Midas audio console with strange oscillation problem

[mkiijam]

This is the relevant section of an AUX return module that is oscillating in a strange way.
1. Turning the SEND LEVEL control causes it to go into oscillation about 3/4 of the way up.
2. #1 is true with no INPUT signal.
3. The TREBLE and BASS controls change the pitch of the oscillation from "motorboating" to past hearing range (testing with headphones)
4. Shorting PIN 2 of IC 5 stops the oscillation.
5. Disconnecting the input trace stops the oscillation, however there is "nothing" on the AUX BUS.
6. Subbing in another channel and it works fine.
7. Placing the bad channel in the other AUX slot still has the issue. This is relevant because if there were noise on that AUX BUS it would follow the "slot"
8. This schematic is NOT the right one but is fairly close. C31 and C32 are not in the channel I am working with.

I'm just stumped at why removing the input to PIN 2 of IC5 stops the issue?

[Audiorepair]

A tone filter is not that far from an oscillator given the output is fed back to the input, so the problem is most probably in the filter circuit.

Forget the fact that if you ground pin 2 of IC5 it stops, this will probably be a red herring.

Check particularly the pots themselves,  that you haven't lost one of their connections.
Then just check every other component, and particularly the PCB traces/solder joints, as this also is a component.

[DrGeoff]

Agree with Audiorepair.
These Baxandall tone controls are shelving types and offer gain into the infra and ultra sonic regions. This can be particularly prone to oscillation if something drifts slightly.
Check for any debris on the PCB (dust, excess solder, flux etc). Also check the op amp, probably a TL072, in case the inputs have had damage over time.

[mkiijam]

All IC's are NE5534's. TDA chips were the old -school version of the NE5534, not sure what the TL is, maybe a TL071? In any case, there is a NE5534 in its place. I would suspect that the NE 5534 sub is the issue, but again, another channel that is laid out in the same way, with the same chips, preforms fine.

All the chips have the needed 47pF comp/balance caps installed, but nothing else. No PSU bypass caps, No feedback caps(!), BUT AGAIN, another channel preforms "okay."

[mkiijam]

The output (PIN 6) of IC5 (NE 5534) has the oscillations.

So unless, and this is an actually question, the filter circuit can feedback somehow and cause the oscillation, wouldn't the issue be confined to IC5?

[Audiorepair]

Don't forget that the schematic implies an ideal board construction and layout.

And note the various stages are connected via quite possibly imperfect ground and various power supplies with insufficient local decoupling.

So a full blown oscillation, possibly at full power supply levels, can enter nearby circuitry via paths you might not imagine from looking at the schematics.


The schematics are not the real world.

[DrGeoff]

All IC's are NE5534's. TDA chips were the old -school version of the NE5534, not sure what the TL is, maybe a TL071? In any case, there is a NE5534 in its place. I would suspect that the NE 5534 sub is the issue, but again, another channel that is laid out in the same way, with the same chips, preforms fine.

All the chips have the needed 47pF comp/balance caps installed, but nothing else. No PSU bypass caps, No feedback caps(!), BUT AGAIN, another channel preforms "okay."

The substitution of 5534's in that circuit can lead to scratchy pot noise (DC input offset is higher). Not sure what it would do to the gain/phase margins but it will be different. You may have to probe with a scope to find out.

[Audiorepair]

Think Occam's razor.  The most likely cause is the simplest.

The filter has become an oscillator.

Fix the filter.

[TimFox]

The NE5534 requires an external 22 pF capacitor at the "comp" pin to be stable at unity gain.
The TL071 does not use an external compensation capacitor to be stable at unity loop gain.
Has a 5534 been substituted for an original 071?

[mkiijam]

So turns out, A previous "tech" modded the console and actually made an oscillator. He misplaced a jumper one PCB hole over and sent the output of this circuit basically back into the input!