDER: The new Model
occurs by mechanical and electronic C-IOO series of instrumentation tape
means when data is processed through recorders is fully described and illus-
DIGITAL CONTROL COMPUTER: reduction equipment at ground sta-
Four-page reference bulletin describes trated in this four-page bulletin, avail-
tion. For copy write VICTOR ADD- able free to qualified members in the
the RW-300 Digital Control Comput- ING MACHINE COMPANY, Gov-
er. Process control, data logging, pilot field. Characteristics, operating fea-
ernment Contract Office, 3900 North tures and complete engineering spe-
plant, and test facility applications Rockwell Street, Chicago 18, il-
for computer control systems are dis- cifications of the new transistorized
linois or use reader service card. recording system are given. Simplicity,
cussed and detailed specifications are Circle 203 on Reader Service Card.
listed.· The computer incorporates an- compactness and dependability of the
alog-digital conversion equipment, C-100 series is stressed. For copy
handles up to 1,024 analog inputs, up write MINCOM DIVISION, Minne-
to 128 analog outputs. Operates with PORTABLE OSCILLOGRAPH: New sota Mining and Manufacturing Com-
wide range of digital input and out- portable two-channel oscillograph pany, 2049 South Barrington Avenue,
put equipment: automatic typewriters package with built-in amplifiers, Los Angeles, California or use card.
Brush Mark II, is fully described in Circle 206 on Reader Service Card.
paper tape and punched card readers,
paper tape and card punches, and on- this four-page, two-color bulletin De-
off devices. For copy write THE signed for applications considered im-
THOMPSON -RAMO- WOOL- practical for direct writing recording
of electrical and physical phenomena FILE-COMPUTER SYSTEM: Book-
DRIDGE PRODUCTS CO. P. O. Box
this recording unit requires no addi- let U 1562 describes company's Model
45067, Airport Station, Los Angeles
tional equipment for operation, fea- 1 File-Computer Data Automation
45, Calif. or use reader service card.
Circle 201 on Reader Service Card. tures pushbutton selection of our System. Features described include:
chart speeds. Low inertia, frictionless, concept of building block construc-
the Mark II records directly onto ink tion, random access to the magnetic
paper. For copy write BRUSH IN- storage drums, advantages of com-
•. DIGITAL MODULES: An eight-page STRUMENTS, Division of Clevite bined internal and external program-
Corporation, 3405 Perkins Avenue, ming, variety of input-output equip-
supplement to this man~facturer's
Cleveland 14, Ohio or use reader card. ment. Also discussed: the parity check,
Catalog T describes the six new addi-
Circle 204 on Reader Service Cord. automatic character and blockette
tions to the Series T group of tran-
counters, checking of computations by
sistorized digital modules. Illustrations
reverse arithmetic process. For copy
and prices for each model are in-
write REMINGTON RAND Division
cluded along with a concise descrip- ANALOG COMPUTER TRANSIS-
of Sperry Rand Corporation, 315
tion of the unit and its applications. TOR: A six-page booklet deals with
Fourth Avenue, New York 10,
The models incorporated in this cata- one phase of the transistor production
N. Y. or use reader service card.
log supplement are: T -Pac Serial cycle. Titled, "An Analog Computer Circle 207 on Reader Service Card.

Research & Engineering, September/October 1968 45

NEW LITERAURE

PROGRAMMING THE FILE-COM- applications. Photographs and dia- write AUTOTRONICS,·· INC., Dept.
PUTER: A· 249-page manual, pre- grams, plus chart of relative charac- # 16; Rt. 1, Box #812, Florissant,
pared for trained programmers and teristics of different types of power Missouri or use reader· service card.
computer operators, explains how to supplies are included. For copy write Circle 210 on Reader Servic(! Card.

