Are you still using your HP 334A? I just spent a few hours refurbishing a unit manufactured around 1972, and I suspect yours is about the same vintage. According to Change #14 in the available Backdating Information, voltage limits and regulation values for A3TP2 were changed from -20.2 V ± 0.3 V to -19.5 V ± 0.5 V on subsequent production runs. Mine still uses A3R37 = 47 kΩ so I had to stick to the older values (-20.2 V ± 0.3 V). Newer production units use A3R37 = 330 Ω and the updated bias information (-19.5 V ± 0.5 V).
I had some serious problems during some bias adjustments until I found out that there were some very leaky electrolytics capacitors upsetting the bias balance in this A3 board. A2 board also had some defective electrolytics but they were basically open so both failure modes (open and short) are expected to be found in such an old piece of equipment. on the other hand, it is surprising that only a few of them have failed, and the remaining still show good (low) ESR as I tested some of them on a sample basis (100% on the A2 board which is the most critical, I think).
If you still interested in such a followup, let me ask a couple of questions regarding the noise level as I am unable to reach anywhere close to the nominal 80 dB fundamental rejection. It looks that the equipment is somewhat plagued by extraneous noise (both mains frequency and high frequency harsh) through unexpected paths. One of them is the shafts (they are concentric) of S2 and S2R8 (Sensitivity and Sensitivity Vernier) which are seemingly purposedly left floating but unlike S1 (Function) and S3 (Meter Range) which have a insulating composite shaft protuding to their knobs, S2 does use a metal shaft which conducts all the way through the wafers of the switch, coupling the noise from your hand directly into where it shouldn't be. This can only be felt during low distortion measurements, of course. I would like to know if your HP 334A also has this drawback.
The second one is even more unexpected. The sliding top cover of my HP 334A is often failing to proper make electrical contact with the chassis, causing rather intense hum (mains, in this case) to appear during low distortion measurements if you touch the conductive the aluminum strip on the front of it. Pressing it agains the chassis may fix the problem for a while, but it is sort of loose so it is often loosing contact again and again. I guess I would have to use some kind of gasket or conductive tape to fix that, but those are hard to find in my country and I wonder if anybody else has ever had such an unexpected problem with a HP cabinet of this vintage.
Any returns will be most appreciated.
Flavio