the standby switch

[R.G.]

I've heard people say that the thumps in SS amps are caused by asymmetric collapse of the PSU rails. If that's the case then fixing the asymmetry should cure the problem.

Thump was originally a problem with single-supply SS power amps, as they needed to charge the output capacitor when they came up, and that current came through the speakers. These were BAD thumps, speaker popping thumps. Asymmetry on the PSU rails is an issue, but you can get a similar effect by the innards of the power amp circuitry not responding symmetrically to the power supply ramps, or by there being thresholds in operation as the power rails come up. Ideally, the input and driver sections would be in operation and trying to hold the output at the reference voltage (ground, usually) before the outputs could abuse the speakers, but that doesn't always work.

Squeegggggs of oscillation can happen as poorly designed amps with marginal stability go through regions of lower and higher gain on their way to the operating point.

Most modern amps don't have much of this as an issue, but...

I have some amps that have never thumped for over a decade, but have only begun to thump now that they have gotten older. with these amps I've also noticed that the thumps seem to be worse when the amp is connected to a lower resistance load.

I would re-cap them. Something is preventing the input and drivers from keeping the output at ground as the power supply moves. Often this is something like a bootstrap capacitor, but a leaky feedback-network cap can do much the same. Also check to see if they now have a getting-worse DC offset on the outputs. Anything over about 10-50mV will be easily audible, and a bum feedback-net cap will do this. So can other things, but I'd start there.

On another front, circuits can slowly degrade bipolar transistors by breaking the base-emitter junction at power on, or, more likely, at power off. If the emitter is tied to a cap that gets charged to more than about 7V in operation, if the input is suddenly pulled the wrong direction by power off, the cap discharges through the base-emitter in reverse. This degrades the noise performance and gain over a period of time, even though the device keeps working. This is one reason old bipolar amps are often noisy. They may not have been that noisy when new, but time and transients get to them.

On Relays and MicroControllers
I know most tube-y people hate microcontrollers, but really guys, there is a set of microcontroller functions that can be thought of as "make this thingy do what I'd do if I could be in there and flipping switches and stuff Real Fast, all the time."

Yes, it's a pain. But I can decide to make a uC that will be powered from the AC line on what amounts to a few uA of "phantom power", and sleep until the power switch changes state. When it changes, the uC can sense which way it changed, then turn power on or off to the power transformer, manipulating speaker isolation relays and such on a millisecond (or even microsecond) level as it goes.

Yes, it's nice to find the side effects of how switches and relays switch to get these things to happen in good ways for the amps, but you really can take control of *exactly* what you want to happen. You can do this in BASIC, test it with a few LEDs, and have something similar running in an afternoon.

Just sayin'...