I had asked Carl Nelson about a particular photo with a circuit layout on a light table, and he told me this story:
The circular structure looks a little like Repus Renez. That was Jim Williams' high frequency RMS voltmeter chip. He took my Super Zener chip, reversed the name, punted the Zener, then used the diodes and heater resistors to make the voltmeter.
Jim presented a paper on the circuit at ISSCC, and I think the Repus Renez name could be seen on the chip in one of the slides. The product, LT1088 is now obsolete, but you can read about it in AN22.
We pulled one of the best pranks on Jim while he was testing the first parts. He kept getting higher and higher bandwidth as he cleaned up the test jig, and we got tired of his bragging every day about some new number. So I waited until he left for the day, unsoldered the test socket, and wired a tiny inductor in series with the input signal, hiding the inductor under the socket. That dropped his bandwidth from 100MHz down to 50MHz or so. He spent the next day tearing his hair out (he literally used to twist his hair constantly) and cursing a lot. I finally took pity on him and suggested that maybe he had picked up a parasitic, but he ignored me. After several more hints, just before quitting time, I suggested in rather strong language that he double check the test socket very carefully. I think he was pissed, but he maintained his composure in keeping with his title of master prankster of them all…
That's not the LT1088 in the photo.
Photo credit goes to BusinessWeek magazine (Paul Margolies)
L-R - Bob Swanson, Bob Dobkin, Carl Nelson, Tom Redfern