Question about the function of a capacitor in a circuit

[xavier60]

Removing output transistors will open the feedback loop. Tack in a resistor from Base to Emitter of one of the missing transistor's pads to keep the loop closed. Use a 1K or else anything from 470Ω to 2.2K.

[fedaalis84]

Removing output transistors will open the feedback loop. Tack in a resistor from Base to Emitter of one of the missing transistor's pads to keep the loop closed. Use a 1K or else anything from 470Ω to 2.2K.

Victory! Or at least a partial one.

I don't have any spare resistors around, so I took a 2.2KOhm SMD resistor out of an old broken phone handset & jumpered it with some wires between the base and emitter of the PNP transistor on the right channel.

Boom, I'm getting 2V on the VOUT- pin on that channel. It should be a small negative voltage. But I'm assuming because the VOUT+ circuit is open on that side, it's leaning toward the positive.

In any case... I *THINK* I'm safe to put the transistors in! Wooo.

Will give that a try next.

[xavier60]

Did you find a reason for the large voltage difference between the VOUTs on the right channel?

[xavier60]

If there really was a large voltage difference between the VOUTs on the right channel, a reason needs to be found even if the problem has seemed to have fixed itself.
It is possibly the reason for the original failure of the output transistors.

[fedaalis84]

If there really was a large voltage difference between the VOUTs on the right channel, a reason needs to be found even if the problem has seemed to have fixed itself.
It is possibly the reason for the original failure of the output transistors.

Nope. Not finding the reason. But there is definitely still a problem, I toasted another set of transistors

I am missing R759 (should be 100ohm) - I put in a ~300 something ohm guy. So not sure of that is the reason.

Voltages were great - everything was great, no protect... Tried out some music on headphones - worked (but the music was scratchy and low volume).

Pulled out the headphones, went into protect. Fried right channel transistors.

Not sure what's up...

[fedaalis84]

You must be right and there's a problem with the biasing circuitry on the right channel. But everything had checked out good on my basic tests.. At least, the transistors were functioning as expected (diodes in the correct direction, etc.), resistors were the correct values...

I think I identified a starting point for an investigation:

R778 is reading ~19kohm resistance (EDIT: resistance to ground) on the emitter resistor side, but anywhere from 200kohm-3mega ohm on the other side.

On the left channel, R728 is reading ~47kohm on both sides, reliably similar on either side.

These are "4.7 ohm" resistors, but with an inductor on the other side of the board it looks like.

It's almost as if the inductor is connected on only one side of R778..

Do you think that could jazz things up?

[fedaalis84]

So I made a few more observations.

Firstly - I think the weird resistance measurements on R778 were just because the positive rail capacitor had some charge in it. So I don't think that's an issue.

I tossed a potentiometer in between the base and emitter of Q754, and have it sitting at around 1.8kohms. Both VOUTs are reading OK voltages.

HOWEVER...

I removed a toasted replacement of R759, so that part of the circuit is open... And the emitter on Q753 is reading +15-20VDC... Its only connection anywhere in circuit is to the collector via C768.

So now I'm thinking C768 is leaking DC from the collector...?

[fedaalis84]

Well I'm done for a while... I let some magic smoke out of IC700 today.

Once all parts come in - I'll replace everything as it should be, and hope for the best.

I think R759 is why the right channel continued to short out - I had put a ~380ish ohm TINY SMD replacement, probably below the 1/4W rating of the original 100ohm there. And that was fried after

If it wasn't that, then it was IC700, which I'll be replacing anyway.

I did my due diligence on the biasing circuitry prior to the output transistors, and after double checking, triple checking, that all looks OK.

If it still doesn't work I think I just need to let it go, maybe keep it on a shelf until I've learned some more (and am willing to spend more money trying to fix it).

[fedaalis84]

Looking like Q751 is not functioning properly actually. The diode readings from base to emitter are reading about twice as high as the left channel equivalent.

Will try replacing & see what happens.

[fedaalis84]

IC700 is out, but everything else is back in, except instead of transistors, I have 1K Ohm resistors from the base to emitter.

I'm getting about 0.5V DC on the left channel, but up in the area of 3-4V DC on the right.

[xavier60]

IC700 is out, but everything else is back in, except instead of transistors, I have 1K Ohm resistors from the base to emitter.

