I had a unit which had a problem with the DC voltage reading not settling. It turned out to be a defective opamp on the voltage reference (U400 which was and Analog Devices AD706). I might not have noticed it if I wasn't comparing it to another 34401A and doing some data logging. The unit passed all self tests. Dr. Frank on the forum helped diagnose the problem.
https://www.eevblog.com/forum/testgear/hp-34401a-dmm-settling-time-comparison-questions/
fwiw, I do have a keithley 2015thd (similar to 2000) and I like it a lot, but I also knew it was not fixable, for the most part, other than swapping out the bad caps (which I still have yet to do..).
most of the comments here are favorable to the hp unit and so maybe I'll try to prefer that one for the next time around. hp does not always make great stuff, but it sounds like this was a reasonable design and implementation, which is great to hear.
(one thing the keithley does have that the hp does not is a thermocouple mode. I am doing some temperature sensor stuff these days and when I connected my fluke banana-plug based tcouple into the keithley I was surprised to see so many digits after the decimal point on that device. a nice extra feature to have, to be sure.)
and only one custom chip
Not quite there is the TF-245 resistor network, both my 2k's had that one blow
thanks
ketil
Talk was about 2001, which dont have TF245
.