Crate GX-212 power amp puzzle

[Enzo]

A drifting 15v rail won't be due to a cap. if anything it will be a zener warming up. The zeners are what sets that voltage.

But you have gross issues, you have excess currents and DC on the output type trouble. That is not caused by one 15v rail being off three tenths of a volt. And any readings you may have taken with IC3 removed would have NO influence from 15v rails at all.

I don't recall, have we yet made the distinction on current draw? When the excess current flows, is that ONLY with a load present, which would mean DC on the output? Or does it also occur with NO load, and both sides are conducting at once? They are completely different problems.

Your little voltages on the empty IC5 pins, 0.159v on pin 7, are just the center point of resistors R51,53,54,56 strung between the two 16v supplies. Now THAT could be the fault of your extra tiny rail voltage. But in the whole system, that gets corrected by the op amps.

Your little voltage climb sounds thermal, so get some freeze spray and chill the zener or the cap or anything else you might suspect and see what makes the voltage drop back where it started.

[Steve Conner]

Enzo, I think the design of this circuit is a bit "meh" in that the idle current actually does end up sensitive to the exact value of the 16V rails. I wouldn't be surprised if the output transistors failed in the first place because the zener unsoldered itself and the -16V rail went high, causing very high idle dissipation.

[DGstrat]

True, I had not considered the temperature effect on the zener. I will see if I can scare up some freeze spray.

So to recap, here's what I have observed in each configuration:

All components installed, no load of any type connected --

Heavy current draw, bright bulb, power supply rails choked way down.

Thats what led me to replace IC3 and IC5 before you guys educated me on their interactions.
After that I was still in the same place. The positive side of the 5532 was driving Q12 and Q13 hard (2v on the base) and the negative side was a bit weaker but still driving (-0.7V with the emitter at +0.56).
I measured over 200mv being dropped by the emitter resistors in this state.

Then I pulled IC3 and that removed the high current condition. The amp can be run without the dim bulb tester in this mode.
I was puzzled by the state of IC5 driving so hard -- then you showed me that the negative output voltage was making him try to pull the amp positive, but with no IC3, it wasn't going anywhere.

Then I pulled IC5 and measured the steady state of the network around D29. Close to right, but a bit on the negative side.

Then I installed a new IC3 (since I hadn't done that combo before). Still no high current condition, but the output has over a volt of DC on it and the opamp responsible for driving the negative side is doing it.
When I short both ends of D29 to ground the output sits much closer to 0V and neither the positive nor negative opamp is trying to turn on the output transistors.

This lead me to question whether the -0.48V on that end of D29 is really correct. And if not, what can influence it? The diode drop is about right, the 4K resistors are sound, so that's why I started questioning the +/- power supplies since the negative is larger than the positive side.

By the way -- thank you very much for the help and analysis

[Jazz P Bass]

How much Vac ripple is on the opamp power supplies.
If there is any, then the low voltage caps are bad.

[DGstrat]

The ripple thing was incorrect. My DMM was showing 80mv of ripple, but the scope shows none on the 16V or -16V rails.

[DGstrat]

Update

So I replaced D25 and D26. This did not make any appreciable difference. +/-16 are a little lower, but they are still almost 400mv apart.

Before I put IC5 back in, I measured -1.8VDC on the output.

I added IC5 back into the circuit to retake those measurements. There is no load on the output, but both sets of transistors are turned on and are pulling down the power rails, as before.

I am seeing approximately 70mv across each of the 4 emitter resistors (a heck of a lot more than the 7.5 that the schematic calls for).
I measure a level of very close to 0V on the output. As you said Enzo, the IC3/IC5 pair do seem to balance everything out.
However, IC3 is driving all 4 transistors to do it.

voltages:

IC3 Pins
1 = 1.85
2 = 0.26
3 = 0.26
4 = -3.0
5 = -.31
6 = -.25 <-- looks odd to me... seems like he's still trying to pull the output further negative but can't
7 = -1.6
8 = 2.8


IC5 Pins
1 = 0
2 = 0
3 = 0
4 = -3.58
5 = 0
6 = 0
7 = 0.435
8 = 2.57

[52 Bill]

I've been following along at home and you have been getting good advice here, but there isn't that much to this power amp. You've already replaced nearly half the parts, and certainly have checked or tested the other half.

