Last time I was at the bench I was working on the DM502. Went through the manual's calibration and test procedure and all was fine until I got to the current measurement tests. They were all wrong but sort of randomly so. The data didn't make sense, so I figured I was doing something wrong and stopped for the night. Then my back injury flared up, keeping me from the bench for a week.
The 502 is a simple instrument. There's one signal path with different front end sections for AC, DC, and resistance measurement, including separate attenuators for current and voltage. I thought maybe the attenuator was damaged, so I checked both. Not the problem. The path on the other side of the attenuators seemed to be working fine, too. Puzzling.
Turns out the problem was with my current generator setup and how it interacted with the DM502. I don't have a proper current calibrator, so I figured I would use my precision voltage source to make one. Set it up, put my 8840A in line, adjust the current, take the 8840A out and put the DM502 in its place.
Problem one: the DVC8500 doesn't do a good imitation of a current source. Problem two: the DM502 current meter resistance is range dependent and varies from 0.1 to 2000 ohms. So the effect of the DM502 resistance in my circuit varied a lot and was quite significant in the low current ranges.
I made some changes to my "current source" setup to take this into account and the current measurements starting making sense. PEBMAD problem solved.
I still need to figure out how to cobble together a AC current source. No good ideas yet, at least with the equipment I have at hand. So that function and the decibel conversion circuit will be checked tomorrow. After that's done, I will start on one of the three units that definitely isn't working.