Bell Labs (Who Really Invented The Transistor?)

Who really invented the Transistor?

Other claims to the invention...

This article first appeared in Radio Bygones magazine, www.radiobygones.com
Reprinted here by permission from the author, Andrew Emmerson

Andrew Emmerson uncovers conflicting claims and some revisionist history.

Before the fall of the Soviet Union the state educators of the old USSR were kept busy rewriting history, either deleting from the roll of honour all reference to heroes of the people now fallen from grace or ascribing the credit for every modern miracle to obscure communist pioneers.

This time, however, it’s the Americans under fire for falsifying history and the subject is the invention of the transistor. The received wisdom is that William Shockley, John Bardeen, and Walter Brattain invented this device in 1947 and of that there can surely be no doubt. But there is and the colourful claims and counterclaims make some fascinating reading.

One fact is not in dispute, that the achievement of Shockley, Bardeen and Brattain was responsible for kick-starting the solid-state electronics revolution and the age of computerised informatics. To decry their role in transforming electronics would be both churlish and crazy, but the claim that they pioneered solid-state amplification has no substance at all.

Received version

Before we go back to the dark ages, let’s examine the standard version of transistor history, courtesy of Andrew Wylie, who has set up an excellent website devoted to early transistor devices. He states:

The transistor was invented at Bell Laboratories in December 1947 (not in 1948 as is often stated) by John Bardeen and Walter Brattain. 'Discovered' would be a better word, for although they were seeking a solid-state equivalent to the vacuum tube, it was found accidentally during the investigation of the surface states around a diode point-contact. The first transistors were therefore of the point-contact type. William Shockley, the theorist who was leading the research, knew at once that this was not what he was seeking: at the time he was trying to create a solid-state device similar to what we now call a junction field-effect transistor.

Bell Labs kept their discovery quiet until June 1948 (hence the confusion about the date of discovery). They then announced it in a fanfare of publicity, but few people realised its significance, and it did not even make the front page of the newspapers. Shockley basically ignored the point-contact transistor, and continued his research in other directions. He modified his original ideas and developed the theory of the junction transistor. In July 1951, Bell announced the creation of such a device. In September 1951 Bell held a transistor symposium, and licensed their technology for both types of transistor to anyone who paid the required fee of 25 thousand dollars. This was the start of the transistor industry that has changed the way that we live, in the Western world at least.

Alien efforts

However, an entirely different origin has been proposed by Jack Shulman, president of the American Computer Company. Frankly, his theory is pretty fantastic but it makes a rattling good read if nothing else. Here's what he says...

I grew up in the household of the head of Bell Labs, so I knew that there was something strange about the transistor because I knew Bill Shockley, and Bill Shockley was something of a witless buffoon. There's no way he could have invented the transistor.

The symbol for the transistor is made up of three pieces: positive, positive and negative; or negative, negative and positive...silicon dioxide doped with arsenic and boron, in 1947. Now, in 1947, doping things with boron was not easy. It required the sort of equipment that even Bell Labs in 1946 did not possess. They had this type of equipment at Lawrence Berkeley Laboratories, but it would have taken thousands and thousands and thousands of man-hours to invent the transistor.

If you look back at it historically, what AT&T was claiming was that one day this "genius", William Shockley, was working with a rectifier; he looked at it and he noticed it had unusual propensities, and there, bingo, he invented the transistor! He figured it out right there!

Anybody believe that story? Me neither. And I knew, because the administrative head of the transistor project was Jack Morton—the man at whose house I was staying to go to school and whose sons I was friends with. He often commented on the fact that it was really a shame that those three idiots got responsibility for the transistor and he didn't.

Mr Shulman goes on to claim that the transistor's real origin lies in technology recovered by the US Air Force from an alien spacecraft recovered at Roswell, New Mexico in 1947. It's extremely controversial stuff and contrary to all received wisdom—but quite amusing of you don't take it too seriously. Let’s move on rapidly, back down to earth and to minerals in particular.

Start of silicon

It was in 1906 that the G.W. Pickard of Amesbury, Massachusetts perfected the crystal detector and in November of that year took out a patent for the use of silicon in detectors. Arguably this was the start of the silicon revolution and it did not take long before experimenters achieved amplification using crystal devices, long before the term transistor was devised.

Solid-state electronics were born even earlier, when Ferdinand Braun invented a solid-state rectifier using a point contact based on lead sulphide in 1874. But it’s to Pickard that the credit goes for discovering that the point contact between a fine metallic wire (the so-called ‘cat’s whisker’) and the surface of certain crystalline materials (notably silicon) could rectify and demodulate high-frequency alternating currents, such as those produced by radio waves in a receiving antenna (what Pickard called a ‘wave-interceptor’). His crystal detector (point-contact rectifier) was the basis of countless crystal set radio receivers, a form of radio receiver that was popular until the crystal detector was superseded by the thermionic triode valve.

