13 Mar
2019
13 Mar
'19
6:44 a.m.
On Mar 13, 2019, at 12:02 AM, Chuck Guzis via cctalk
<cctalk at classiccmp.org> wrote:
...
This is a bit interesting in that Brattain, Bardeen and Shockley are
credited in the popular press as having invented the transistor.
However, that was a bit overstated; they had to re-word their patent
application to state that they'd developed a "junction" transistor, when
a patent search turned up the fact that a Hungarian immigrant named
Julius Lilienfeld had obtained a patent on a field-effect transistor in
1930--a full year before he obtained a patent on the electrolytic
capacitor (ever heard of those?. Dr. J applied for the patent in 1926,
which is a bit mind-boggling, when you consider that tubes like the
UV20A1 were introduced in 1924. It's those field-effect transistors
that are widely used today, not Shockley and chums' bipolar cousins.
My father had an article about those FETs in his files somewhere, unfortunately it was
lost years ago but I remember it. I think the semiconductor used was copper oxide.
Interesting that they actually found this. There are many examples of the same thing
being patented several times, or a thing being invented long after it was first built. An
example of the former is frequency modulation (patented in 1927 by Idzerda, then in 1935
by Armstrong). An example of the latter is Abraham Lincoln's patent for what a
Dutchman would recognize as a "camel" -- a device for carrying ships over shoals
that goes back a century or two from Lincoln's patent.
Such is history and those who write accounts of it.
Names like
Atanasoff and Zuse are consigned to the dustbin of history, while Eckert
and Mauchly get the historical mention.
That's what makes it interesting to dig into the less known corners of technology
history. I've enjoyed poking into Electrologica, which did a number of things early
on, perhaps earlier than others that are well known or at least around the same time. The
core ROM I mentioned is an example (it looks vaguely like Ken Olsen's design that
became the Apollo Computer ROM, but the operation is different and somewhat more
efficient, with a shorter read latency). They also seem to have been the first to deliver
interrupts in a commercial computer as a standard feature. But because of location and
limited sales, few people even know the company's name, let alone much about it.
paul