QSC PLX2402 O/P stage heating

[Nyqusit5]

Hi,

I am currently repairing a PLX2402 power amp. Q28 (o/p transistor) was shorted its R88+R91 ballast resistors were open. After replacing them, The amp works for a few seconds but some output transistors slowly conduct more and more current (until my current limiter reduces the supply voltage s much that the amp shuts off and starts again. I replaced all output transistors thinking some of them would be the origin of the problem but the situation hasn't changed. I have checked all ballast resistors and they seem ok.
3 o/p transistors conduct increasingly higher current after start-up: Q35, Q37 and Q40. They also heat up more than the others. What could cause this? Q27 (MJ15032) seems fine.

Schematic attached (I had to uplaod a screenshot because the service manual is too big)


Thanks in advance!

[mhuss]

Are the diodes (including the 10v Zener) in the Q20/22 circuit all OK?

[g1]

If you go as far as replacing all the outputs, you should also replace the drivers (Q21 & Q22). If the driver is breaking down it could play havoc with the bias current of the outputs.

[Nyqusit5]

Wrong schematic...

Aufully sorry. i posted another circuit I am working on. This is what sleep deprivation does to you! Here is the correct one.


Again, sorry for the waste of time and thank you for your input.

[mhuss]

Have you tried swapping Q27 with one of the 'hot' ones?
Did you check the resistance of the four sets of emitter resistors to make sure they're very close to equal?
Are the replacement transistors well matched?

[g1]

Ok, Q26 & Q27 then.
It starts up ok then runs away, so whatever the faulty part is will probably read fine under static conditions.

[52 Bill]

I'd be more interested to know how the idle bias is set and what it is set to after the repair.

[Nyqusit5]

The bias is set by VR43. I have it at minimum now and it is easier to identify which op transistors are unstable because the heating up process is slower. The most unstable is Q29, then Q34. I have changed the ballast resistors for Q29, and also changed Q29 but it still behaves the same way (no change).
The only way I have found to slow or reverse this behaviour is to cool Q22 down. So I changed Q22 (2SA1837), but there still is no difference...