Troubleshooting noisy audio mixer preamp.

[Yamin]

Hey guys,
I recently got a soundcraft EFX8 mixer in for general servicing. All the faders were in pretty bad condition so almost all the channels were scratchy and noisy. Replaced them all and all the channels are working fine except for one.
So the channel in question produces a high frequency hiss, its not consistent and changes in range. So here are things I have done so far,
1. Replaced the gain pot(P500A).
2. Jumped the connection on the "insert".
3. Replaced the quad op amp IC500A.
None of the above rectified the problem. Before replacing the opamp I did scope around and here is what I found- I can clearly see the trace jumping around at pin 1 of IC500. I checked pin 2 and pin 3 and I did not measure any anomalies there. I have done other troubleshooting things like measuring the voltage rails, looking for dired joints, checking for proper ground connections and tap test. I have also compared measurements against a working channel but can't see anything obviously wrong.
One other thing I would add is while I had the op amp desoldered and out of the circuit I did scope around pin 2 and pin3 pads while I was running a 1khz test tone. The DC offset at pin 3 was 2V higher than pin 2. I am not sure if that's significant or not.
I was hoping you could give me some other things to look for. I have attached a pdf of the schematic of the input section.
Thanks

[magic]

Regarding differential balance, pin 3 is loaded by R507 to ground while pin 2 is loaded by R508 to pin 1, which was open circuit. Try again with the chip installed - pin 1 should be at ground and the inputs should be equal.

You could check the transistor circuit for balance: PNP bases should be equal, emitters should be equal to a few mV max.

Possibly a failure of some resistors could cause one transistor to be starved of bias, increasing noise. One or more transistors may be bad. Perhaps a diode or electrolytic capacitor developed high leakage current.

Maybe try shorting pin 2 or pin 3 to ground. This will likely cause output offset, but if it makes the hiss disappear, you will know which signal half is to blame.

[Yamin]

Regarding differential balance, pin 3 is loaded by R507 to ground while pin 2 is loaded by R508 to pin 1, which was open circuit. Try again with the chip installed - pin 1 should be at ground and the inputs should be equal.

You could check the transistor circuit for balance: PNP bases should be equal, emitters should be equal to a few mV max.

Possibly a failure of some resistors could cause one transistor to be starved of bias, increasing noise. One or more transistors may be bad. Perhaps a diode or electrolytic capacitor developed high leakage current.

Maybe try shorting pin 2 or pin 3 to ground. This will likely cause output offset, but if it makes the hiss disappear, you will know which signal half is to blame.

I took some voltage measurements today. Checked against a working channel. I took measurements around the transistors and all the measurements which I took were identical. It was a bit hard to reach since the smd transistors were very close proximity to the input jacks. So I might have missed some measurements.
However I found one major variance - that is at pin 1 of IC100 (sorry in my first post I grabbed a screenshot of another channel). I measure -0.4V of offset. The test signal is also offset by that amount. All the other channels were way below that some were in the order of -0.02V the highest I measured in a good channel was -0.1V. I don't know if this is significant or not, what could be the fault that is causing this discrepancy.

Could you also please explain the procedure you suggested, shorting pin 2 or 3 to ground (marked in the picture)? What am I looking for and what could I deduce from the result? Do you mean a direct short to ground without any resistors? Would that affect the IC?

Thanks again.

[magic]

Sorry, maybe not so fast. Grounding pin 2 is nonsense, it will make the TL074 work as comparator. Grounding pin 3 makes an ordinary inverting amplifier, and probably causes it to output several volts DC offset, maybe even hit the rail.

What is the normal voltage on pins 2 and 3 and what are the  -V1 and +V2 supplies?

However I found one major variance - that is at pin 1 of IC100 (sorry in my first post I grabbed a screenshot of another channel). I measure -0.4V of offset. The test signal is also offset by that amount.

What do you mean by "test signal"? The final output of the preamp?

An offset here could come from a few sources. Inequality of any of the resistors pairs: R100,R102, R101,R105, R103,R106, R107,R108 or maybe bad transistors, for example very low β.

[Audiorepair]

A common problem with such circuits that employ them are the protection diodes, D103/104.
These often get damaged, and become noisy or intermittent causing DC problems.

Remove them both and see if that cures the problem, they are not required for the circuit to work properly.

[Audiorepair]

Also, the forward voltage of the faulty diodes measured in circuit will often be different to all the other good ones on the board, further reasons to remove them and see.

