70's Peavey Musician 400 split waveform

[Randall]

I have a real ratty sounding Peavey. I put a 400Hz sine thru it, and I get this unsymmetrical wave on the speaker. All eight output transistors test the same out of circuit. I'm not sure how to proceed with this one.

001.mp4

peavey-musician-power-amplifier-schematic.pdf

[The Dude]

Scope it back to where it looks like a sine wave to narrow down where the distortion starts. What are the voltages on the output transistors?

[Enzo]

Disconnect the speaker, now what does the output look like?

Looks like the pos half can't drive.

[Randall]

With the speaker disconnected I see no signal at all, which seems strange. The +/- rails are at +43v/-43v on the outputs.

[Enzo]

ALlow me one free shot... ANy chance the scope is still connected to the speaker instead of the amp? DOn't ask me why I think of it.

[Randall]

No, the scope is connected to the speaker jack. With the speaker unplugged, for some odd reason I lose signal all the way back to the SPS761 diff pair. I have a clean sine on the base of the first transistor, but nothing on the Emitter or Collector.

With the speaker plugged in I get a subtly rounded sine on the collectors on both SPS761s, and a distorted wave on the emitters that looks like the one on the speaker output. Seems maybe that diff pair needs replacing?

The loss of preamp signal with the speaker disconnected is lost on me.

[Enzo]

Plug a cord into the jack but no load on the other end. Just thinking maybe a shorting jack? It makes no sense all signal goes away unloaded by loaded it makes something.

[J M Fahey]

Amp on, no load, no signal, measure Vbe for all transistors from Q1 to Q13

[Randall]

Good to hear from you Juan.

Vbe for transistors is a little tricky since no components are designated, and some are obsolete, but here is what I believe is correct:

Q1 - 0.7v counts slowly up over a 30 second period
Q2 - 0.5v
Q3 - 0v
Q4 - 0.4v
Q5 - 0.5v
Q6 - 0v
Q7 - 0v
Q8 - 0v
Q9 - 0v
Q10 - 0v
Q11 - 0v
Q12 - 0v
Q13 - 0v
Q14 - 0v
Q15 - 0v

I'm not sure why all the 0v measurements, but I did probe the rails a couple of times to make sure my meter was actually working. These readings were done with no signal and no load. I tried it with a dummy plug as enzo suggested, no difference. They are not shorting jacks.

[J M Fahey]

You´re welcome.

We have a ton of bad Vbe voltages there

In principle I guess you lost the positive rail reaching some points where it´s needed.
Cracked tracks? Broken connectors or solder? Some open resistor?

OR Q3 might be very shorted.

I know parts are not labelled , so I added my own; use a fine tip marker number by each transistor.

Small detail: just in case, we need Vbe *polarity* too.

Next time put black probe on emitter, red one on base, so NPN will be , say, +0.6V, and PNP will be -0.6V or so.
Those are expected healthy Vbe voltages ... which clearly we do not have here.

Next tests:



a) measure Q3 Vce

16) to 25) are all voltage to ground , black probe to ground, red probe to measured point.
remember quoting polarity.
Guess 200VDC scale will be fine.

[Randall]

Q3 Vce, black on collector, red on emitter = -37v, yet in circuit it measures 2.5M ohm collector to emitter.

16 = - 0.7v
17 = 0v
18 = 43v, that 1K cement resistor measures 11M ohm in circuit
19 = 43v
20 = 43v
21 = 43v
22 = 43v
23 = -3.2v.v
24 = slowly counts up to -3.2v
25 = -3.2v

These measurements were before I found the open 1K 5watt at points 17 and 18. Once I replaced it, there is an improvement, but something is still off.

Here is what I am getting now.

[Randall]

I made those measurements again after replacing the 1K 5watt.

16 = -0.7v
17 = 1v
18 = 20.9v
19 = 42.6v
20 = 42.6v
21 = 42.6v
22 = 42.6v
23 = -0.033v
24 = -0.033v
25 = -0.033v

[Enzo]

Yes, you aer missing the bottom half of the waveform.

Is it still doing the distorted under load but ZERO output with no load?

[Randall]

Actually no, now the wave looks fine with no load. I should have mentioned it looks good under load up to a point as it is turned up, then the bottom flattens out and looks like it has an oscillation on it. I tried to post another video of it, but somethings amiss with that.

[Enzo]

OK, good waveform unloaded means the amp can make the output voltage, but when it is loaded one side collapses, meaning that side cannot put out the current at that voltage.

SInce you have THREE outputs in parallel on each side, I tend to think they are not the problem, so look further left.

Just to satisfy an old man, look at Q14, Q15. Each one has a diode. Unsolder one end of each diode and lift it from the circuit. That will disable the current limiters. ANy change?

And I always check the resistors for open, if you have not already. Especially those associated with Q4,5,6,7.