Transistor Products

History of Clevite Transistor Products

Founding in 1952

Transistor Products was incorporated mid 1952 in Rahway New Jersey as a new subsidiary of Purolator Products Inc. A brief announcement advised that a license had been obtained from Western Electric and that Scientific Specialties Co of Boston would produce specialised equipment for quality transistor manufacture. [Tele-Tech 1952] Purolator produced an extensive range of fuel and oil filters. Their rationale for entering the electronics industry is unclear.

Roland B Holt, its President and former Director of the Nuclear Physics Laboratory at Harvard University headed the research. Holt owned Scientific Specialties Corp which produced laboratory equipment. Transistor Products first premises were in Brighton, Boston, Massachusetts.

Its founding took advantage of the commercialisation of the transistor following its invention by Bell Laboratories. Transistor Products were one of some 35 licensees who had paid a one time license fee of $25,000 and who attended a Symposium hosted by Bell for eight days in April 1952. The purpose of the Bell Symposium was to transfer Bell’s semiconductor know-how developed over the five year period from the end of 1947 when it invented the transistor through to 1952 when it began to make licenses to its technology available to US and foreign companies. [Burroughs 1953 (1)]

Research and Development was led by Dr Richard Johnston, a recent graduate from Harvard. Mr Ed Quirk, also a Harvard alumnus, had been recruited from Western Electric, Allentown as Production Manager. Mr. Quirk was responsible for organizing and managing all manufacturing and engineering operations for transistors and crystal diodes. [Radio News 1958]

These were key recruitments and together with the Bell license, formed an unhelpful exclusive dependency on Western Electric technology and a fruitful relationship with Harvard University.

Progress by 1953

By March 1953 the company had a staff of about 25 persons including consultants. They had produced a remarkably wide range of prototype products including gold bonded diodes, point-contact and grown junction transistors and photo devices produced from germanium made by the company. Development had begun on alloy junction power transistors.

In his Oral History Dr Neville Fletcher recalls working for the company in the Summer of 1953. At that time Transistor Products was producing germanium from germanium dioxide. This was zone refined and used to grow crystals, including grown junction NPN crystals, for making grown junction transistors. [Ward 2003]

In 1953 the company was producing point-contact transistors under its Western Electric license using a case style identical to the already obsolete Western Electric Type A design. They did, however, claim to produce a superior transistor to those from Western Electric due to their advances in forming technology. Transistor Products were endeavouring to find customers for their point-contact transistors noting that in order to mass produce them, they would need to move to a product more suited to mass production. They hoped that bead encapsulation (pioneered by Western Electric) would prove suitable for scaling up production.

Left: Transistor Products Type 2D Point-contact Transistor courtesy R McGarrah

Page from an early brochure by Transistor Products Courtesy Jack Ward

Nine point contact transistors were advertised in the September edition of Electronics (see below) for applications in “switches, amplifiers or oscillators.”

Amplifier Types

2A 2B 2D 2E

Switching Types

2C 2F 2G

In addition special types were offered to the standard range. Note that the available data only covers seven transistors. Neville Fletcher recalls:

“As with most Transistor Products, the point contact transistors were made on a fairly “manual” production line and then were adjusted and had the point contacts “formed” (by discharging a capacitor) at the end. After this, they were tested and sorted into different categories on the basis of breakdown voltage and current amplification factor. There were, I think, three bins for the final product, and the transistors were given different numbers on the basis of their performance. While I was familiar with the production line at the time, and remember things such as the nice little machine for putting the kinks in the point contact wires, I was not personally involved,” [Ward 2003]

Gold Bonded Diodes

In 1953 Transistor Products had a capacity for 1000 gold bonded diodes per day and were planning on increasing this to 2000 per day. These were all in the form of a plastic bead 3/16ths in diameter. They also produced a high voltage version with two conventional diodes in series mounted within a single plastic case. They were not producing hermetically sealed types.

Above: Transistor Products Gold Bonded Diodes from an Early Brochure Courtesy Jack Ward

Junction Transistors in 1953

In 1953 the Transistor Products range was relatively advanced with four junction transistors on offer. These were all grown junction types based on Bell grown junction technology. Transistors of this kind were very difficult to produce and the first commercial junction transistors were alloy junction types pioneered by General Electric and RCA. Transistor Products reliance on Western Electric’s point contact designs and grown junction technology stunted their early growth. According to reports at the time (Burroughs) they knew little about alloy junction transistors and were dismissive of their prospects.

These transistors were promoted for “audio and low speed switching applications where dependability, long life and minimum space requirements are important.” The transistors were all "X" or experimental types indicating that the company was promoting its products prematurely. The four transistors were:

X-22 X-23 X-27 X-28

Most contemporary listings mention only the X-22 and the X-23 and the status of the X-27 and X-28 is unclear as they are not mentioned in most data books. Transistors have been identified from Garner 1953, Turner 1954 and Kristalldioden- und Transistoren. The X-27 and X-28 from Burroughs 1953.

