How to test a Neve BA438 opamp?

[Golgoth]

Hi everyone,

I am having issues with BA438 opamps.

The audio is coming from a T1452 followed by a 10k pot to the BA438 opamp, then going to BA283 + LO1166.

When I remove the opamp and plug a jumper between the input and output pins (even tho the audio is extremely loud, I just did that to confirm the card was fine) audio is coming through.
When the opamp is plugged in I get almost nothing, a few clicks, dark crunchy bits of sound, with very low gain even with the pots cranked all the way.

The +24V rail is doing great, grounding is doing great (zero hum, and very little hiss even with the BA283 at full gain with the jumper between the pins).

What could I check next? I'd like to be able to test the opamps themselves if that's possible.

Thanks for your help!

 

[Coop]

You could try measuring the voltages at the transistors and compare them to a working BA438. Could be a cold solder joint. Had that on a BA283 before.

 

[Golgoth]

You could try measuring the voltages at the transistors and compare them to a working BA438. Could be a cold solder joint. Had that on a BA283 before.

Thanks again for the help with the BA283, it works like a charm!

I unfortunately don't know if I have a good one or not, when I received the unit I am restoring I decided to dismantle everything for cleaning before turning it on as it was very dirty inside.

The underside of all the 438s I have look very good tho. I believe I'll need to select one of them to disassamble and test the components individually..

 

[Coop]

Great to hear that

Well I would just test everyone out and see if one passes audio with no issues. Then check the voltages and compare it with the others. Not much that can go bad. Maybe just a cap or a resistor.

 

[Golgoth]

Great to hear that

Well I would just test everyone out and see if one passes audio with no issues. Then check the voltages and compare it with the others. Not much that can go bad. Maybe just a cap or a resistor.

I tried all 8 of them without success unfortunately.. I think I'll go to the electronics store and get everything needed to rebuild one of them.

It will be handy indeed to have one known good unit to test the others against!

 

[Coop]

Mhm, that's weird that all 8 of them are dead. You are sure there is nothing else wrong with the board the amp plugs into?

 

[Golgoth]

Mhm, that's weird that all 8 of them are dead. You are sure there is nothing else wrong with the board the amp plugs into?

Yes that's what I was thinking too, they seemed ok to the previous owner and he is pretty trustable. I doubt the package was struck by lightning during transit.

Regarding the card, I measured correct B+, and successfully passed audio with a jumper between the input and output pins. I will try to reflow the pins solder joints. I remarked these opamps have a thread for a screw, does that mean they need to be seated very tighly and deeply inside the pins?

I don't really know what to check next on the card..

 

[Coop]

Mhm, well I would reflow all solder joints. I somewhere read that due to the heatsinks of the metal camp transistors, those solder joints are prone to get bad over time.

You could also test all the transistors with a multimeter in diode mode.

 

[Golgoth]

Mhm, well I would reflow all solder joints. I somewhere read that due to the heatsinks of the metal camp transistors, those solder joints are prone to get bad over time.

You could also test all the transistors with a multimeter in diode mode.

I will do both and report back!

 

[Whoops]

I tried all 8 of them without success unfortunately..

If it was just one, I would change the 2 tantalum capacitors and then desolder the 5 transistors and test them.
But as you are having the same problem with 8 of them, I thing the problem is elsewhere and not in the Opamps.
I don't see how could 8 opams fail and fail in the same way

 

[Golgoth]

If it was just one, I would change the 2 tantalum capacitors and then desolder the 5 transistors and test them.
But as you are having the same problem with 8 of them, I thing the problem is elsewhere and not in the Opamps.
I don't see how could 8 opams fail and fail in the same way

I am currently testing the components on one of the opamps. Thankfully BC441 and BC461 are okay (they seem pretty hard to find in the UE). I'll check the tantalums next.

What else do you have in mind about what could be wrong if not the opamps?

 

[Matt Syson]

As some Neve circui#ts use multiple pins for either ground or supply it is likely a ground or 'decoupled' supply is missing for the 438 positions. Of course there is an output capacitor involved that is not on the sub board and that could be at fault or a damaged track. the 438 shows 2 'input' pins although in reality one is a connection point to add a gain set resistor (and capacitor). This meant that the SAME little boards could be plugged into different positions in the signal paths and the gain that was required was 'pre set' by components NOT on the sub board. Of course most chip OP amps are 'effectively similar in that the circuit gain depends on external components and circuit design.
It is highly unlikely you would have 8 sub boards with the SAME fault.
Matt S

 

[Golgoth]

Salut @Matt Syson ,

Thanks a lot for these informations. I just finished checking every transistor and capacitor on one opamp and they are all fine.

What would you check on the board?

 

[Whoops]

I am currently testing the components on one of the opamps. Thankfully BC441 and BC461 are okay (they seem pretty hard to find in the UE). I'll check the tantalums next.

For old tantalum caps I don't test them at all, I just replace them all with new ones.

What else do you have in mind about what could be wrong if not the opamps?

Nothing more, just what I wrote. Tantalum caps and checking the transistors, thats it.
Your problem doesn't seem to be related to the Opamps at all

 

[Golgoth]

Nothing more, just what I wrote. Tantalum caps and checking the transistors, thats it.
Your problem doesn't seem to be related to the Opamps at all

Matt made a very good point regarding the card and the multiple pins being the probable issue. Now that I've checked the components and that they all are validated as good this is what I need to check next.

 

[mjrippe]

Here is a drawing with the voltages you should see at each pin. Also the schematic of the 438 on the off chance you do not have it already.

 

[Golgoth]

Here is a drawing with the voltages you should see at each pin. Also the schematic of the 438 on the off chance you do not have it already.

That's very precious information, thank you very much!

Here are my measurements according to your numbers:

6 = 23.3V
5 = 12,2V
3 = 0V

1 = 0V
7 = 8,9V
2 = 10,5V

7 & 2 seem a little on the low side correct?

 

[mjrippe]

That's very precious information, thank you very much!

Here are my measurements according to your numbers:

6 = 23.3V
5 = 12,2V
3 = 0V

1 = 0V
7 = 8,9V
2 = 10,5V

7 & 2 seem a little on the low side correct?

They do indeed. Are you trying a few different 438s at different positions in the board? If so, and the readings are the same, I would look for leaky caps on the motherboard.

 

[Golgoth]

They do indeed. Are you trying a few different 438s at different positions in the board? If so, and the readings are the same, I would look for leaky caps on the motherboard.

I did, the readings seem to follow each opamp..

I will replace the caps and resistors on the motherboard as it is only a dozen components in total and can not hurt. I have also bought everything needed to replace every component on a BA438 and will do that in a second time if recaping the motherboard does not help.

 

[Golgoth]

@mjrippe I replaced all caps and transistors on one opamp. We can consider it as good. I then replaced all the caps on the motherboard where the opamps are plugged and retook measurements.

6 = 23.3V
5 = 12,2V
3 = 0V

1 = slowly dropping from 7V (new problem!)
7 = 8,9V
2 = 10,5V

This time the values are exactly the same for both opamps on the card (newly populated one and randomly selected one on a two opamp card). But pin 1 now has voltage..

What could it mean? It does not seem to come from the components being faulty.

The card is not connected to anything except V+ et V- (V- being connected to the chassis in the end).