Weird fault in Marshall 100V model 8100. Advice appreciated!

[J M Fahey]

Agree with the above posts and confirm: all amps "wake up stupid" (hey, me too ) at least for a fraction of a second, while all caps reach *expected* voltages.
If load pulls rails way too low it may never recover and stay there, nothing "broken" there.
Solution is to start without load if on limiter or with it if straight into the mains outlet,where you plug only after you checked there is no DC on the output..

[allante666]

I realise that theres usually a bit of a surge while they wake up. If I start without load on either limiter or direct, its fine, works perfect. If i plug load in while its running its still fine but if I start from cold with a load on then it puts the -ve rail straight onto the speaker output and stays there. It does the same thing with or without a limiter. pulling the speaker out and reconnecting it cures it.

[allante666]

ok so I followed Enzo's advice and swapped the TL072 as I too have a drawer full! Lifted d4 and d5 to disable limiters and also checked the mentioned resistors. Due to the outputs being swapped around previously I replaced them with new TIP142/147's and replaced the TIP 31 with the correct TIP29.

Still the same..

The only time it will start up with a load on and not put DC on the output is if IC8 is out of circuit (no sound obviously)

[Jazz P Bass]

This sounds suspiciously like that 'Ashdown' thread awhile back.

http://music-electronics-forum.com/t38478/

[Jazz P Bass]

Here is the schematic: Marshall-Valvestate80-80W-8080-8100-8412-Schematic.pdf

I wonder if that capacitor (C55) on the feedback path is to blame.

[allante666]

I thought about the cap but it measures pretty much spot on. I lifted one end to measure it.

[The Dude]

I've also seen this when a filter cap starts to fail, but before it fails completely. It takes longer to charge than it's opposite. You might also measure the output at startup and see if there is any AC content along with the DC. Since the output DC is positive, I'd suspect the cap on the negative rail. I'm not saying this IS the problem, but could be and it's worth checking.

[allante666]

Ill give it a coat of looking at! Not something I had thought of but as you say worth a look. I've been cabinet building today so haven't had chance to get back to it.

[allante666]

Update. So, after having sat on its own for a weekend I come back to it to find its behaving totally differently! I power it up with no load as before but this time I get the limiter lamp half lit. adding a speaker it mmakes it light up fully. Not what it was doing before!
So I go back to basics and check power rails to find my 15 volt rails are being pulled down. Traced this to IC3, and IC9, Oh, and IC10! Without these in circuit my 15v are stable and solid. I figured it could only have been a spike on the rails which would have knocked out all 3 chips but the only thing i found was on one of the 15 volt smoothing caps was not a dry joint but a crack in the solder probably from the factory. I wondered with me turning th epcb over and resting it upside down I may have caused the cap to disconnect. Would this have been enough to damage the chips?

After resoldering it I find that (although I dont have IC9 in circuit as I dont have one but I think its only in the reverb circuit) now, when I power up with a load (or without) it powers up as it should. No limiter lights, no DC on th ethe speaker socket.

However! not all is as shoudl be as I find that if I use the boost channel with full gain, when I try turning the volume up I get a high pitched squeak. I'm guessing this is something to do with an earth but just wondered if it could also be one of the other IC's thats a bit iffy?

[allante666]

Update again!
High pitch squeak only happens on one particular guitar so I guess that's a pretty easy diagnosis. (in fact only that guitar and only when the neck pickup is being used) Only fault on it now is that the clean channel is really low volume. I'm suspecting one of the op amps but that's a job for another day!