program the Univac file-computer. GENERAL ELECTRIC COMPANY,

The book is divided into eight prin- Rectifier Department, Lynchburg,
cipal topics to facilitate study of spe- Virginia or use re'ader service card.
cific areas of programming and in- Cire/e 209 on Reader. Service Card. ANALOG-PUNCHED TAPE SYS-
corporates over 175 illustrations and TEM: This manufacturer's Model ZA.
b:ock diagrams. Of value to business:" 750 Analog-Punched Tape Data Sys-
men considering using this company's tem· is described in a four-page bro-
CLUTCHES AND BRAKES: Electro-
file-computer for data processing. For chure. Principles of operation, output
magnetic clutches and brakes of a
copy write REMINGTON RAND DI- format, technical specifications and
new miniature line are outlined with
VISION, Sperry Rand Corporation, construction are given. Designed to
specifications in company's twenty-
315 Fourth Avenue, New York 10, meet requirements for a rugged, eco-
six page booklet. Schematic diagrams,
N. Y. or use reader service card. nomical, low speed data processing
dimensional data and minimum per-
Cire/e 208 on Reader Service Card. system, model ZA-750 contains plug
formance curves are provided for ten
patch-board permitting quick and
models. Characteristics and general
easy changing of program format. For
data is given on last page. These units
copy write ELECTRONIC ENGI-
DC POWER SUPPLIES: A six-page feature high torque rating, rapid re-
NEERING COMPANY OF CALI-
folder describes manufacturer's cus- sponse, zero backlash, - are light
FORNIA, Sales Department, 1601 E.
tom built DC power supplies for com- weight and moderately priced, accord-
Chestnut, Santa Ana, California .
puters, aircraft, military and special ing to the manufacturer. For copy Circle 211 on Reader Service Card .

ENGINEERS-MATHEMATICIANS

SYLVANIA OPENING NEW DATA PROCESSING

LABORATORY IN SUBURBAN BOSTON
"Programmed" staff expansion offers exceptional growth opportunities
to engineers and scientists in Digital Computer Activities , .... - ........
..............
If you have the potential to assume increasing technical responsibilities in this field,
look into the expanding analysis and development activities at Sylvania's
newly formed DATA PROCESSING LABORATORY where Sylvania
-,.,..- -- -- --
....
........

-- --
engineers are engaged in the High Speed Data Processing Phase of
the Ballistic M}ssile Early Warning System- known as
BMEWS - in addition to a number of other projects;
• SR. DEVELOPMENT ENGINEERS
Assignments here are of a nature to bring rapid .-,"'"
professional growth to the talented and _ ;''- - • SR. SYSTEM ANALYSTS
creative man. The environment .-
encourages initiative and .-,- • SR. PROGRAMMERS
original thinking. .... - -
• MATHEMATICAL ANALYSTS
------------:::. --
."""
Positions are open for:
--- -... --- ...
• ENGINEERS WITH EXPERINECE IN DISPLAY
OF RADAR DATA HANDLING
11

Please send resume in complete (Previous digital data processing experience

confidence to: ERLING MOSTUE - ... -_ is important for most assignments.)
Interview and relocation expenses paid by Sylvania. Inquiries wi~ be ........ -_
answered within 2 weeks. Convenient Saturday interviews arranged.

WALTHAM LABORATORIES Electronic Systems Division

-
-- -... -... -......
~SYLVANIA~SYLVANIA ELECTRIC PRODUCTS INC . 100 First Avenue - Waltham 54, Massachusetts
--
Circle 12 on Reader Service Card.

