Discrete opamp trouble shooting.

[pucho812]

Would be useful to build up a socket where you can plug in an op-amp module to test before installing back into the full system.

That did cross my mind. The setback is not having a schematic. Tracing one will take some time.

 

[ccaudle]

Tracing one will take some time.

You don't have to trace out the schematic with bipolar transistors, high probability is you are going to find a device with base-emitter junction not at the right voltage (junction failure, or bias circuit failure), or you will see from collector or emitter voltage that either the expected current is not flowing (device failed open), or too much current is flowing, or the voltage between collector and emitter is too small (device failed short or low resistance).
When you have incorrect DC levels on a bipolar circuit like that there are only so many places operation can go wrong, likely either a failing device which is not caught by your device tester (e.g. only fails at full voltage, or partially works so the tester detects it as a working device, but the parameters are too far out of spec to work properly in circuit), or there is a hidden fault somewhere, such as a break where a trace connects to a via.

Speaking of hidden fault, this solder joint looks suspect. Was that just from removing and re-installing devices and the via did not get fully filled? Or is there something wrong with the plating in that hole?

 

[dmeek]

I can draw a schematic if you can post a few close-up photos, including the foil side.

 

[pucho812]

You don't have to trace out the schematic with bipolar transistors, high probability is you are going to find a device with base-emitter junction not at the right voltage (junction failure, or bias circuit failure), or you will see from collector or emitter voltage that either the expected current is not flowing (device failed open), or too much current is flowing, or the voltage between collector and emitter is too small (device failed short or low resistance).
When you have incorrect DC levels on a bipolar circuit like that there are only so many places operation can go wrong, likely either a failing device which is not caught by your device tester (e.g. only fails at full voltage, or partially works so the tester detects it as a working device, but the parameters are too far out of spec to work properly in circuit), or there is a hidden fault somewhere, such as a break where a trace connects to a via.

Speaking of hidden fault, this solder joint looks suspect. Was that just from removing and re-installing devices and the via did not get fully filled? Or is there something wrong with the plating in that hole?
View attachment 127085

Thanks for that breakdown. Yes that joint was from testing when I had the opamp pulled. It was desoldered after the photo before testing in a unit. I found a failed transistor on one of the discrete boards. It was beyond hot to touch. once replaced I have that one discrete opamp working as it should. 3 left to work out what is what.
I have confirmed traces are on all 3 that are left. I get continuity from solder pad to solder pad. But you never know and I will gladly look again.
Also the two open solder points below that are + voltage and ground. There are the same on the other side for - voltage and ground.

So I last left off today with one of the discretes going from working to outputting dc tripping the protection circuit and then back to working.

 

[ccaudle]

if I test between pads with no transistor soldered in, I get .55vdc diode

Are some transistors in parallel for lower noise?

 

[pucho812]

Are some transistors in parallel for lower noise?

Yes. Got another discrete section working today.
Two left.
Today I did the fine tooth comb approach. I had done it before but not to this extent.
I removed every transistor and tested it again.
Dmm testing.
Reinstalled every transistor and boom the section is working.
Time to do the other two.
Edit: One left.


Thanks for the words of advice and wisdom

Edit: all discrete section testing ok now. No more dc at the output. Proper level at the output.
thanks again for the words of wisdom

 

[ccaudle]

Reinstalled every transistor and boom the section is working.

Do you conclude there were likely non-obvious bad solder joints? You re-installed the same transistors, correct? Did not buy any new components to install?

 

[pucho812]

Do you conclude there were likely non-obvious bad solder joints? You re-installed the same transistors, correct? Did not buy any new components to install?

No, I did not reinstall the same transistors. In removing each one, and doing a thorough test with a DMM, I did find faulty ones that were replaced. Although with that one photo, I can see where you would suspect a wonky solder joint or two.

 

[ccaudle]

doing a thorough test with a DMM, I did find faulty ones

Interesting. Previously you mentioned a peak electronics dca75, do you find that the dca75 did not do a thorough job in testing the transistors?

 

[pucho812]

Interesting. Previously you mentioned a peak electronics dca75, do you find that the dca75 did not do a thorough job in testing the transistors?

It was one of a few different tests I did.
When hfe is 5 it’s clearly something not right when the expected hfe is 300.
But the fact the dca read it shows a working transistor

 

[JohnRoberts]

Its usually the last thing you check....

JR

 

[pucho812]

Its usually the last thing you check....

JR

I say the same about my pantaloons

 

[Walrus]

I say the same about my pantaloons