I was asked by Modemhead if I had a copy of the original Electronics Magazine article about the 8020. I did a little digging and found a copy. So this scan is a copy of a copy, but the text is readable.
My original boss at Fluke was Stan Jones, who was the project manager of the 8020. Also Norm Strong had a huge influence on me and taught me much. Although known as someone who occasionally took his lunch in liquid form, Norm was a certified genius. The article detailed the chip development conducted with Intersil. Two chips came out of that effort, the ubiquitous Fluke 429100 which was used for many years in 3 1/2 digit meters at Fluke. At the same time, a chip was developed for driving LEDs, a few of which I have in my parts drawer. However, the LED chip was never used.
The ICL7106 and 7107 were released by Intersil about a year after the 8020 was announced. What Intersil did to get around exclusivity of the Fluke chips was to mask off the internal switch that allowed the Fluke chip to switch between 200mVFS and 2VFS. It's this internal switch that prevents a 7106 from being used as a sub for the 429100 chip. As I mentioned in an earlier post, when Fluke got a hold of the 7106, it still had the Fluke Logo embedded in the Silicon. I wasn't involved with the lawsuit later filed, but it was settled out of court because frankly, Fluke could not afford to lose their relationship with Intersil. However, this did inspire Fluke to put in their own Silicon Fab at the new factory in Everett WA. My project, the 8060, was the first recipient of this, resulting in the 8060's MAC.
Another important Fluke partner was Caddock Electronics. They came up with the precision input divider in their very precise Thick-film technology. I later leaned heavily on Rich Caddock himself for design help during the design of the 8060, which had even higher precision thickfilm networks in it. Thickfilm was preferred over thinfilm as, at least at the time, thinfilm resistors were far less robust for transient spikes.
My personal involvement with the 8020 consisted of working at the separate manufacturing plant in Everett training the cal and repair techs. I probably fixed about a 1000 8020s myself. I remember that a big early problem was that many of the polypropylene caps used in the integrator were exhibiting too much dielectric absorption. This would cause what we called turnover error. That is, for example, +1.000V would read -.998 volts just by "turning it over", as in flipping the calibrator voltage polarity. So if your old meter doesn't read the same negative as positive, it's most likely a cap problem (as long as you have made sure the board is really clean). The other major source of DMM error is leakage caused by contamination. So earlier posts that recommend an IPA scrub are spot on. Any old electrolytic caps should be replaced, but be careful to clean thoroughly before and after soldering. I spent many hours testing 8020s and derivatives in humidity chambers. The largest contributor to failure due to humidity / temperature was always board contamination. Fluke spent a fortune on clean water systems for board scrubbing.
Always happy to answer questions if I can remember. This was over 30 years ago. Stan Jones died too young in the 90s. I lost contact with Norm Strong, I hope he's still kicking. I have happy memories of all us engineers going out at break and launching model rockets. An old Fluke story for Amphour.