46 The Magazine of CRTR T!CJN

 NEW LITERATURE

TRANSICON DATRAC: Six-page TAPE RECORDER/REPRODUCER: graphic model of the MC-5800 com-
folder illustrates this company's mod- Model FR-I00A, modular magnetic puter clearly illustrating the unique
ularized, transistorized, reversible ana- tape recorder/reproducer for instru- packaging wherein the computer may
log-to-digital and digital-to-analog mentation, is fully described in this be instantly "unzippered" from the
converters, and alarm limit monitors. fourteen-page, four-color booklet. confines of its cabinet for unob-
Line drawings and a description of Photographs and· information explain structed access during maintenance.
the models is given. Folder contains how this latest equipment is being Other features of the computer are
ten pictures of Transicon Datrac plug- used and back page lists company described including circuit logic re-
in building blocks and specifications equipment specialists where specifica- quired for building-block flexibility,
on analog-to-digital, digital-to-analog tions may be obtained. For copy write adaptability for high speed repetitive
converters and alarm limit monitors. AMPEX CORPORATION, Instru- operation, bivariable function genera-
For copy write EPSCO, INC., Equip- mentation Division, 860 Charter tion, complete automatic problem
ment Division, 588 Commonwealth Street, Redwood City, California check. For copy write MID-CEN-
Avenue, Boston 15, Mass. or use card. or use reader service card. TURY INSTRUMATIC CORP., 611
Circle 212 on Reader Service Card. Circle 215 on Reader Service Card. Broadway, New York 12, New York.
Circle 216 on Reader Service Card.

ANALOG COMPUTER: A fully il-

lustrated 18-page brochure and folder INTEGRATED DATA SYSTEM:
BASIC SWITCHES: A revised fliir-
describes MC-5800 analog computer. Systems for gaining instrumentation
ty-two-page edition of Basic Switch.
Fourteen pages of specifications and comput,d data in rapid time is de-
Catalog 62c, with illustrations and
descriptions are included in a four- tailedm this two-page folder. The
diagrams, gives details of over 200-
page folder. The folder is a photo- manufacturer's Automatic Data Re-
catalog listings of basic switches and
related devices. Page twenty-seven
contains technical information on in-
stallation, mounting and proper se- NEW PRODUCT FEATURE
lection. Switches detailed include the
new high-precision roller lever switch;
the adjustable actuator switch for fine
adjustment; the "pulse" switch for
securing electrical impulses without
complicated actuating mechanisms.
For copy write MICRO SWITCH, A
Division of Minneapolis-Honeywell
Regulator Company, Freeport, Illi-
nois or use reader service card.
Circle 213 on Reader Service Card.

DATA ACCUMULATOR: New auto-

matic inspection data accumulator for
tinplate, model GE 302, is covered
in this four page booklet. Designed NEW ADVANCED DESIGN FOR
specifically for an industrial environ- MAGNETIC MEMORY DRUM
ment, maker contends this new mag- A small, lightweight, aluminum, low cost magnetic drum head, .3122 dia. by 1~" long, and
netic drum system reduces accus- .024 track width, has been developed by Data Storage Devices Division of Midwestern Instru-
tomed number of electronic circuits; ments, Tulsa, Okla., and is now in production. The OL-12-E features balanced low impedance
windings, low record current, and high playback voltage for use with transistorized circuits.
increases system reliability by reduc- Bit densities of up to 200 bits per inch at 1 mil spacing; read and record information at high
ing complexity. Booklet contains il- frequencies. Other features include all-metal construction, continuous operations at high tem-
lustrations, diagrams and specifica- perature, milled flat perpendicular to gap. The precision features of this head are typical of a
tions. For copy write GENERAL complete line of magnetic heads manufactured by Midwestern, including digital tape heads
capable of 2,000 bits per channel per inch with 100 percent resolution, video heads to 4.5 mc,
ELECTRIC COMPANY, Computer and heads for all types of analog and binaural audio. Data Storage Devices Division is geared
Department, 1103 North Central Ave- for volume production of a complete line of standard and special application drum heads, mag-
nue, Phoenix, Arizona or use card. netic tape heads and magnetic memory drums. For further information contact Midwestern
Circle 214 on Reader Service Card. Instruments Data Storage Devices Division, P. O. Box 7186, Tulsa, Okla. (Adv.)
Circle 13 on Reader ServIce Card.

