A really great dive into the history and players leading to where we are today. He was on the Odd Lots podcast in late October to talk about the latest legislation, that's also worth a listen.
If nothing else this article appears to explain to me why the control point is called the "base": in the original point-contact transistors the emitter and collector poked into the top; the control was on the bottom -- the base -- of the germanium slab.
This is weird when you see the canonical transistor schematic symbol, but if you rotate it 90º anticlockwise the base will be at the base of the symbol and make sense.
Perhaps everybody else knows this already, but it was aha! to me.
The symbol and the names both make more sense for point contact transistors than BJTs. Like the article says, I had never heard of point contact transistors' different operating principle.
FETs have a schematic symbol that describes their operating principle pretty well - they have never had a shift in meaning.
The names continued to make sense for the alloyed transistors (most of the germanium transistors were alloyed) and for some of the earlier diffused silicon power transistors, the so-called single-diffused transistors, of which the most well known was the RCA 2N3055, which was used by many ancient audio amplifiers.
In those alloyed or single-diffused transistors, the fabrication also started with a semiconductor crystal that was the base of the transistor, in which emitter and collector electrodes were created, but by alloying or by diffusion, instead of point contacts.
Only when the double diffused transistors, especially in their planar version invented at Fairchild, became the dominant kind of bipolar transistors (after 1960), the name "base" for the control terminal became no longer appropriate.
While "gate" was inherited by the FETs from the vacuum tubes and gas tubes, the names "source" and "drain" mean exactly the same thing as "emitter" and "collector", so there was no reason for introducing these 2 alternative terms.
>The first patent for the field-effect transistor principle was filed in Canada by Austrian-Hungarian physicist Julius Edgar Lilienfeld on October 22, 1925, but Lilienfeld published no research articles about his devices, and his work was ignored by industry. In 1934 German physicist Dr. Oskar Heil patented another field-effect transistor. There is no direct evidence that these devices were built, but later work in the 1990s show that one of Lilienfeld's designs worked as described and gave substantial gain. Legal papers from the Bell Labs patent show that William Shockley and a co-worker at Bell Labs, Gerald Pearson, had built operational versions from Lilienfeld's patents, yet they never referenced this work in any of their later research papers or historical articles.
All the early experiments with semiconductor devices have been plagued by irreproducibility, because it was not understood how much even minute amounts of impurities or crystal defects can affect the results.
The reason why the transistor and the other semiconductor devices have been discovered quickly after WW2 was that during the war there has been a huge effort to improve the purification and the growing of semiconductor crystals like germanium and silicon, with the purpose of making large quantities of detection diodes for radars.
At the high frequencies used by radars, the devices used in low frequency radio receivers, e.g. vacuum diodes or improvised point-contact diodes, such as those made on galena, were ineffective.
After the war, the research on semiconductor devices could use the newly available high-quality crystals, on which reproducible devices could be made, so the invention of the transistor became an unavoidable consequence.
Like the nuclear industry, the semiconductor industry is also a result of the large amount of money spent in USA during WW2 for improving manufacturing techniques.
With normal corporate spending that is limited by short-term profits, becoming able to make reproducible transistors might have needed many decades, if ever happening. Making pure semiconductor crystals is not something for which a lone inventor could afford to build equipment.
A really great dive into the history and players leading to where we are today. He was on the Odd Lots podcast in late October to talk about the latest legislation, that's also worth a listen.