There is a huge body of information on updating and modifying the Dynaco PAS-3x preamp, including kits to replace all the circuitry. You may have decided a less extensive rebuild suited your needs, or maybe just replaced degraded power supply capacitors. One thing that's often suggested is replacing the carbon composition resistors with metal films, but the benefits are rarely demonstrated in a clear manner.
I recently made this change to a properly functioning PAS-3x and was astounded at the improvement in noise level. The following sound file contains 5 seconds of line stage noise with the carbon composition resistors, followed by 5 seconds of noise after replacement with RN60D metal films. It's recorded at a fairly high level so you can appreciate the change, but the gain and all recording conditions were identical; I just switched between a modified and unmodified channel. The two channels had matching noise levels before and after modification, i.e., both improved by the same amount.
44.1 kHz 16-bit .wav file of preamp line stage noise, about 2 MB
In actual listening at normal distances and levels the unit went from having a constant (and annoying) level of background hiss, to completely silent. There is a trace of hum present due to ground loops in Dynaco's somewhat poorly thought out ground scheme. This can be remedied with a bit of rewiring, but the right channel will usually have some excess hum when compared to the left. This appears to be due to proximity to the transformer and AC line wiring, though there may be yet more grounding issues involved. When examined with a spectrum analyzer, both channels have the same level of 120 Hz contamination and harmonics, but the right channel also has a 60 Hz component not seen on the left, thus my belief that's it's related to the AC mains and transformer.
Note that only a couple resistors were out of tolerance, though they were the plate resistors and had risen in value 20% or so. My guess is not every PAS-3x is quite this noisy, and there are probably only one or two resistors that are really critical. Given the desirability of replacing all the carbon composition parts, it wasn't worth the effort to determine the exact noise contribution of each resistor.
Interestingly, the line stage benefits the most from the resistor change. Dynaco had already used several precision resistors in the phono stage, and replacing all the phono resistors didn't yield any measurable benefit. It's still a good idea because the carbon composition parts will drift in value over time and the higher precision metal films will give more closely matched gain.
One other interesting but unrelated observation concerns the infamous Black Cat capacitors. I measured a few of these after replacing them with polypropylenes and was surprised at how good they were. The values were very close to nominal. The dissipation factor was the same as a Mylar. Based on various web pages, I think they ARE Mylar! The DC leakage was less than 2 uA at full rated voltage. The capacitance didn't change with voltage, unlike all electrolytics. My conclusion is that if your Black Cats are in good\ shape, and you're happy with the sound, let them be. As long as they don't develop some problem like DC leakage, they're a much higher quality capacitor than the detractors would lead you to believe. I would not, however, extrapolate that to other similar looking caps like Black Beauties and Bumble Bees.
last edit January 23, 2014