Gerald B. Herzog

Oral History Jerry Herzog


Oral History Jerry Herzog



One of the things that we desperately needed was some instrumentation that could look at nanosecond signals. I discovered some literature by a gentleman by the name of Sugarman on sampling oscilloscopes and proceeded to build a transistorized version that was a plug-in for a Tektronix oscilloscope. This unit was, I believe, the first of its kind to use avalanche transistors to generate a very short pulse. The design was copied widely throughout RCA divisions working on the contract and eventually copied, in a modified form, by Tektronix. However I warned them that avalanche transistors were in short supply that would deliver the necessary short pulse. Subsequently Tektronix had to abandon the approach and build the unit with vacuum tube circuits to generate the short pulses.


Shortly thereafter the tunnel diode was discovered and the agency handling our contract recommended that we look into using tunnel diodes for the high speed circuitry that was needed. Ed Herrold pointed out that two terminal negative resistance devices were hard to work with and had never been successful in the past when trying to be used for practical applications. However the agency was adamant that we work with what was then called the tunnel diode balanced pair, proposed by Goto of Japan. I was not enthusiastic about the approach and when I was offered a European fellowship by the Laboratories, I spent my time at Cambridge University doing an analysis on their computer on the so-called balanced pair of tunnel diodes.






I showed that there was an inherent instability in the circuits, and this was confirmed by experiments back at the laboratory that showed that the circuit would go into uncontrolled oscillation under certain conditions. The approach was eventually abandoned in favor of working with transistors. The agency had specified early in the contract that we would not be allowed to use transistors in our approach to high speed circuitry because other people in the field would be pursuing that line.



My Role in The Development of CMOS.


When I returned from England in 1961, I wanted to work on transistor or semiconductor memories. However, I was part of the computer research organization run by Dr. Rajchman, the co-inventor of ferrite memories, and as he pointed out, semiconductor memories could never be as cheap as memories made from mud namely, ferrite material. However, I argued that people were interested in associative memories in those days, which required some logic to be performed as part of the memory function, and only semiconductor memories could do that. So, therefore I started a group using thin film transistors, and also set up a computer aided design group so we could lay out unipolar transistor arrays. This work continued until 1965, when the Air Force decided there should be an industry-wide effort on large scale integration.



Go To Herzog Oral History, Page 5



COPYRIGHT 2004 by Jack Ward. All Rights Reserved. http://www.transistormuseum.com/