Power Design 2001a, Ripple and Accuracy

[rwgast_lowlevellogicdesin]

I have been rebuilding a PD2001A, I just rebuilt the heater, and then the reference board with new caps and low noise transistors. Ive also replace 90% of the caps on the amplifier and AC input section. The supply is working very well but I have a few questions for anyone with one of these supply's about its noise an precision.

First of all the PSU puts out a very nice signal most of the time, I have measured it with a siglent DSO, and an old HP scope with 100uV resolution and differential inputs. I seem to be in the .800uV to 1.5mV range when it comes to ripple, loaded and unloaded. When I set the current regulation everything is fine, but if the load attached to the PSU exceeds the current limit the ripple gets much worse. It jumps from 800uV to about 8-10mV with about a 120hz sine wave pattern. Is this common, what is the best way to resolve it?

Second the accuracy and calibration is dead on perfect (for the most part) from what I can tell by comparing it to a dmm check and with a few 20,000ct meters. The only accuracy problem I have is with the volts selector. Every time I add a volt it adds 1V plus a few mV, here is what my readings are for each of the 10 voltage selections
0.000
1.000
2.001
3.002
4.002
5.003
6.004
7.00  can't measure mV past 7V with DMM
8.01
9.01

So as you can see the mV roll up with the voltage selection, so im guessing whatever is going on is needs more than just trimming one of the resistors. Should I attempt to add a high value trim pot in parallel with each resistor on the voltage knob?

[dacman]

If the 2001A accuracy is +/- (0.1% + 1 mV) then the meter does not have the precision (nor accuracy) for the check.

[David Hess]

I do not have a manual for the 2001A but if it is like the 2020, then the current limit is a Vbe design and more for safety than precision.