dc voltage at speaker

[J M Fahey]

@ Roly: the basic idea about grounding the amp output rail and letting float the PSU centerpoint is the following:

If you want to make an amp where an Op Amp directly drives the output transistors (Darlington or MOS), you can't get more than 15V Pk either side ... because the Op Amp only stands so much.

*One* trick to reach more (and which carries its own set of problems), is to float the Op Amp supply (as you see, there's no escape from some kind of floating rail or another) and make it track the speaker out voltage.

Fender is a big fan of that , just look at any of their >50W amps.
They use a speial Op Amp which stands +/-25V rails, feed it around +/-22 or 23V and float (bootstrap) its supplies so when providing +40V peak, its +rail reaches +44V and the - one reaches 0V ... never surpassing then almost 50V end-to-end it can stand.

2 problems with this:
1) Higher voltage OpAmps are hard to find and very expensive (tell *that* to the Company Accountant)   )
2) anyway there's a parameter calles "common mode voltage swing" which tells you how far away from ground can *both* (+ and -) inputs be from ground .... so even if you successfully float your Op Amp supplies, you tear its inputs to pieces ... no kidding.

But with this floating rails configuration, emitters stay tied to "ground" so they can't be more than a couple volts away, and their bases are 1 or 2V away at most.
What do we need now?
+5V to +5V peaks?
Driven from *any* +/-15V fed Op Amp?
Piece of cake.
That's the main point. The first amps were this was seen were some QSC amps.
*Very* simple, yet scary powerful amps, driven by puny Op Amps.
Something which was important for them, is that they use complementary PNP/NPN transistors, grounding their collectors.
They bolt transistor cases straight to the heat sinks, with grease but no mica.
Not to save a few cents, but to lower heat resistance.
http://www.qscaudio.com/support/library/schems/Discontinued/USA%20Series/USA900.pdf
Simple and elegant.

[Roly]

Well it was pejorative Enzo, until JM explained the reason I see the output coupled to the strays of the PSU and the need to filter the output from the 15 volt rails.  I've had a few of the older Peavy's across the bench but never encountered one of these before.  {In large part because most of the gear that is brought to me is older stuff like valve gear that younger techs seem unwilling to work on, leading to a situation where they don't get valve experience and I don't get to see this sort of stuff.}

Thanks for the clarification JM.

I'll leave this one to you guys and follow with interest.

Happy New Year.

[Enzo]

Peavey will be back after the New Years, contact customerservice@peavey.com and ask for the schematic you need.  Look on the serial number plate, and get the exact model name.  If it says STudio Pro 112, make sure to include any extra numbers like "Export" or "'04" or whatever.  Include your serial number to help them.   And do not add anything to the name.  If it says STudio Pro, don;t add the 112 unless that is actually printed on the front, for example.

[cigarbox2]

I think that i found the problem pin 4 of the power amp ic that is bolted to the heat sink had broken where it comes out of the ic itself so was not clearly visible while in circuit. This morning i was checking to see if i could figure out if it was coming from the pre or power in the process i noticed if i put pressure on the board it would stop especially when pressing near the ic. So i removed the board and the heat sink and saw the broken pin.

[cigarbox2]

I finally got around to installing the ic after powering up within seconds pin 4 fried. Pre amp board is disconnected it made no difference before so i did not reconnect. I was able to get correct schematic from peavey. I will post. Any suggestions would be appreciated.

[cigarbox2]

here's the schematic.

[J M Fahey]

Well, pin 4 is -35V *and* it's also connected internally to the LM3886 back metallic tab, so I guess you put a new chip but didn't use the proper insulating micas+grease or Silpad, so you shorted pin 4 to ground.
I also guess you used a larger than normal fuse, or it would have blown before pin 4 solder melting (which indicates a GROSS overload).
So:
1) put a new LM3886 there (the one there is most certainly fried)
2) properly insulate it. Don't forget the grease unless you use a rubber Silpad.
3) put the proper fuse where it belongs.
4) build and use a lamb bulb limiter (search SSGuitar)
5) turn the amp on, no speaker connected nor signal applied and measure +/-V rails and DC voltage at speaker output.
Thanks.

[Loudthud]

I would add to what JM said that before you apply power to the amp, you measure the ohms from pin 4 to ground. You may get an unstable reading but not a short or reading less than 1 ohm.