Research & Engineering, September/October 1958 47

 Data Processing Digest is a monthly
How to service containing carefully pre- NEW LITERATURE
get cost- pared abstracts on management'~
newest tool-Electronic Data Proc-
saving essing. Each month you receive
ideas for complete information on actual ac-
complishments in data processing
ELECTRONIC programs. DPD editors review
articles from almost 100 periodi-
DATA cals monthly-plus all important cording Monitoring System, designed
PROCESSING
books-to give you the facts and
key ideas.
for acquisition and processing of data
from rocket or jet engines, is out-
••
,
h

fREE examination copy of DPD lined with a system block diagram il-
Let us send you the latest issue of lustrating data How and typical com-
DPD for Free Examination With- ponents. A small, general-purpose dig-
out Obligation. If the Digest meets ital computer is an integral part of
your needs, keep the issue and O.K. the system. For copy write CONSO-
our bill of only $2 per month (an- LIDATED ELECTRODYNAMICS
nually in advance) for payment. If CORPORATION, Systems Division,
not, return bill under no obligation. 300 North Sierra Madre Villa, Pasa-
dena, Calif. or use reader service card.
Circle 217 on Reader Service Card.

CANNING, SISSON & Associates

1140W S. Robertson Blvd., Los Angeles 35, Calif.

Send me a Free Examination Copy of the current DPD

Name _________________________________________
ANALOG, DIGITAL UNITS: In an
eight-page General Electric booklet,
Company· ______________________________________
computers and computing systems,
Street------------__________________________
analog and digital, are described.
City'_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ State __________
Covered are such subjects as, "GE
..................................................... .. "
\,
100 Electronic Data Processor for
Circle 14 on Reader Service Card.
Banking . Applications," "GE Transis-
torized Magnetic Ink Character
Reader," "GE 310 Data Acquisition
System for Process Monitoring," "GE
302 Automatic Inspection Data Ac-
cumulator for Tinplate," "GE 309
COMPUTER ENGINEERS Gage Logging System," "GE 307
Miniaturized AC Network Analyzer,"
Positions are open for computer engineers cap- "GE 306 Analog Computer," "GE 308
able of making significant contributions to Economic Dispatch Computer," "GE
advanced computer technology. These positions 301 Heat Rate Computer," "GE In-
are in our new Research Center at Newport dustrial Card Reader." For copy write
Beach, California, overlooking the harbor and GENERAL ELECTRIC COMPUTER
the Pacific Ocean - an ideal place to live. These
DEPARTMENT, 1103 North Central
are career opportunities for qualified engineers
in an intellectual environment as stimulating as Avenue, Phoenix, Ariz. or use card.
Circle 218 on Reader Service Card.
the physical surroundings are ideal. Qualified
applicants are invited to send resumes, or
inquiries, to Mr. L. T. Williams.
Positions Open: Areas of Interest:
Systems Engineers Computers & MINIATURE CLUTCH AND
Logical Designers Da ta Processors
Programmers InputjOutputEquipment
BRAKE: Four page leaHet on manu-
Circui t Engineers Storage Units facturer's new line of miniature
Mechanical Display Devices clutches and brakes. Applications in-
Engineers Computer Components
Applications Solid State Devices clude drive for tape reader; position
Specialists Electromechanical type servo for analog and digital com-
Sales Engineers Equipment
puter. LeaHet gives dimensional. data,
AERONUTRONIC SYSTEMS, INC. specifications and minimum perform-
a subsidiary of Ford Motor Company ance curves for electro-magnetic
1234 Air Way. Bldg. 26, Glendale, Calif. • CHapman 5-6651 brake model BF-125; clutch model
C-125 and brake-clutch model MC-
125. For copy write AUTOTRONICS,
INC., Dept. #16, Rt. 1 Box #812,
Florissant, Missouri or use card.
Circle 219 on Reader Service Card.

