Discrete opamp trouble shooting.

[pucho812]

what’s the best way to trouble shoot a discrete opamp.
So far testing all components and all are working but the opamp is not.
I measure 1vdc lower than the supply voltage at the output. This is no bueno.

 

[Matador]

Which one? Is there a schematic?

 

[pucho812]

sadly no schematic. as this is not your typical 2520 or 990 ish design. I may have to trace one out or generate a layout and schematic.

it's a SIP 12 PIN connector that runs off +/-30VDC. It's the last little piece of a puzzle I am working on.
All components on the PCB are thru hole and they all measure as working. All resistor values measure correctly, all transistors measure as working with my transistor checker. Nothing else on the PCB other then a little grey ceramic capacitor that has the number 303 on it. All traces measure with continuity and none of them go where they shouldn't. so no PCB faults.

 

[elskardio]

Out of curiosity… are you testing the DOA out of circuit?

 

[JohnRoberts]

what’s the best way to trouble shoot a discrete opamp.
So far testing all components and all are working but the opamp is not.
I measure 1vdc lower than the supply voltage at the output. This is no bueno.

Maybe test it like an IC op amp. Is the output valid for the state of the input pins....

If the output is pegged high is the + input higher than the - input, that would cause a valid output high state.

JR

 

[pucho812]

Maybe test it like an IC op amp. Is the output valid for the state of the input pins....

If the output is pegged high is the + input higher than the - input, that would cause a valid output high state.

JR

The discrete opamp as it sits sees .xx vdc at input of either +/- of the differential input. So well below 1vdc

 

[pucho812]

Out of curiosity… are you testing the DOA out of circuit?

Yes. In circuit I get the same measurements.

 

[JohnRoberts]

The discrete opamp as it sits sees .xx vdc at input of either +/- of the differential input. So well below 1vdc

I don't follow.

What voltage is at the + input and what voltage at the - input?

If + input is higher than the - input, the output pegged high may be correct.

JR

 

[pucho812]

In circuit, I get 1vac at input, and less then .1vdc same input. Output i measure roughly 28vdc which is close to the +/-30vdc psu feeding it.
I am sure I am doing something stupid but on a working unit the same discrete opamps with same test have .1 or less at the output of each doa.

 

[Brian Roth]

@pucho812 ..... With no audio signal, what DC voltage do you measure between the + and - inputs of the DOA?

Bri

 

[pucho812]

@pucho812 ..... With no audio signal, what DC voltage do you measure between the + and - inputs of the DOA?

Bri

in circuit, around .1XX VDC I get about the same if I have the opamp stand alone.

 

[Brian Roth]

Hmmmmm....IIRC that much offset with an opamp chip will cause the output to slam to a rail.

Maybe a defective input transistor on the DOA???? Just "winging it" here.

Bri

 

[pucho812]

Hmmmmm....IIRC that much offset with an opamp chip will cause the output to slam to a rail.

Maybe a defective input transistor on the DOA???? Just "winging it" here.

Bri

I may have much less than that. I'll test again with known working doc's in circuit and see.
I thought about that. it's why I tested all of them and they all test out. I did a the usual test with a multimeter along side testing with a peak electronics dca75 which is a handy tool for checking everything you ever wanted to know about transistors.

Confirmed I have .000x dc on the inputs.

 

[Hubbub]

Testing an op-amp open loop is next to impossible. The easy way is to use maybe a 10:1 resistive divider between output and minus (-) input to close the loop. For +/-30V supplies I'd try 9K1 and 1K resistors. With the positive (+) input grounded via a 1K resistor the output should be about 10X the DC offset voltage of a good amplifier, so definitely below 100mV output. If good at DC apply 2V p-p (+1V to -1V, DC coupled) triangle or sine to the positive (+) input via a 1K resistor and see how the output swings. Output should be same shape with no visible distortion, just 10x bigger.

 

[lens42]

Hook it up as a simple follower or a -1 gain inverter. Then stick in a DC input. Any op-amp’s job in life is to make its + and - inputs the same voltage so start there. If the inputs are not the same (in a closed loop circuit) some probing should tell you why. Troubleshoot at DC first to not overcomplicate the problem.

 

[Hubbub]

Hook it up as a simple follower or a -1 gain inverter.

But be wary, some op-amp designs are not stable at unity gain and may oscillate at unity to low gain. Also, with +/- 30V power the input common mode range may be a lot less, so the 10:1 divider is much more likely to work properly and give a fair test of the op-amp.

 

[pucho812]

did some testing last night.
Where the opamp solders in on the motherboard, I have whet I believe to be the negative side of the differential input tying to the output of the opamp.
I’ll take photos and post.

 

[pucho812]

Here is a photo of there opamp in question. Here it is used in circuit. It's different than the ones on the I/O board. I have worked out the following so far.
1.PSU unregulated is +/-50V, regulated it is +/-30V and further discrete regulation to provide +/-15V for the chips. PSU behaves the same in both units.
2.I confirmed both main boards are identical and measure the same as far as voltage test points.
3.I have the pinout for the discrete opamps on the I/O board which are different then this opamp. so I was able to work out which one has signal for the in and which ones have signal for the output to LF, MF, and HF. I confirmed if I hit the mute switches for the respective output I lose signal.
I confirmed the I/O board is working without issue in a second unit. I can swap I/O boards and they behave the same in a working unit.
4. I confirmed both filter boards work in a working unit.
There is a globe safe that mutes all outputs should conditions change.
When I install these things it's when the safe mode kicks in. So it has me thinking it has to be these things.
There are 8 in total in a unit.
I confirmed from signal probing which ones are for which and I thought I had a pinout from probing but I may be mistaken. The first two that are not installed at the moment are for the LF, the 4 that are there are input from the I/O board and then output of the MF and the last two that are not installed are HF.
As both main boards measure the same, it stands to reason that the fault is with the opamps I pulled. It also stands to reason that they are the culprits as when installed it trips the protection circuit(Global mute).
I thought I had the complete pinout of these as I can measure where I get VDC, probe where I get audio before it hits these and probe where I should have audio coming out. Sadly there is no documentation to go off of so it's been a slow process.

2 out of the 4 trip the

 

[elskardio]

Based on what I'm reading, you have a second unit that's fully working?

Why not remove a single opamp from the working unit and test one by one the opamps from the faulty one?

 

[pucho812]

Based on what I'm reading, you have a second unit that's fully working?

Why not remove a single opamp from the working unit and test one by one the opamps from the faulty one?

I have. It’s how I got down to the 4 I have which cause the unit to misbehave.
It was one of the first things I did. It’s what got me a working unit out of the two units and got the second unit to get out of the global protection mode. Which it’s smarts, the output automatically mutes when it sees dc on it. Not bad for a crossover.