Transistor Products X-23 Courtesy J Ward and J Hoppe

Photo Devices

The same year the company made an experimental batch of 50 photo-diodes based on the Western Electric

M1740 which dated back to the Bell 1951 Symposium. [Powers 1951] The company offered to scale up production if there was demand. Burroughs evaluated samples in March 1953 finding these diodes appeared suitable for photo transfer from punched tapes. [Burroughs 1953 (2)] These photo diodes were probably X-4 devices advertised in the September edition of Electronics.

Right M1740 photo diode courtesy R McGarrah

The company also made a X-25 grown junction NPN photo transistor first advertised in September 1953 and described in more detail in the November Radio Electronics supplement of Radio and Television News that year.

Report from Radio Electronic Engineering supplement of Radio and Television News November 1953 page 22

Above Transistor Products Journal Advertising in Electronics for September 1953

Transistor Testing Instrument

Transistor products needed to produce test instruments for their own use but offered this to other interested parties.

They branded their instrument the "Transtester." It was designed to measure the small signal parameters of a transistor at any reasonable set of fixed bias conditions. Despite the price of $2150 it did not offer dynamic testing.

Picture credit: Electronics September 1953

Purchase by Clevite Corporation

Transistor Products had a relatively short history as an independent company. In 1953 Clevite Corporation purchased 51% of the company with an option to purchase 100%. [Radio and Television News 1953, Burroughs 1953 (1)]

Clevite Corporation was founded in 1919 as an engineering company and during the World War II had produced bearings for the aviation industry. Post War they sought to diversify into electronics and began with the acquisition of Brush Development Corp in 1952. Brush was also one of the first companies to obtain a Bell transistor license.

Power Transistors 1953-1955

The history of the development of Transistor Products’ power transistors has been extensively described by Joe Knight at the Transistor Museum site [Knight 2007].

Work on power transistors was initiated by Neville Fletcher who worked at the company while undertaking his PhD at Harvard University on the theory of energy levels in semiconductors. His oral history has been recorded by Jack Ward [Ward 2003]. He began at Transistor Products in the summer vacation of 1953:

“As my vacation project, I decided to try and develop a power transistor using a fairly intuitive approach. There was just Shockley’s Book Electrons and Holes in Semiconductors and a few manuals that the company had from Bell Labs but otherwise you went on your general understanding and feelings. The transistor that I made that eventually became the X-78 actually during that vacation had an output of rather more than one watt which was ten times higher than any of the smaller alloy transistors that you could get in those days.”

Early versions of the X-78 courtesy Joe Knight

Somewhat in the style of the times the prototype was quickly advertised early in 1954 through advertising in the January edition of Electronics as shown below. Neville Fletcher confirmed the relatively fast track to marketing the new device: “The “X” means “experimental” and the “78” doesn’t mean anything! The shape of the mounting is rather fortuitous, as I simply made the original from scrap metal and the people in the drawing prettied it up without changing anything much. As I recall, the transistor mount was later redesigned in a better manner and the transistor itself was surrounded by bent copper-sheet cooling fins.” [Ward 2003]

Transistor Products advertising for the new x-78 in the January 1954 edition of Electronics

Editorial in the June 1954 edition of Electronics identified this transistor as the X-78 a pnp alloy junction type for use in Class B audio output amplifiers requiring an output of 2 watts. Matched pairs of transistors were available for this purpose. The X-78 was a conventional alloy junction structure of the kind first developed by John Saby at General Electric.

According to Fletcher, “making alloy junction transistors in those days was a pretty simple operation. You worked at relatively large dimensions…half a millimetre up to a few millimetres. All the electrodes were shaped by placing indium or tin, whatever the thing was, in the slots in a milled graphite electrode, putting the whole thing in something rather like a chemistry department muffle furnace, cooking it for just the right sort of time: you had to get that right because you were dissolving the emitter in from one side, collector in from the other side and you aimed to stop with only a couple of thousandths of an inch between them. That was one crucial part of the actual process.” [Ward 2003]

The alloy die made in this way is shown in the micrograph below where the small insert on the left hand side is the X-78.

Transistor types: (left to right) X-78 standard power transistor (without base electrode); (2) single-bar-transistor; (3) double-bar transistor; (4) multi-bar transistor. The upper row shows the emitter-base side of the transistors and the lower row the collector side. [Fletcher 1955]

Fletcher worked part time at Transistor Products until September 1955 when he returned to Australia. During this time he developed the theory and practice of alloy junction power transistors and in particular the inter-digitated design where the emitter was formed into a bar or comb like structure with the base connection as interlacing fingers. Collector dissipation of up to 100 watts was possible by increasing the area of the collector in combination with cooling fins of varying scale. An example of a double bar structure is shown on the right (courtesy J Knight).

An alternative structure had been published in 1952 by Robert Hall working at General Electric. In his approach, the collector and emitter were constructed in an inter-digitated geometry both on the same side of the germanium die which served as the base. [Hall 1952]

The work done by Fletcher in commercialising these designs earned him the title “the father of the inter-digitated transistor” in an article published in Electronics entitled “Solid State-fingers in the die.” [Tatum 1968]

Work on power transistors was funded by the Signal Corps Engineering Labs for applications to replace vibrator and dynamotor type power supplies. [Tele-Tech 1955]

Range in 1954

This is best described in journal advertising for this year in the March and April editions of Electronics. This shows considerable development of the diodes range but nothing new in transistors.