By its nature the crystal rectifier was a passive device, with no signal gain. But radio historian Lawrence A. Pizzella WR6K notes anecdotal stories of shipboard wireless operators in the second decade of the 20^th century achieving amplification using a silicon carbide (carborundum) crystal and two cat’s whiskers. He cites a taped interview made in 1975 with Russell Ohl at his home in Vista, California in which claims of signal gain were made. This is an excerpt from Ohl’s testimony:

He gave me a copy that he had of… I think it was The Electrician. It was a British magazine, one of these big-paged things, you know. In it was a translation from a Russian paper in which they had used carborundum with two contacts and a battery supplying one of the contacts and had gotten a power gain of ten times. And this was way back in the 1910s, so the fact that you could get a power gain had been known, but it was never put on a controlled basis. I knew about it because an operator of the Signal Corps back in 1919 had told me that some of the operators used carborundum as oscillators for receiving. When I had seen this article that Curtis gave me, I was not astounded because I had known about this before I ever saw the article. I had heard about it. I knew a former first sergeant in the Signal Corps who had lived in, the boarding house that I lived and he was an expert radio operator. He told me a great deal about the use of crystal detectors on ships. He told me that professional operators carried two crystal detectors with them. One of them was made of carborundum them and one of them was something like galena or something of that sort. He said the carborundum was used for two purposes. They used it in the harbour when they were close to a transmitter to prevent burnout. They also used it at long distances with two points. One point was excited with a battery and they were able to get long wave oscillations out of it and in that we were able to be in long wave telegraph stations.

Ohl, it should be noted, was the man who invented the silicon solar cell in 1941 and discovered during World War II that semiconductors could be doped with small amounts of impurities to create useful new properties. Born in 1889, he was bitten by the radio bug at the age of 16 and devoted much of his life to making simple radio receivers employing semiconductors. His accidental discovery of the P-N barrier in his work at Bell Telephone Laboratories led to the development of solar cells.

Oscillating crystals

A fascinating letter to Wireless World in May 1981 under this title came from Dr Harry E. Stockman of Sercolab (Arlington, Mass.) Then 76 years old, he had lived through the era under discussion and provided a valuable summary of ‘prior art’ preceding the re-invention of the transistor. His letter had been triggered by a ‘Sixty Years Ago’ item in the same periodical) recalling an article by W. T. Ditcham on crystal oscillation in its May 1920 issue.

This effect, he stated, was discovered by Dr W. H. Eccles in 1910, and remarked: “It is hard to realize that it took about ten years for practical active crystal-diode circuits to appear, in spite of Ditcham's reminder—circuits that included both RF and AF amplification. The last one, at the time, was totally unknown to most ‘affectionados’, one of them being the author of this letter. Most of the credit for creating practical devices [of this kind] goes to O. V. Lossev of Russia, whether or not he knew of Eccles' pioneer work a decade earlier. He should have known about it; one has the right to expect that he as a qualified scientist was familiar with the world's scientific literature.”

Clarification comes from Lawrence Pizzella, who explains how these experimenters created successful amplification techniques using mineral crystal devices. Lossev, he says, used zincite and a steel cat’s whisker with bias to make an oscillator and even a low-power transmitter in the early 1920s. This was reported in considerable detail in the September 1924 issue of Radio News and in the 1^st and 8^th October 1924 issues of Wireless World. Hugo Gernsback, the editor of Radio News, named this the ‘Crystodyne’ and predicted that crystals would someday replace valves in electronics. All details needed to duplicate these circuits to make a tunnel diode oscillator are in these articles. A German book by Eugen Nesper described an oscillating detector circuit in 1925 too, using zincite material and a bias voltage of 8 to 14 volts.

With so much information in print it’s inconceivable that the Bell Labs team were unaware of these techniques. But in any case Pizzella says Russell Ohl showed William Shockley his radio using crystal amplifiers several years before the transistor’s alleged invention in 1947. Shockley is also quoted (in Crystal Fire by Riordan and Hoddeson) as saying that seeing Ohl’s radio convinced him that an amplifying crystal could be made.

First FET

Another experimenter of this era who deserves far greater credit is Dr Julius Lilienfeld of Germany, who in 1926 patented the concept of a field effect transistor (FET). He believed that applying a voltage to a poorly conducting material would change its conductivity and thereby achieve amplification. Lilienfeld is rightly noted for his work on the electrolytic capacitor fame but according to Stockman should be recognised also for his pioneering work on semiconductors.

