Mackie SRS1500 rebuild problem

[mnow]

Just rebuilt an SRS1500, amp idles well but when signal is increased it starts clipping on the negative side and hence the DC voltage on the speaker terminals rises to something like 7 volts.
Seem to be a bias issue as signal is increased. Still looking for what I call a "dynamic" resistor (a resistor that changes value as load is increased) but no luck yet. Clipping starts at the input stage of the power amp but am still trying to ascertain if this is cause or effect from something downstream.

[MarkusBass]

Just rebuilt an SRS1500, amp idles well but when signal is increased it starts clipping on the negative side and hence the DC voltage on the speaker terminals rises to something like 7 volts.
Seem to be a bias issue as signal is increased. Still looking for what I call a "dynamic" resistor (a resistor that changes value as load is increased) but no luck yet.

Why do you think it is related to bias? For me it looks like failed semiconductor. Have you check all transistors and diodes?

Mark

[g1]

This is with or without load?
If it is with load, does the signal still collapse with no load?

[Jazz P Bass]

I would work my way back to the physical input to the power amp.
You need to determine if it is a preamp or a power amp issue.

Also, this is a Class G rail switching design.
So you need to determine if the collapse is at the high level switching level.

[mnow]

First test were done without load just to make sure nothing bad going on.
Finally added the load and sure as heck, got a switching problem going on
now signal looks like something from an SCR circuit. Got a bunch of noise coming
out of 1 of the opto-couplers while the other put out a nice square wave.
(Kind of jumped the gun on this one I guess)
I believe this is actually a class G circuit as well.

[Jazz P Bass]

Yep.
It is a Class G output section.

[mnow]

Got the load on. Amp idles just fine but as soon as signal is applied, speaker out runs from +1V to +15V.
Clean wave up to the power amp but notice an imbalance in Q23 and Q29 (possible effect from DC on speaker line)
and output is severely clipped on the negative side even at Q25, Q28 (again possible effect from output).
U2 opto seems to have much noise coming out compared to U1 so need to order replacements before proceeding.

[mozwell]

Im not familiar with these amps, (& not an amp tech) but at lower levels the output power is provided by a high & low side of npn & pnp, like a normal class AB amp.
When the output voltage to the speaker goes higher, the rail switching turns on.

So lets divide & conquer.
It seems from your tests like all the semiconductors check ok, buts lets double check
With power OFF & all power supply rails discharged, measure all diodes & transistors. This includes the LED's

Now we can power on
First, check all the power supplies, measure & record their voltages, V1+, V1-, V2+, V2-
Then check the voltages across all the zeners D21, D27, D33, D34, D29, D31, D5, D8 not all of these are referenced to 0V, just measure across each zener & see what volts are there.
Only when the power supplies are all ok can we go further.
Note check the fuses F1 & F2 are ok

All good, then apply an input signal so that we get maybe 5VAC across the speaker terminals. This is a low level output, which will "check" the low power part of the amp, before the rail switching operates.
We need to see a nice clean output here.

If we don't get a clean signal, it might be the regulators on the V2+ & V2- rail, while you have output of 5VAC, measure the V2+ & V2- rails, are they steady DC, if not, we may need to look at the circuit around Q4 & Q5

Measure & post your test results here, so the good people of this forum can point you in the right direction


*** edit *** if the mosfet drivers are not ok, they can be disabled easily, short out pin 2 to pin 3 of each optocoupler. If you still get the noise on your output, remove each optocoupler & short D1 and D4 to disable the rail switching mosfets

[mnow]

Never mind!! Stupid mistake of rushing the job.

Found D5 shorted effectively shutting off drive to Q1!!!
Odd though that it would upset the DC out at lower levels.

Thought I had checked that diode before but must have missed it obviously.

Thanks for all the input anyhow and keep up the good work here.