Mister Transistor's Historic Semiconductors

The British Thomson-Houston company Limited was yet another of those classic British electrical manufacturers with a long and labyrinthine history. It is perhaps most famous for having built the world's first prototype jet engine for Sir Frank Whittle in 1937. The company was created in 1896 with funding from GE in the USA, and thus was in some ways similar in its origins to Standard Telephones and Cables Ltd.

However, in 1926 BTH became one of several (potentially competing) UK electrical companies to fall under the umbrella of a holding company called Associated Electrical Industries Ltd (AEI), which went on to buy further UK industrials. This unusual situation persisted for some decades, but in 1958 the BTH name was replaced by AEI on their products. The unusual nature of AEI and the general decline in heavy industrials led it into financial difficulties and in 1967, AEI was absorbed into the still-expanding GEC and disappeared as a brand.

AEI logo

Transistors and other semiconductors were a relatively minor part of BTH/AEI's product line, however the company was certainly a middle-rank manufacturer by UK standards. However, information about this aspect of their industrial history is scarce. The change of name from BTH to AEI occurred just after the company started manufacturing germanium transistors, and so both sets of initials can be found on early devices. (They always painted their transistors, usually in a bright colour, but never printed their logo on them, although many of their diodes bore it). The company's main industrial plant was based in Rugby, but their semiconductors were made in Lincoln. If you know anything about this, or have other information about BTH/AEI's semiconductor history, please

I do not possess all the transistors below, in fact I am seeking many of them. It is noted in the text where I am seeking examples of any particular type: if you have some for sale or exchange, please Conversely, I am happy to help anyone looking for information on these devices. I have an extensive collection of original data sheets and books, and can provide some information on most BTH/AEI types. Technical characteristics are also available on Maurice Woodhead's excellent Vintage Radio & Electronics site.

The first reference to a transistor made by BTH is rather mysterious. In his 1954 book, "Crystal rectifiers and Transistors", the author E.Molloy and consulting editor M.G. Say show a cutaway coaxial transistor with the point contacts at both sides just below the apex of a triangular crystal. This is a rather unusual form of the point-contact transistor, one that was never commercialised, but did explain some physics of the operation of point-contacts. Molloy states that his image (which looks more like a model than a real device) shows an experimental transistor which B.T.H. "originated in 1948". My image is an enlarged scan of the centre of the device: the printing dots of the book can be clearly seen. Molloy goes on to state that the germanium die is about 1/16 of an inch square. The next reference that I have found to BTH transistors is to junction types in Betteridge's book published in 1954. What happened in the intervening six years?

Two correspondents have come forward to shed light on BTH's important role in the early semiconductor industry. Harry Metcalfe reveals an exciting connection between BTH and Bell Laboratories' work on germanium.

A BTH apprentice at the time, Richard Groom, has provided me with a fascinating account of what he believes happened.

These two accounts seem to agree that BTH did not see any future in the point-contact transistor, and simply stopped work on transistors in 1950. Later, when others had proved the viability of the junction transistor, they resumed work on those. In view of this, the further statement in Molloy's book that "some thousands" of this co-axial type were made seems most improbable. However, I would be extremely interested in any surviving specimens, so if you know of any, please I am also very keen to tie up the the two histories recounted by Metcalfe and Groom, so if you can shed any light on them, do .

The next reference to the earliest BTH transistors is also mysterious. In the 1954 book "Transistors and Crystal Diodes", B.R.Bettridge refers to three BTH transistors, JT1A to JT3A. These are also listed in the 1959 book "British Transistor Manual" by E.N.Bradley, in the section on obsolete types. I have no further information on these devices and I am very keen to obtain examples of them. If you know anything at all about these transistors, or BTH's early semiconductor research, please

JT4X transistor

I do have one BTH transistor type in the JT series (BTH did not use this prefix for their commercial devices, switching instead to the GT prefix). The image shows a power transistor marked "BTH RES LAB" and "J.T.X.4/C". This is clearly a research device, although strangely I have come across two of them from completely different sources. The number 4 would seem to imply that this device follows on from JT3, although the extra X almost certainly means 'experimental', and the suffix is C rather than A. Furthermore this is the only BTH power type known to me: they made no commercial power transistors. Again, if you can explain any of this, please

The low-power AF/IF types GT1 to GT3 are described in Bradley's book (1959), and the evidence suggests that these were BTH's first commercial types. Perhaps they were modified versions of the JT1A to JT3A. As with all early germanium junction types, they had a rather low collector voltage and power dissipation, 9 volts and 125 milliwatts respectively. However, they were unusual in having a relatively high cut-off frequency of 800 kHz to 1MHz, thus allowing them to be characterised as AF/IF and low-frequency switching types. The three types are distinguished by increasing gain and cut-off frequency from GT1 to GT3.

Also found in Bradley's book are the trio GT11 to GT13 which employ the same small can as GT1 to GT3, and have the same voltage and power limits, but have significantly higher cut-off frequencies of 3 MHz, 5 MHz and 7 MHz respectively. BTH characterised them as suitable for "intermediate and high speed switching applications. Also used as a self-oscillating frequency changer and i.f. amplifier".

Interestingly, BTH transistors, presumably out-of-spec examples, seem to have been made available cheaply to the home construction market. These "red spot" devices were unmarked and unpainted and were normally used at audio frequencies in amateur projects. They are clearly BTH types, possibly rejected GT1 to GT3.