Says Stockman, himself a distinguished author of many books and papers on semiconductor physics, “He created his non-tube device around 1923, with one foot in Canada and the other in the USA, and the date of his Canadian patent application was October 1925. Later American patents followed, which should have been well known to the Bell Labs patent office. Lilienfeld demonstrated his remarkable tubeless radio receiver on many occasions, but God help a fellow who at that time threatened the reign of the tube.”

David Topham GM3WKB adds that Lilienfeld followed his 1925 (Canadian) and 1926 (American) patent applications for a ‘Method and Apparatus for controlling Electric Currents’ with another granted in 1933. Says David: “US patent 1,900,018 clearly describes the field effect transistor, constructing it using thin film deposition techniques and using dimensions that became normal when the metal oxide FET was indeed manufactured in quantity well over 30 years later. The patent (and subsequent ones) describes the advantages of the device over ‘cumbersome vacuum tubes’.”

More prior art

The website of Dr Robert G. Adams states that he designed a crystal amplifier at the age of thirteen years, when he lived at Hastings, New Zealand. A photograph of his set-up is shown on his website along with the diagram reproduced here with acknowledgment.

Connections to the two crystals made use of the then-available vertical cantilever type cat’s whisker holders, providing stable connections to the central junction and input and output points. Two different methods of interconnection between the two crystals gave no apparent difference in performance. Adams stresses that it never occurred to him to pursue any patent action simply because the invention was already in the public domain. In his view it was obviously unpatentable by anybody (Bell Labs notwithstanding).

Someone who built a similar amplifier of this kind is Canadian radio amateur Larry Kayser (va3lk / wa3zia), who spotted a circuit for a ‘novel’ crystal radio circuit that exhibited ‘amplification’ published in Gernsback’s magazine Radio during the 1932-1934 period. This, he recalls, used two cat’s whisker probes on a lead-mounted galena (PbS) detector. He says he was able to duplicate this action in the early 1950s as a young hobbyist and whilst the degree of amplification was nothing like that of the first commercial transistors, it was at least in the order of 3dB or a bit more.

History repeated

That was then but this is now. American radio amateur Nyle Steiner K7NS was determined to prove or disprove these claims for himself—and has succeeded in spectacular fashion. On his website he posts technical results, photographs and curve traces of several experiments in which he has demonstrably achieved oscillation with iron pyrites and even transmitted his voice over the air (a circuit for a broadcast band iron pyrites negative resistance oscillator is given there).

“Success with this experiment has been a very exciting experience for me as it represents the ability to build a simple homemade active semiconductor device. It is almost like making your own homemade transistor,” he states. “This is an actual realisation of some very old, and esoteric 1920s experiments by Eccles, Pickard and Lossev, that were so vaguely reported in a few articles that I have often wondered if in fact it had actually been done. Even so, I have always had an extreme fascination with those reports of being able to produce a continuous wave RF signal from a crude semiconductor material back in the very early days of radio.”

Other experiments of his show an oscillator based on zinc ferrite and an N-type negative resistance device, similar to a tunnel diode, created by touching a piece of galvanized steel wire against a piece of aluminium. As Nyle says, “This project may not be very practical but I find it to be a very exciting experience.

Historic conclusion

The more you study the history of invention, the fewer example you find of entirely new devices conceived and perfected by one individual in isolation. History loves heroes and people prefer simple stories, regardless of inconvenient facts.

It’s perfectly clear that Bell Labs didn’t invent the transistor, they re-invented it. The fact that they totally failed to acknowledged the pioneer work done by others can be explained by human nature—pride, arrogance, ignorance or plain self-interest. It’s perfectly true that the world wasn’t ready for previous incarnations of the transistor but that was no reason for denying that Lilienfeld patented the original solid-state triode oscillator/amplifier well before others claimed all the credit. But that’s life; it was not the first time and doubtless not the last.

Resources on the WWW

Lyle Steiner’s amazing experiments

http://www.sparkbangbuzz.com/els/iposc-el.htm

Fringe science?

http://www.nuenergy.org/

The Transistor Museum™

http://semiconductormuseum.com/Museum_Index.htm

Four point-contact transistor manufacturers

With the Bell Laboratories Invention of the transistor, and its manufacturing arm Western Electric producing them, here is a list of both Western Electric and other four point-contact transistor manufacturers.

Western Electric: 2N21, 2N23, 2N25, 2N26, 2N67, 2N110
RCA: 2N32, 2N33
CBS-HYTRON: PT-2A
General Electric: G11
Raytheon: CK703, CK716
Sylvania: GT-372 and later copied RCA’s 2N32

Pickard’s patented crystal detector of 1906 kick-started the silicon revolution.

The crystal amplifier devised by New Zealander Dr Robert Adams in 1933, showing how terminology has changed over the years.

We Offer Personalized One-On-One Service!

Call Us Today at (651) 787-DIAL (3425)