Have you considered going back and double checking everything that you have already done just to make sure that no new parts have been damaged or have been inserted in the wrong way, etc.? Or maybe you already have done this.

[Jazz P Bass]

What's up with the opamp (IC3 & 5) power pins (4 & 8).
And they come off two separate sources.
Something ain't right.

[J M Fahey]

What's up with the opamp (IC3 & 5) power pins (4 & 8).
And they come off two separate sources.
Something ain't right.

Don't worry about IC5, it's fed from "normal" rails, like most others.

The key lies in IC3, which is fed from floating lines, which track the speaker output.

It only has "+/-19V" at idle, with signal the rails follow it, but keep a 39V difference beteen pins 8 and 4 , thanks to D10 ..

In another post, Steve Conner mentioned that "Op Amps" do not have ground pin, nor need to be referenced to it, they are quite happy floating.

Anyway, this particular Crate design stinks of new designer who wants to prove the Boss he's hot.

Nothing against, but sort of wasted in a *guitar* power amp.

Fender also uses a floating Op Amp, in that case driving TIP142/147.

[Jazz P Bass]

Yeah I get the fact that the two IC's have separate sources.
What I was questioning is the voltages in the OP's last post.
Here is a copy.
IC3 Pins
1 = 1.85
2 = 0.26
3 = 0.26
4 = -3.0 ????
5 = -.31
6 = -.25
7 = -1.6
8 = 2.8 ????


IC5 Pins
1 = 0
2 = 0
3 = 0
4 = -3.58 ????
5 = 0
6 = 0
7 = 0.435
8 = 2.57 ????

[Enzo]

OK, so we have not mentioned the bias control in a while. Now if things seem stable, what does that control do now? If you are using a bulb limiter, does that control make it brighter or dimmer? If you are plugged straight into the wall, I like to monitor mains draw on a metered outlet, same question, affects draw?

[J M Fahey]

Yeah I get the fact that the two IC's have separate sources.
What I was questioning is the voltages in the OP's last post.
Here is a copy.
IC3 Pins
1 = 1.85
2 = 0.26
3 = 0.26
4 = -3.0 ????
5 = -.31
6 = -.25
7 = -1.6
8 = 2.8 ????


IC5 Pins
1 = 0
2 = 0
3 = 0
4 = -3.58 ????
5 = 0
6 = 0
7 = 0.435
8 = 2.57 ????

Yes, they *do* look very low, but for diagnosis I would like to know what the voltage is at the main rails feeding the Op Amp ones .

My sad experience with Op Amps is that most , if fed less current than a certain minimum, they fold down/become stupid/whetever and collapse their own rails .... which *might* be the case here.

[Jazz P Bass]

Voltage readings from post #5.
Copy:
IC5 has +16.1 on pin 8, -16.4 on pin 4, 0V on pins 5 and 6, yet 14.5V coming out pin 7... I have replaced it (because the previous one was also showing a high output voltage on pin 7).

IC3 (when removed) has +20.7 on pin 8, -20.5 on pin 4.

[Enzo]

Juan, remember that IC3 does not take its power from 15v power supply rails. It has its own drops and that 39v zener.

[J M Fahey]

Yes, thanks, I meant so when I spoke of "floating rails".
Don't know if I was clear enough, the point is that I didn't want to call them "+/-15V", "+/-19V" or whatever "fixed value", because that would be misleading, they actually track the speaker audio, thanks to R67 and both 22uF caps.

Call them "bootstrapped rails" if you wish, I won't disagree