Information about the three types GT31 to GT33 is quite difficult to come by, as they do not appear in most transistor data books from the late 1950s. However, they are described in the 1960 "British Semiconductor Guide" as a trio of low-power AF and general-purpose devices, with increasing gain from GT31 to GT33. Interestingly, the booklet shows them as made by AEI, but those in my image are still branded BTH. The fact that they are green, as are GT1 to GT3, suggests that this colour means "audio frequency", while light blue means "intermediate frequency". If you have any of these for sale or exchange, please

The four types GT40 to GT43, although apparently a series ordered by cut-off frequency, and hence gain, were in fact released as the trio GT41 to GT43 in about 1960, with the lowest-performing type GT40 being added in about 1961. They are IF amplification and high speed switching types, and BTH/AEI seem to have specialised in this higher-frequency regime, used mainly in signalling and computing applications, rather than the technologically easier audio frequency market. They have quite a low maximum collector voltage of 9 volts DC, 15 volts peak.

(I do have a number of GT42's, which are still soldered on circuit cards, and hence I do not show one here. They are in identical pale blue-painted cans).

The GT4x transistors also come in a slightly larger encapsulation as shown here, identifiable by a -B suffix on the type number. Strangely, no reference to this is made in any of the AEI data booklets that I possess. Only AEI branded examples seem to have this form, which presumably allowed a higher power dissipation. If anyone can provide documentation on this larger form, please

The types GT44 to GT47 have the same gain and frequently limits as GT40 to GT43, but with an increased maximum collector voltage of 25 volts (DC and peak). As my image of the GT45 shows, it is likely that all of these also came in the 'B' variant. If you have any of these types for sale or exchange, please

My extensive library of data books and sheets fails me when it comes to this device, marked AEI GT50, and the larger GT50-B, of which I have several examples on a circuit board. Does the yellow colour indicate a different application from the above types, switching perhaps? If anyone has data on this transistor, please

This little transistor is an oddity because it is branded Hivac, another UK electrical company that specialised in thermionic valves (vacuum tubes). According to Bradley's "British Transistor Manual" of 1959, Hivac sold two transistors, this type XFT2 and the type TM1. However, it is quite clear that this device uses an identical encapsulation to the BTH types GT1-3 and GT11-13. Is it a re-branded BTH device? Hivac does not appear to have been one of the companies owned by AEI. If you have any Hivac transistor, or know anything about Hivac's semiconductors, please

Here are two more unusual transistor types, marked AEI TR1A and TR2A. What is really odd about them, apart from the fact that I cannot find them in any data book (and I have a *lot* of data books on germanium transistors) is that they use the black-painted 'top hat' encapsulation that is unique to early Ediswan transistors. Ediswan fell under the AEI umbrella, but why did AEI brand these two transistor types as their own?

I am grateful to John Hills of Oregon in the USA who has explained that these TR1A and TR2A transistors were supplied to Tri-ang (Lines Bros) for use in their Tri-onic kit sets. These sets were probably only produced from 1961 until 1963. John's 1961 "A" set contains two top hat transistors marked "A.E.I. TR1A" and "A.E.I. TR2A", exactly like mine. These kits contained components in plastic housings with connecting pins that plugged into a series of circuit boards, Tri-ang designated the "A" set transistors as TR4 and TR5. TR4 is the TR1A and is described as being the RF transistor, which John believes to be an Ediswan XA101 or XA102. TR5 is the TR2A and is described as being the AF transistor, which John thinks may be an Ediswan XB102, XB103 or possibly XB104. John has other Tri-onic kits and some have transistors marked "A.E.I. TR1" and "TR2" but in metal cans with plastic sleeves that are probably later Ediswan types. Even later kits contain Mullard OC series transistors. In 1962 Tri-ang introduced the kit "A/B", This added five more printed circuits and other components including another RF transistor in the TO7 package, probably a Mullard AF11n type.

These CG series types are point-contact diodes in glass capsules that are very irregular in shape and sometimes variable in length, suggesting some were hand made. They form a series from CG1-E to CG12-E (with gaps) and have a variety of colourful rubber or plastic sleeves. They are all general-purpose or detector diodes with various inverse voltage limits. They are described as obsolete types in the 1961 Wireless World Radio Valve Data book.

BTH also appears to have made the above CG series point-contact diodes in a larger format, in glass or ceramic bodies with metal end caps, and with the suffix C rather than E. However, I have no data on these types, so if you can provide any, please

BTH also made miniature diodes, also in glass bodies with coloured sleeves, with the suffix H in the data books, although it does not seem to be printed on the devices. The CG60H series were point-contact diodes for general-purpose use as detectors, modulators, frequency converters, pulse separators etc. The CG80H series were germanium gold-bonded diodes for computer and data processing applications.

The GJ series of germanium junction rectifier diodes used a metal body with a glass insulator around the anode. The series included GJ3-M to GJ6-M (M for metal?) with an apparently random selection of maximum reverse voltages, all with 1A maximum forward current. I also possess one example of GJ7-M, which is not included in any BTH/AEI data book. BTH also sold pre-wired assemblies of these diodes such as full-wave rectifiers.

BTH/AEI made a variety of other diodes and Zener diodes, both germanium and silicon, in a variety of shapes as shown here.

BTH/AEI Semiconductors