wrote:
On Sun, 29 Oct 2017 05:28:38 -0600, wrote:

On Fri, 27 Oct 2017 20:25:42 -0500, Ken wrote:

wrote:
Actually lytics are worse when cold, the opposite of your
situation.

The caps I was thinking of were disc caps across the diodes, not
electrolytics elsewhere in the circuit. What puzzled me about the
30 minute delay was that electrolytic DO behave that way. If the
problem was the caps across the diodes, that would make sense since
they seem to be the source of the noise.

One of my latest additions to the shop tools is a digital infared
thermometer. They are made to measure the temperature of anything,
but are sold for automotive use, such as finding a hot brake drum,
or hot spot on an engine. But I have found this thing is pretty
accurate on electronic components too. It has a built in laser
pointer to highlight the thing it's pointed at. So it will tell the
temp of a heat sink, or CPU, or even an opamp or other chip or for
that matter, any component.

If you have one of them, or get one, try to get the temperature of
different parts on the chassis or PCB when you first turn it on,
then do the same after a half hour. Certain things will get warmer
like heatsinks, power resistors, etc but if there are passive
components or semiconductors getting quite warm or hot, you may want
to check them further. They used to sell some sort of stuff in a can
to cool parts. It might have been freon??? I used it many years ago,
I have not even looked to see if they still sell it, but if it's
still sold, that could help see if the sound changes when a
suspected part is cooled.

I did not read this whole thread, but I think you said those disc
caps across the diodes were open. I assume you mean the power
rectifier diodes. I cant understand why those would open, unless you
had a lightning surge in your power lines. That makes me wonder if
those diodes could be a little flaky. Usually diodes are either good
or bad, but lightning surges can do strange things and nothing in
electronics has a 100% definite solution. Regardless what the books
say, sometimes strange things happen that have no real explanation.
IR thermometers are pretty handy but can give wildly false readings.
This is because of the emissivity of the surface being measured. For
example, bare aluminum will measure muich cooler than it really is
when the temp is measured with an IR thermometer while black anodized
aluminum will measure pretty accurately when the IR thermometer is
used.

Any hum results from vibration or oscillation between wires, but usually within a wire (with too small of an AWG) or between wires. Re-connecting a load side wire or replacing it with a larger size AWG could help.