Russ Theisen solved a major problem at IBM regarding Computer Technology IBM 1964 [The bad tunnel diodes caper]



Russell E. Theisen LSM IEEE



One of my first jobs after graduating with a BSEE was with IBM in Endicott, N.Y.



I was hired to help build the first Solid Logic Technology IBM 360 computer.



There were many task that I was able to help with, but one of my jobs was performing interface between hardware and software programmers, And Quality inspectors and hardware designers.



We had a large order of tunnel diodes that our incoming quality inspection was rejecting. These diodes were 15ma mili-amps devices that cost well over $40 dollars each at the time.



This type of problem would be costing months of delay in building a fast switching device to test the high speed logic modules.



I asked to speak to the Quality Inspector that had rejected the whole batch of them.



I asked him, how are you testing these diodes, that you rejected.



He showed me by using a Triplet 630 NA multi meter where he would check the front to back resistance of the diode.



It was obvious what the problem had been. And I yelled stop what you are doing. You are destroying these 15 ma tunnel diodes by you testing.



He assured me that he was using the “Clock thing” he called his multimeter, to measure the front to back ratio just as his procedure instructed for testing diodes.



I asked him what is the specification of the tunnel diode that you are testing?



He said he did not know.



I told him that the Triplett 630NA multi-meter on the RX1 scale was causing over 150 ma to flow through the tunnel diode each time he measured it with his meter. And that was over 10 times the current that the tunnel diode was designed to handle. He had just blown over 2 dozen tunnel diodes just like a blown fuse.



I got his supervisor and told him the problem and the solution was to make sure the range multiplier used was to be set at RX100 or more to make the front to back measurement. That was only 1/100 the current placed on the tunnel diode and that setting would not blow the diodes.



We were able to get the Quality Acceptance test procedure corrected to allow testing tunnel diodes with out blowing them.



I found that IBM people were just learning about transistors and semiconductors because their experience had been with tubes and relays and mechanical machines.



We all had learned many new lessons while we developed this new technology as we went.