Sir . . . . . newjeeper
Thought that you had not came back yet, but it was just the fact of the thread transitioning to its second page.
Still looks like there are no great breakthroughs.
Referencing to your very top page one of your supplied schematics of the power amp portion.
As per Sir dorkes procedure:
2.Try the same "noise test" with Line-In (J3-in amp schematics):
insert a "shorted plug" into J3 (i.e center pin connected to outer one,
without a cable, just the plug opened and shorted with a short wire).
I think that you said that the audio problem was STILL present using that procedure and that the amp wouldn't amp, even with signal put into the J3 line input connector/jack .
Try this now. . . . .
Note that below the schematics J3 line in and J2 line jacks that there is a 4 pin J1, RA type of connector.
That carries a signal ground and a +and - 45 down to your lower boards to be voltage dropped down for the other boards + and - 15 v supplies and the last connection being the carrying in of a processed line audio signal from the other boards and lastly, a somewhat "exotic" grounding mode.
UNPLUG that connector and test the unit for your audio condition.
That way we will know, dead certain, that no audio is getting into the power amp or if any loading fault on the developed + and - 15 VDC supplies could be affecting the power amps +and- 42V supplies.
The power amp should then be as quiet as a mouse peeing on a cotton ball. Excluding any frontal gaussian or shot noise potentially coming from Q1-Q2 semis.
If you still have your noise-es-es-es then . . . . .try this possibility:
While you are at it, take voltmeter in hand . . .a . .la . . .DC mode and power up the amp and meter from your speaker plugs ground shell to any of the possible 4 connections that R35-36-37-38 emitter resistors share.
If multi stage biasing is proper and balance out, the +-47 differentials will place close to O VDC,on that audio power buss, then try subbing in a speaker as an amp load, even an old clunker if you are leery.
I am seeing Q1 and Q2 transistors along with cross mates Q3 and Q4 as differential pairs with input signal coming in thru C3 (?) .15 ufd to Q1-2 bases.
There is also Q1-2 base connection to "ground" via R3(?) 47K. Also include HF shaping network C1-R2.
Move over to companion pair Q3-4 and you will find the 47K resistor being there as R40 and the like HF network R41-C23.
When you follow that line to the right, you find there being no LOW Z DeeCee path to ground. The pre driver, driver and output transistors are not saturated by conduction while being at a no signal condition.
Only when you finally reach the end, and get past the L1 HF choke and its shunting resistor and connect to the R?10 ohm and series C22 .1 ufd (ultrasonic filter) do you even get a decent LOW Z AC path to ground.
By your having a speaker plugged in, it is completing a LOW Z DeeCee path to ground(s) to couple back into that R40 47K. (C 23's right lead says . . . . ME TOO !)
Additionally, BOTH types of grounds are being made at the speaker jacks ground shells.
Thasssit . . . . check it out . . . . .
73's de Edd
.