[Yamin]

Sorry, maybe not so fast. Grounding pin 2 is nonsense, it will make the TL074 work as comparator. Grounding pin 3 makes an ordinary inverting amplifier, and probably causes it to output several volts DC offset, maybe even hit the rail.

What is the normal voltage on pins 2 and 3 and what are the  -V1 and +V2 supplies?

However I found one major variance - that is at pin 1 of IC100 (sorry in my first post I grabbed a screenshot of another channel). I measure -0.4V of offset. The test signal is also offset by that amount.

What do you mean by "test signal"? The final output of the preamp?

An offset here could come from a few sources. Inequality of any of the resistors pairs: R100,R102, R101,R105, R103,R106, R107,R108 or maybe bad transistors, for example very low β.

The voltage rails are +/-15V. I can't remember the voltages at pin 2 and 3 at the moment but it was the same as other channels (I shall double check that). At pin 1 I measure -0.4V DC with or without and input signal. All the knobs are set identical to a working channel.

Is there are way to diagnose what is causing the offset, is the only option to remove all the components and measure them individually?

Thanks

[Yamin]

A common problem with such circuits that employ them are the protection diodes, D103/104.
These often get damaged, and become noisy or intermittent causing DC problems.

Remove them both and see if that cures the problem, they are not required for the circuit to work properly.

Also, the forward voltage of the faulty diodes measured in circuit will often be different to all the other good ones on the board, further reasons to remove them and see.

I shall try that, the diodes are in a place very hard to reach. I would have to desolder the input connectors to take the diodes out. Is there a way to double check that for instance with freeze spray?

Did you mean the forward voltage measured in diode mode with the DMM?

Thanks

[Audiorepair]

Freeze spray doesn't always work, but does often enough to be worth having a go first.

And yes, using a typical meter in diode mode.
You might be able to measure them easier by measuring  between the common -V1 and the electro caps instead, the diodes are connected to these points.

[magic]

Is there are way to diagnose what is causing the offset, is the only option to remove all the components and measure them individually?

Power off and measure resistors in circuit.

Power on and measure voltages at all terminal of transistors.
They should be about equal on all channels.
They should be very equal for both differential branches of any channel.

Output offset means that either some resistors connected to opamp pins aren't equal, or currents through the two Sziklai pairs aren't equal.

[schmitt trigger]

Back when I repaired audio equipment, a “hissing” channel almost always would be traced to a defective semiconductor.
Therefore I agree that the protection diodes would be the first suspects. However a leaky transistor junction would have the exact same effect.

And I hear you loud and clear. Many board manufacturers believe that their products will NEVER require to be repaired. And what should be a straightforward troubleshooting job becomes a major disassembly task.

[magic]

Hmm, without any disassembly you can measure voltages on CN101 terminals. If they are significantly negative, replace the diodes. If they are significantly positive, something is amiss with the PNPs or maybe resistors.

Does the 0.4V offset change if you turn the gain pot?

Electrolytic capacitors may leak too, and I see that they are used for mic signal coupling. Was phantom power enabled or disabled in your tests? Does it make any difference?

[Audiorepair]

Most of these circuits now use 4 protection diodes, one between each input electrolytic and both power rails, this only uses 2 to the -ve power rail.
Cheapskates.

This wasn't always that common, the Mackie VLZ didn't have them, for instance, which made it a pain to find which of the 4 transistors were faulty.
Nowadays, these transistors seem to be protected by the protection diodes. I can't remember the last time I had to replace one, though I have replaced lots of protection diodes.
Which is a lot quicker and easier, especially as they tend to show up using diode mode in circuit.
And sometimes I just change them anyway if they are accessible, it's a 5 minute job to give yourself a known known while fault finding.

I have also found the first and/or second op-amp downstream of this can also be damaged. 
I don't know if perhaps the diodes get zapped and then later the op-amp, or whether everything happened in one go.


So, now I pretty much never suspect the transistors, but go for the protection diodes first, then the op-amps.

But that's just my experience of things.



Edit: Oh, and electrolytics always get checked anyway, cos we know what they're like.

[Yamin]

Hmm, without any disassembly you can measure voltages on CN101 terminals. If they are significantly negative, replace the diodes. If they are significantly positive, something is amiss with the PNPs or maybe resistors.

Does the 0.4V offset change if you turn the gain pot?

Electrolytic capacitors may leak too, and I see that they are used for mic signal coupling. Was phantom power enabled or disabled in your tests? Does it make any difference?

I can't recall that CN101 being very "out" then rest of the channels. Sorry in advance for the noob question, but I'm trying to figure out how a leaky diode can make CN101 got negative.

The 0.4V offset does not change much with gain knob. Its same for all other channels too not much change when turning the gain knob.

