Quote from: phatt on September 22, 2016, 09:56:58 AM
I agree it is likely part of the issue,,,When I said check ground/common at R61 & R68 Check they are all connected to each other. You should have continuity between all points that go to common.

Then,,, check DC voltages at points shown.

As you have had bad connections, drifting resistors then check resistors at those nodes for drift or open circuit.
R63 & R65 supply power to the emitters of Q11 & Q12.
Obviously if no voltage at emitters then the amp won't go.
Use the working amp as cross reference.

I'll do that tonight, yes, I've been comparing with the working channel which is a great resource.

Seems like it may indeed be a grounding issue. When the pots are secured to the chassis the behavior changes, and that strikes me as something that shouldn't happen if the board is properly grounded.

Quote from: phatt on September 22, 2016, 07:12:35 AM
I'd Check that R61 (68k) is in fact going to ground.
The Base of Q11 needs a common (ground) reference, otherwise the circuit won't work.
Verify Common at R68 also.
Phil.

R61 measures 67k with 0.2 ohm to ground, R68 measures only 500 ohms when it's a 1k resistor so I'll swap that out -- again 0.2 ohm to ground on the ground side. I swapped out R68, no change.

Quote from: gbono on September 22, 2016, 02:38:02 AM
What signal do you have at the collector of Q13?

No signal at Q13, the scope loses trigger and the line drops off.

Using a signal generator I ran a square wave through and it seems to lose the signal (compared to CH2) at R63, in other words, past Q11 and Q12, which I guess I'm going to test again.

Replaced Q11 and Q12 and I still lose the signal at R63, it may be lost before that, but I definitely have it at R62. See attached portion of the circuit. Any thoughts?

Update: Seems I lose the signal at the emitter of Q11, R64 tests alright, C34 was OL so I replaced it with a 12pf I had on hand, no change, D3 tests alright.

Quote from: gbono on September 21, 2016, 03:37:02 PM
Is the problem in the preamp or the power amp section? It appears that the problem was ch2 and now it's ch1??

If you don't have a scope you can make a "signal chaser" https://www.youtube.com/watch?v=IEcyDzYAsuU

Pretty sure the problem is with CH1, it only affects CH2 when the CH1 speaker is plugged in and the short continues over to CH2 -- if that makes sense. If there's no speaker plugged into CH1 then CH2 works perfectly fine. I have a used a signal probe/chaser on effect pedals but I was wary of using it with such high voltages around, I don't want to fry another amp trying to test this one. I do have a scope, an old Tektronix, but I'm not fully proficient at it.

Quote from: phatt on September 21, 2016, 08:40:30 AM
Oh dear
I think we need better minds to look at this one.
Enzo, JMF, Teemu, anyone care to comment? xP
As I'm sure you know I don't wish to turn substatica's Amp into landfill with wrong advice.
Phil.

Thanks for all your help. I'm pretty confident that the transistors are all good now. I guess I'll go over all the resistors and caps on that board again.

The speaker works fine when used on the working channel and measures 7.5 ohms, so I don't think it's that.

Though I just tried it again and I get DC hum on the good channel when a speaker is plugged in on the bad channel.

It's changed again, now the bad channel is measuring 20 VDC at the speaker outputs with no speaker.

Speaker seems alright. When I measure resistance across the good CH2 speaker outputs where they connect to the board I get an expected 7 ohms and change, when I measure resistance across the problematic CH1 speaker outputs where they connect to the board I get a slowly climbing and falling resistance that climbs and falls, 100 ohm, 200, 70, 50. Does that mean anything to anyone?

Replaced both output transistors, Q1, Q2 with Sanken 2SC2921-ND's. Things were still being wonky. Pulled all the transistors off the board and tested them. They all tested good so I soldered them back in place. Now CH2 plays fine, the LBL bulb lights up then dims, voltage at CH1 output is 0.01 VDC or so -- in other words everything seems fine but I'm getting no sound out of CH1. I have to test that speaker, will report back when that's done. Feels like I'm getting close to cracking this nut.

Here's my 5 minute limiter...

I swapped in a 2SC1624 for the 2SC1628-Y temporarily. Looks like Q1, 2SD845, an output transistor? is completely shorted. Any suggestions for a replacement? Does that mean I have use a new matching pair? Once I pulled Q1 the amp powers up with the LBL and CH2 plays, I heard another snap, I'm guessing from Q2 again, but it still tests fine.

Update: Seems like a Sanken 2SC2921-ND will work for the 2SD845, though a little pricey at $9 per. Still trying to find a replacement for the C1628-Y, NTE190 is supposedly equivalent but the only source I've found wants $21, yeesh.

Replaced Q1, Q2.

Q14 is bad, but it's 2S1628-Y, any chance I could replace that with one of the 2sC1624's I picked up? At least temporarily?

http://alltransistors.com/transistor.php?transistor=12170
http://alltransistors.com/transistor.php?transistor=12177

Otherwise I have to order some parts. The local shop didn't have any 2S1628-Y.

Quote from: gbono on September 16, 2016, 04:01:28 PM
It is possible the output transistors have been or were originally damaged but look at the DC voltages at the base, emitter and collector of Q1-4 (wouldn't hurt to look at the same junctions of Q14-15. See this link if you need some help in understanding how to trouble shoot transistor circuits: http://www.allaboutcircuits.com/worksheets/bjt-amplifier-troubleshooting/

I sometimes use the LBL to help establish what is getting hot in the circuit: a cheap IR thermometer is a good tool to use and scanning the board before and after power is applied sometime gives a clue as to what is not happy on your board.

I have a cheap IR thermometer -- I'll give that a go

It seems indicative that now that I've replaced D1, R72, R73, R74, Q14 and Q15 the fuses pop right away. Not sure what that's indicative of, but seems telling?

Quote from: phatt on September 16, 2016, 08:37:37 AM
From that picture; If that arc is constant then your molex plug has corroded and is now Arcing inside the plug. very bad!!! Cleaning won't bring the transistor back to life,,,
   Washing a blood stained shirt won't fix a bullet wound :crazy2:
As I said Before ,, Your output transistors are likely dead and if you are unsure how to test them then just replace all 5 actives and save yourself the fussin.
Then retest with light bulb first, and no speaker until it's stable.
Oh and the molex plug will need to be replaced as the arcing will have destroyed the contact inside by now.
That may have been what killed it in the first place. 8|
Phil.

The arc is not constant, I think since I replaced some components it's changed where the short it occuring. I didn't clean the transistor to fix the transistor, only to get rid of debris and corrosion that may have contributed to the short. There was some sort of corrosion on the case around that molex, and one of the leads on the transistor seems a bit corroded, that's why I took it off and cleaned the case and the transistor. The transistor tests fine, as far as I can tell.

I'll test the rest of the output transistors, by output you mean Q1, Q2, Q13, Q14, Q15? Trying to squeeze in this troubleshooting after work, after the kid goes to bed doesn't get many transistors changed out nightly