48 The Magazine of DRTR1\4i~Ti C 1'J

 XPLORE
ANEW.
UNCHARTED
AREA IN
INFORMATION
HAN DLI NG At-ReA's NEW
Astro-Electronic Products Division, Princeton, New Jersey

This new division makes ava ilable immediately cre- translation and information retrieval; and digital
ty ative opportunities in an advanced and challenging computer design and engineering.
area of automatic data processing. The long-range To participate in this program applicants must
research program now beginning will involve the so- have advanced technical degrees, or equivalent
ro- lution of logic problems and the programming of experience, and a record of achievement. A govern-
itS
deductive rules (non-numeric) and systems engi- ment security clearance is required of all project
lry
neering. personnel.
Technical proficiencies are needed in all branches These opportunities offer stimulating positions in
my ~
be
of information handling including the following an atmosphere conducive to creative activity as part
HY areas of specialization: information systems design, of the Princeton research community. For a personal
development, analysis and synthesis; programming interview, communicate with Mr. D. D. B,rodhead.
nts
gic

e
research and logical design; automatic language ..

~QD
RADIO CORPORATION
OF AMERICA
Astra-Electronic Products Division
Call Collect, WAlnut 4-2700
Or Send Resume to Dept. PE-13
e
Princeton, New Jersey
Circle 16 on Reader Service Card.
 In use now, as a part of several systems, the Bendix G-15
has proven itself the ideal digital computer for data reduction
and control. Its high speed ... versatile command structure ~ ..
widely varied methods and means of input and output ...
small physical size ... and its low cost, all co~tribute to the reason
why the G-15 is being selected for use in more and
more on-line applications.
The G-15 is the fastest general purpose computer in the low
price field. For real-time control applications, this speed is often
important. In at least one case the G-15 has beenchosen
for real-time computation where only a million dollar computer
has ever been used before.
The versatility of the G-15's basic programming system con-
o Process controls tributes heavily to its .ability in on-line applications. Commands
€) Missile tracking and impact prediction systems are available for shifting with tally, extracting and assembling of
E) Engine test stand data reduction and control words, overflow indication, branching, block data transfer,
o Wind tunnel data reduction and control and many other special functions valuable in on-line use.
Perhaps of greatest importance is the computer's unique variety
o Real time position determination for mobile
of input-output possibilities; The basic G-15 includes an
missile launching sites
electric typewriter for input-output and control, as well as a
o Machine tool controls and many other open
paper tape punch and magazine loaded high-speed photoelectric
and closed loop systems
tape reader. Punched card, and magnetic tape units are
available and all may be connected at the same time through
the computer's buffered input-output registers.
Other devices such as A to D or D to A converters may be
connected simultaneously or in place of the above mentioned
accessories, and operated under control of the computer.
Finally, information can be directly written on or read from the
memory drum, llllder control of special external devices.
~

. Note that all of these methods of input and output

can be utilized without any modification of the computer.
Connectors are provided on the rear of the G·15 for
each type of input and output described.

If ruggedness is required, the G-15 can prove an enviable record.

Two of them have. been in use for well over a year bolted
directly to the deck of a ship, and have shown a performance
record to be envied by any comput~r based on solid ground.
During a recent six months period the average up-time for all of
the over 100 Bendix G-15's in the field was 95.4%.
The G.-15 is compact, too, occupying just six square feet of
floor space. Of course, it can be used as a powerful general
purpose computer, as well as for on-line applications. A variety of
simplified programming systems is available, including the
renowned INTERCOM 1000, which can be used after
four hours training or less.
The reasons are many ... but the fact is that more and more G-15'~
are being leased or purchased for on-line use. If you would
like to discuss your own requirements, we would be pleased to
work with you. Write Department E-6 for information.

DIVISI.ON OF BENDIX AVIATION CORPORATION

I
L.OS ANGEL.ES 45, CAL.IFORNIA

Circle 17 on Reader Service Card