Transistors for 1955

Late in 1955 Clevite completed its purchase of Transistor Products.

By mid 1955 Clevite also completed its purchase of Intermetall, Dusseldorf, Germany. Clevite had engaged with Intermetall since late 1954 undertaking due diligence on their semiconductor devices manufacturing capability. Part of this process was the evaluation of Intermetall’s OC33 and OC34 low powered alloy junction transistors which were under development in Dusseldorf.

By early 1955 Intermetall were able to mass produce the OC33 and OC34 transistors in small volumes although the earliest public record of this appears to be in May of that year. [Burgess & Gebert 2015]

At this time Clevite did not have comparable low power AF pnp junction transistors thus Intermetall technology was attractive. The USA market offered opportunities for transistors of this type in portable radios. Accordingly Clevite closed a deal on Intermetall and qucickly added the OC33 and OC34 to the Transistor Products range. The acquisition of Intermetall closed a serious gap in the Clevite product range and reduced their dependency on Bell Laboratories.

“Clevite Corporation of Cleveland has purchased Intermetall GmbH of Dusseldorf , Germany, the second largest company in Europe engaged in the development and manufacture of transistors and diodes. The German firm will be operated in conjunction with the parent company’s Transistor Products Inc subsidiary in Waltham Mass. The Waltham division's corporate name is being changed to Clevite Transistor Products.” [Radio and Television News 1956]

In September 1955 Tele-Tech provided a listing of the transistors being produced in the USA. The following were listed for Transistor Products. Note the inclusion of the two new Intermetall transistors. [Tele-Tech 1955]

Point Contact Types2A 2C 2D 2E 2G 2H

2N32 2N33 2N50 2N51 2N52 2N53

Junction Types

X-22 npn X-23 npn OC33 pnp OC34 pnp

Power Types

X102 X107 X120 X122 X125

Picture, courtesy of Joe Knight showing an internal view of a X-107 power type with its cooling fins removed

Transistors for 1956

In December of 1956 Electronic Industries and Tele-Tech published a "1957" chart of transistor specifications.

This shows that the rationalisation of the product ranges of Transistor Products and Intermetall was complete and the "experimental" designation had been dropped. The transistors were now branded CTP or carried JEDEC “2N” type numbers. The range now included silicon transistors developed by Intermetal.

New Silicon Types (Intermetall)

The 2N260, 2N260A, 2N261, 2N262, 2N262A are the Intermetall types OC43, OC44, OC45, OC46, OC47. These Intermetall types were subsequently renamed OC430, OC440, OC450, OC460, OC470 to avoid duplication of the type numbers used by other European producers. [Gebert 2014]

Germanium Range

References

Burgess M Gebert W 2015 History of Intermetall Semiconductors Radio Museum

http://www.radiomuseum.org/forum/history_of_intermetall_semiconductors.html

Burroughs 1953 (1) Technical Memorandum Trip Report: Transistor Products March 9^th 1953 courtesy Burroughs Corporation Collection of the Charles Babbage Institute

Burroughs 1953 (2) Study of Transistor Characteristics Progress Report 17 March 30^th 1953 courtesy Burroughs Corporation Collection of the Charles Babbage Institute

Burroughs 1953 (3) Study of Transistor Characteristics Report R-16 August 17 1953 Research Division of Burroughs Adding Machine Company courtesy Burroughs Corporation Collection of the Charles Babbage Institute

Electronics Design September 1953

Electronic Industries 1956 Electronic Industries 1957 Transistor Specifications Electronic Industries and Tele-Tech 15 12 53-8

Fletcher N 1955 Some Aspects of the Design of Power Transistors Proc IRE 43, 551-59

Garner L E 1953 Transistors, Their Practical Application Coyne Chicago Illinois

Gebert W 2014 The first German Silicon Transistors Radio Museum Forum

Hall R 1952 Power Rectifiers and Transistors Proc. IRE 40 1512-18

Knight J 2007 A Survey of Early Power Transistors TransistorProducts Inc Germanium Power Transistors

Kristalldioden- und Transistoren Taschen-Tabelle 4th Ed. 1963

Philco Transistor Data Manual 1954

Powers D 1951 The Present Status of Transistors and Transistor Applications October 1, 1951

Radio and Television News May 1953 p24

Radio and Television News 1956 February 164

Radio and Television News 1958 Within the Industry April 1958 121

Tatum J 1968 Solid State-fingers in the die Electronics February 94

Tele-Tech 1952 Tele-Tips June 10

Tele-Tech 1955 Radarscope May 1955 66

Tele-Tech and Electronic Industries 1955 Transistor Specification Chart September 1955 1-7

Turner 1954 Transistors Theory and Practice Gernsback Library Inc

Ward J 2003 Oral History of Neville Fletcher Early GermaniumPower Transistor Development