Adcom GFA 5500 troubleshooting

[dwmorrin]

I'm looking at an Adcom GFA 5500 with the following complaint: "made a bad humming noise, blew fuse"
My friend changed the fuse before bringing it in to me, and does not remember which fuse blew. Looking at them, L. Ch. + fuse is visible different than the others in make.
At this time, he does not know the make and model of the speakers he was running these into. They were something in his recording studio, but he insists that it was a proper load.
So far, in my bench test, the amp is working.
No fuses have blown at full output, but I haven't run it into a load for too long because I got a little worried checking the bias, and the fact that the heat sinks are uncomfortably hot (although it is a BIG amp).

Here the following voltages for the source resistors on the output devices of the Left channel:
(No input signal, amp left on for about 15 minutes prior to readings, all readings in mV)
R52 52
R50 86
R48 97
R46 116
R44 88
R43 90
R45 90
R47 93
R49 81
R51 70

The right channel has similarly uneven readings, ranging from 50mV up to 100mV.

Service manual bias adjustment is to monitor R52 for a voltage drop of 50mV +/- 5mv, which is has, however, the other voltages should be around 50mV as well, or not? The 116mV is especially disturbing.

Fuses are all holding, input and output voltages agree with spec sheet all the way up to clipping, but I am a bit worried about those voltage readings, and preventing future fuse pops.

Both channels heat sinks get uncomfortably hot with no signal applied.

Should I be looking for a closer range of source resistor voltage drops, or am I worrying about nothing?
Any suggestions for further testing of this unit are much appreciated.
Schematics and service manual attached.

[Jazz P Bass]

What you see in those readings is what you would see if you where matching the mosfets for idle current draw.
The 52 & the 116 seem a tad out of range.

As to the heat, is there any measurable Vdc on the output?

[dwmorrin]

Didn't see any dc on the output. I was checking it, never saw more than a few mV. There is a dc offset trim pot available.

Thanks for pointing out that it's more or less just R52 and R46 sticking out from the pack. At first, just seeing 52mV, then 86mV, then 116mV... I though it was all over the place.
Maybe if everything gets colder, they'll line up better.

[dwmorrin]

Turning the bias pot colder, I still get a pretty wide spread of mV across these source resistors.
Now the first 4 read:
R52 25mV
R50 45mV
R48 55mV
R46 69mV.

Does it make sense for the voltages to increase like that? My experience with biasing solid state amps is limited to a few bass amps. Nothing this powerful. Any guidance is much appreciated!

[Jazz P Bass]

Maybe you did not understand my previous post.

What you are measuring is the idle current draw of Each mosfet.

In a parallel setup like this , ideally, you want the mosfets 'matched' for draw.

Within reason.
The ballast resistors help the current sharing but they can only do so much.
If one mosfet wants too much current, it will hog it & get hot or fail.

Check the readings on the other ouput section.
Make a list of all of the mosfets.
Maybe, by swapping items around, you can get a pretty good balance.

[dwmorrin]

Thanks Jazz.
Thats what I was thinking, just wanted to be sure before I get messy with this thing.
I am familiar with the idea of current sharing... which is why I figured they should be close... but wasn't sure how tight the mV readings need to be. The service manual just lists +/-5mV for the first resistor, so should that tolerance apply for all? As in, all of these mV readings should be between 45-55mV?

The other section is similar in wide mV readings, but in a different order.

It's gonna be tough to desolder all the MOSFETs, so I wanted to be sure that was the next move.

How would you logically swap the MOSFETs around? Seems like an awful lot of work if I were to just try every permutation.

[Jazz P Bass]

Ideally, the factory would go through a batch of mosfets (using whatever method, but similar to what we are talking about) and match them for current draw.
Did they or are the mosfets drifting?
Hard to tell.

The amp came in because of a blown fuse.
That could have been caused my a mosfet 'hogging' all of the current (instead of sharing it)

In my mind, you should only be concerned with the oddball transistors.
The really high or the really low.

Can you take the time to list all of the readings? (both sides)
As I said, you may be able to fudge it.