[Yamin]

Most of these circuits now use 4 protection diodes, one between each input electrolytic and both power rails, this only uses 2 to the -ve power rail.
Cheapskates.

This wasn't always that common, the Mackie VLZ didn't have them, for instance, which made it a pain to find which of the 4 transistors were faulty.
Nowadays, these transistors seem to be protected by the protection diodes. I can't remember the last time I had to replace one, though I have replaced lots of protection diodes.
Which is a lot quicker and easier, especially as they tend to show up using diode mode in circuit.
And sometimes I just change them anyway if they are accessible, it's a 5 minute job to give yourself a known known while fault finding.

I have also found the first and/or second op-amp downstream of this can also be damaged. 
I don't know if perhaps the diodes get zapped and then later the op-amp, or whether everything happened in one go.


So, now I pretty much never suspect the transistors, but go for the protection diodes first, then the op-amps.

But that's just my experience of things.



Edit: Oh, and electrolytics always get checked anyway, cos we know what they're like.

I'm thinking its time to take out the input jacks, only way to better probe and take out components. Also on close inspection there seems to be some kind of white dirt substance, like dried up flux collected there. All over the input board actually.

[magic]

I can't recall that CN101 being very "out" then rest of the channels. Sorry in advance for the noob question, but I'm trying to figure out how a leaky diode can make CN101 got negative.

By leaking current from R109/R110 to the negative rail.

The 0.4V offset does not change much with gain knob. Its same for all other channels too not much change when turning the gain knob.

Sorry, I'm an idiot again. Of course there is no DC gain because the pot is AC coupled.
However, do you have access to the terminals of C100, any voltage across it? On other channels?
It should be very low.

[Yamin]

However, do you have access to the terminals of C100, any voltage across it? On other channels?
It should be very low.

Yes its accessible from one end of the R104 resistor and the emitter of TR101. I'll check this.

[Yamin]

So guys an update, I did some further troubleshooting.

However, do you have access to the terminals of C100, any voltage across it? On other channels?
It should be very low.

I didn't measure any voltage across this cap and its the same for other channels too.

Also, the forward voltage of the faulty diodes measured in circuit will often be different to all the other good ones on the board, further reasons to remove them and see.

I checked the diode and it shows the same forward voltage drop as the others. Same reading in the opposite direction (comparing with other channels).

Some other findings:
1)Collector of TR100/101 and base of TR102/103 measures at -10.01V on the noisy channel. All the other channels at the same point measures at -9.96 to -9.98V. Gain knob or injecting a signal or not didn't have any effect. Again I'm not sure if this is significant.
2)I placed the scope at base of TR100 and I didn't see the hiss/pop noise there. However at the collector of that transistor and base of TR102 I sort of saw the noise signal but it was very hard to see and not as significant as seen from the output of the opamp.
3)Placing the probe at base  of TR102 produces a noise that the same on all channels. But on the bad channel the hiss/pop sounds seem to get louder if I tap that point from the tip of the probe.
4)Maybe a pointless point but - I measured varying of voltage (millivolts) at the negative lead of C115 with respect to ground. This wasn't the case for the good channels. I did add a cap parallel to it and it didn't do anything with the noise. Besides, signal going to the "insert" jack has noise on it.

I have a small video clip of the noise, here is the link:
https://www.youtube.com/shorts/DaNaWfkYpPM

The owner of the mixer has got a gig coming up in couple of days, he is ok sacrificing one channel. But I would like to identify the fault for future reference too. Hopefully I get to diagnose the problem with your guys help.

Thanks for the help again.

[themadhippy]

Being that its souncraft id check,then double check that 10 ohm resistor

[Yamin]

Being that its souncraft id check,then double check that 10 ohm resistor

Funny that you mention it. I did check that,measures around 11 ohms, others measure at around 10.5ohms. I'm guessing that ok. Are they known to go faulty? Does it cause this kind of noise?

[MarkT]

Its definitely suspicious to be 10% out.

[themadhippy]

Are they known to go faulty

personally ive had 2 soundcraft desks develop faults that was down to a 10 ohm resistor and know of a couple  of others who have had similar problem.However these were all on desks from a good few years back.

[Yamin]

Its definitely suspicious to be 10% out.

Are they known to go faulty

personally ive had 2 soundcraft desks develop faults that was down to a 10 ohm resistor and know of a couple  of others who have had similar problem.However these were all on desks from a good few years back.

Thanks guys something to keep in mind for future reference as well. I wanted to ask will a faulty R104 cause this kind of noise?