Output Transistors Hot

[Rick Erickson]

See above messages about testing without a load & DC offset voltage:
Check the diode string CR20-23. You might also try pulling the IC (U12/13) out of the circuit to see if it is causing an offset or current draw through the power transistors. Also check current limiters Q6 and Q7 for shorts or damaged junctions, and diodes CR32/33 (CR48/49). I don’t think increasing resistor values is the answer to you problem.

RE

[Dave G.]

Thanks Glen, I learn something new every day. today I learned those 4 diodes are heat sensing, I thought they were in holes in the heat sink to help dissipate heat. while i'm confessing my mistakes I replaced them with a substitute from Mouser. I got the first set for the other channel from MCM, sounds like i need to order the real thing they fill more space in the holes too.

The insulators are rubberized should i use heat sink compound too? There was none in there when i first took it apart but someone had it apart before me.

I turned it on with no load and only read 3mv

[Dave G.]

Thanks guys I Have a feeling my problem might be that I replaced CR 36-39 with a substitute instead of the real thing. I used BYV26D when I replaced CR20-23 at that time I didn't know there was a problem with the other channel.
Theres no dc offset so i'll order those diodes and get back with the results.

[Mars Amp Repair]

yeah dave,
my recollection is that those bvd diodes are 2 diodes in series...I'm not looking at the schemo, but I don't think the originals are of that variety...that certainly would throw the biasing off....I don't have an issue with the rubberized insulators, but perhaps someone else can comment...good luck! glen

[Enzo]

If you replace dual diodes with single diodes, that will pull the bases closer together and the amp will run cooler but with added Xover distortion.

[Mars Amp Repair]

ahhhhh, a wise guy ehh? nyuck, nyuck, nyuck!! g

[Enzo]

I had a PV2K that had one channel running hot after a rebuild. I could not for the life of me get the channel to cool off, and all the parts were good. A call to the factory got me this advice: Oh just short across one of the diodes.

The point of the bias string in the solid state amp is to set the opposing bases just at the verge of turn on - class B land. If the bases are ever so slightly farther on than that, both sides start to conduct and current rises - they get hot. SInce the bias is set by the number of diodes, shorting across one is a simple way to reduce the idle current.

So if you have some diodes with higher junction drops, the bases will be farther apart and thus more "on." Replacing a dual with a single will indeed bring teh bases one drop closer and thus cooler. You most likely won't hear the Xover difference.

I wasn't joking.

[Steve Conner]

Haha, now we know how SS amps got their reputation for poor sound quality, it was Peavey tech support

You're really "supposed" to use a circuit called a "rubber diode" there. It uses a transistor plus a trimpot to create what is more or less a diode with an adjustable volt drop. The trimpot ends up setting the idle current of the output stage, and if you stick your rubber diode to the same heatsink as the output trannies, then it will roughly hold that idle current setting.

We're not talking hi-fi here, though, and that trimpot is one more thing to be tweaked on the production line, so I guess I can see why Peavey (and a good few other manufacturers) decided to leave it out. However, different kinds of diode have different forward voltages, so you may have trouble finding ones that work just the same as the ones PV installed. If I was fixing an amp like this for my personal use, I'd stick a rubber diode in there and twiddle it while looking at the error signal in the NFB loop, because I like my own stuff to be as good as I can make it. If I was being paid to do it, though, I'd do what Enzo did, time is money

The other day I was repairing a hi-fi amp that had its rubber diodes just dangling in mid-air about 1/8" away from the heatsink. The output stage was working, though, so I left it alone. If it ain't broke...

[Enzo]

The thing is a tiny bit of Xover is not audible, and the larger the signal around it the less so, since the notch is th same size no matter what. The larger the signal the lower the percentage. Sorta like teh voice coil rub that is easy to hear at low levels but can't hear when played loud. And if the amp is already a quarter volt high on the bias, removing one drop makes it only a quarter volt low. And even that ignores the correction from the feedback circuit.

[Mars Amp Repair]

Sorry Enzo....wasn't trying to implicate that you were joking....that's good info....g

[Dave G.]

Monday I replaced CR36-39 with the right diodes and put heat sink compound under the transistors, and the bias current seemed to be under control. Tonight I finally had the time to plug a guitar in and test it, and it sounded ok.
After I put it back in the case and plugged in the reverb tank it sounded better. (I still prefer the glow of glass bottles in a guitar amp.) So I can finally get this out of the studio and I might even get the time to play. I want to thank all you fine folks for your generous help. I learned from my mistakes and from you.(At my age I've learned alot from mistakes) God bless Ya, Dave G.