Amp Head blowing fuses

[craig4739]

I am at my wits end with my Marshall 8100 head. It blows fuses. I removed the power darlingtons TR8 and TR10 and the fuse held. I have traced a good signal through to the output of IC8 pin #1. Pin # 7 shows a good output for a brief few seconds on power up and then becomes grossly distorted. Lifting the RC combination of R81 and C 48 restores a good output signal on Pin#7 of IC8. Not sure where to go from here. As I said, the output transistors TR8 and TR10 are removed. No apparent shorts to ground. Power supply checks out ok

Any advice would be most welcome.

http://www.amparchives.com/Amp%20Arc..._8080_8100.pdf

[nickb]

I am at my wits end with my Marshall 8100 head. It blows fuses. I removed the power darlingtons TR8 and TR10 and the fuse held. I have traced a good signal through to the output of IC8 pin #1. Pin # 7 shows a good output for a brief few seconds on power up and then becomes grossly distorted. Lifting the RC combination of R81 and C 48 restores a good output signal on Pin#7 of IC8. Not sure where to go from here. As I said, the output transistors TR8 and TR10 are removed. No apparent shorts to ground. Power supply checks out ok

Any advice would be most welcome.

http://www.amparchives.com/Amp%20Arc..._8080_8100.pdf

(1) Do not plug the speaker in when testing.
(2) Check four all voltage rails are present and correct.
(3) Most likely one or more of the other transistors are bad. You need to remove and test them, especially TR4-7 and TR9. Did you test TR8 and TR10?
(4) Also, while those transistors are out, check the resistors in the power amp.
IC8 will not give a valid output until it's DC conditions are restored.

[craig4739]

All voltages check out ok. Original TR8 and 10 were blown. Replaced ones did not work (amp buzzed for a second and then went quiet) I removed TR8 and 10 and they both checked out good. Guess I didn't leave them in long enough to blow. TR9 is suspect only because it is the only transistor common to both channels....what does this damm thing do anyway?

[nickb]

All voltages check out ok. Original TR8 and 10 were blown. Replaced ones did not work (amp buzzed for a second and then went quiet) I removed TR8 and 10 and they both checked out good. Guess I didn't leave them in long enough to blow. TR9 is suspect only because it is the only transistor common to both channels....what does this damm thing do anyway?

TR9 sets the voltage difference between bases of the output transistors and so sets the quiescent current.

ALL the components have to be right or bad things will happen. Also, it sounds like you had the speaker in -bad idea as that leads to burned out speakers- maybe that's why it went quiet Execute steps 3 and 4 and don't skip anything.

[craig4739]

Thanks all,
I'm off to try your recommendations.

Craig

[J M Fahey]

Agree and add: build and use a lamp buln limiter.

[craig4739]

TR 4,5,6,7,8,9,&10 all check out good. Will be checking the resistors later today. Possible that it might be D4 or 5 or TR 11 or 12? Getting close to throwing this thing in the trash

[km6xz]

Are you testing with a speaker or load? If so, you ARE going to damage the amp worse, and might have taken out the speaker also. We can't give advice if you ignore the most basic recommendation repeated above concerning testing without a load. Even if there is a shorted transistor, without a load the amp can be run and trouble shot but with a load, only more severe damage can result until the amp is stable and working perfectly without a load.

[craig4739]

Testing with no load

[km6xz]

If it is pulling current from the output transistors with no load, either both upper and lower side or shorted or they are both turned on at the same time(bias circuit).

[nickb]

TR 4,5,6,7,8,9,&10 all check out good. Will be checking the resistors later today. Possible that it might be D4 or 5 or TR 11 or 12? Getting close to throwing this thing in the trash

These components are there to try and protect the output transistors so I don't think they could cause the problem you are seeing.

As was said in post #10, either the both output transistors are shorted or there is a problem with the bias cct. The bias is primarily set by TR9, R74 and R73. If TR9 is open, R74 open or high, or R73 short or low then the bias will be way to high.

If it turns out all resistors are OK then I'd ask what method you are using to check the transistors.

The only other thing I can thing of is if C49 and C48 were very leaky but that is very, very unlikely; but never say never, right?

[craig4739]

Transistors were checked with a dmm. results for TR9 were

B to E .578
B to C .578
E to B OL
C to B OL
C to E OL

[craig4739]

Just went through and check out all components in the power section. Everything checks out good except.........
R118, R96, and R120 are supposed to be .33 ohms (5 watt). they come in at .7 ohms

Are these discrepancies enough to make a difference?

[nickb]

Just went through and check out all components in the power section. Everything checks out good except.........
R118, R96, and R120 are supposed to be .33 ohms (5 watt). they come in at .7 ohms

Are these discrepancies enough to make a difference?

Nope, that's not it. You really need a specialized meter to measure under an ohm or two with any accuracy. It wouldn't cause it anyway.

Something else to check would be that the output transistors and TR9 are all insulated from the heatsink. The collectors of the output transistors are connected to the case so they use insulting bushes and washer to prevent shorting. Check they are OK with meter.

Failing that I would try to bring the power amp in a limited way. This means restoring everything and putting those transistors back (bummer I know), with the exception of the output ones and temporarily wire linking the base and emitter pads of TR8 on the PCB.

This should allow DC conditions to be established without too much current. Power it up then you go round with your meter and measure the voltages at every* transistor lead and power supplies and post them back so we can see what doesn't make sense. You might find it easier to extend the wires to the board.

*edit: by 'every' I don't mean every component on the board, just the power amp section of course ( I thought that was obvious but maybe not..)

[J M Fahey]

Basically, measuring part by part is slow and stressing, both to you, the part, but much worse to the PCB, because pads and tracks get torn and destroyed with so much soldering and pulling.

Soon you end with an unrepairable mess.

Of course, it does pay measuring the big suspects: power transistors and their drivers, but not *everything* on the PCB.

As suggested above, replace clearly shorted transistors, plug the amp into the lamp limiter (no load, all controls on 0), and start measuring voltages.

Beginning with + and - rails (nominally +/-40V or thereabouts), +/- 15V , DC at speaker out.

Unexpected values will start to point to real culprits.

This "long path" ends up being the shortest one