I'm getting about 0.5V DC on the left channel, but up in the area of 3-4V DC on the right.

I would expect to see some negative voltage coming from the protection  circuit. It should be the same on both channels. Did the Emitter resistors get damaged on the right channel?
Do you mount the power transistors well?
Thermal paste must be used. The transistors have to sit tight and flat against the mounting surface. The leads must be soldered last of all.

[xavier60]

On second thoughts, I think it's stray leakage. Are you using extra soldering flux?

[fedaalis84]

When I did have the right transistors in, I did Mount them with thermal paste and also use the little mica squares. And Soldered them in as the last step.

The emitter resistors on both sides still check out good. The protect circuit absolutely does put out some negative voltage, it brings the protect voltage (should be 3.3V) down to -50 or so millivolts.

I think the problem on the right channel is the biasing transistor Q751.

I had pulled Q752 out of circuit and it tested fine. But it did have some thermal paste between the collector and emitter. I cleaned that off, put some fresh thermal paste on the back and remounted it.

[fedaalis84]

On second thoughts, I think it's stray leakage. Are you using extra soldering flux?

Yes, and I haven't cleaned it up in a while. The first few times I made some changes oh, I would brush it off with acetone. But as I started going to the back of the board more often, I started leaving the flux instead of cleaning it.

[xavier60]

When you get the parts, as a precaution, check the bias voltage before fitting the output transistors as before.
If I have doubts, I temporarily fit 10 to 47 ohm resistors in place of the emitter resistors when the output transistors are fitted. They act as fuses and hopefully will protect the output transistors if there is a bias problem.
Unless IC700 was faulty the whole time, it would be difficult for it to be damaged by other faults. The VOUTs source/sink a limited amount of current which makes them inherently immune to overload.

[fedaalis84]

I am going to do a good cleaning of the board and see what I get. Without IC700 in, I'm pretty sure there should be ZERO voltage on any of the bases...

Edit:
Or maybe some negative voltage from the protect circuit like you say.

[fedaalis84]

 I am using some cheapo water soluble flux I got cheap on Amazon. Which I've read is maybe the worst type of flux if you don't clean it sufficiently, leaving behind a soapy, conductive residue if incompletely cleaned.

[fedaalis84]

I tested with a DC6V power supply, and current is leaking, from seemingly anywhere on the board to the bases for both transistors on the right channel. It does not leak to the bases on the left channel.

Putting power at the biasing circuitry on either side gives similar voltage and current on both sides.

So I suspect you're right, and too much flux and improper cleaning have caused some leaky current.

I tried cleaning the board thoroughly with distilled water (my flux is water soluble and I have no deionized water on hand), followed up with 50% ISO (again, not a great option, but all I have on hand).

The board is drying, but we'll see where that gets me.

[fedaalis84]

Still issues on right channel after drying. Same deal.

0.6V on left channel, no current.

Almost the positive rail on the right, about 6 microamps.

Interestingly, one of my 6800uF filter caps is reading -70 to -80 microamps on the top metal part of the cap. The other 6800uF cap (and all other caps on the board) read 0 current. I also read -40V on the top.... that sounds kind of, not good..

[xavier60]

Although  there should be no leakage, it is not enough to cause problems with the output stage biasing.
Leakage around IC700 might cause DC offset problems, but will not damage the output stage.
Leakage could affect the protection circuitry directly.

[fedaalis84]

Although  there should be no leakage, it is not enough to cause problems with the output stage biasing.
Leakage around IC700 might cause DC offset problems, but will not damage the output stage.
Leakage could affect the protection circuitry directly.

The voltage drop across base and emitter on Q751 is still out of whack compared to its left channel counterpart (~1V right channel vs. ~0.6V left. So I'm hopeful replacing that may yet resolve the issue.

[fedaalis84]

With the filter cap issue, it sounds normal. Supposedly the negative leg of capacitors are generally not insulated from the aluminum can.

[fedaalis84]

Still going strong on this bad boy.

Just popped IC700 in today.

Can't do any testing until I get the display wiring re-done. (I wound up messing up the flat cable, being impatient, and soldering in detatchable jumpers instead... but still need to finish that).

Ohhh what a mess this beast is.

But learning a bunch in the process.