Oscillating Preamp

[smilan]

HI, I have a DIY mixer that was built for a musician with a very special instrument by another technician. The mixer works but when there's some interruptions in the mains supply like a refrigerator that start to work and drawing a lot of current the preamp start to produce this noise as you can hear in the video.

The only way to stop it is to turn off the mixer and turn it on again.
The PSU schematic attached to this message.
Do you think that by increasing the value of the 1000uF caps I'll be able to eliminate this noise, so in case of momentary low input current the caps will be able to fill the gap?

 

[Whoops]

From what I can listen in the video that noise seems like what is normally refered to as "motorboating" noise.

https://en.wikipedia.org/wiki/Motorboating_(electronics)
I don't think that problem will be solved by replacing any capacitor for higher value, I think something failed in the circuit and needs to be fixed.

Start by checking the voltage rails, are the voltages correct?

Do you have a schematic for the mixer?

 

[smilan]

I don't have the mixer in front of me now, I'll get it next week and try to see what's hiding inside...

 

[JohnRoberts]

I don't click on links, but "motor boating" is a LF oscillation commonly associated with power supply issues. A common path for the oscillation causing feedback is perhaps a collapsing PS rail that gets into the input, but there can be several different vectors (bad ground solder joint, etc).

I don't see anything obvious wrong with the PS schematic (i did click it) but didn't spend a lot of time.

Is the trigger a mains voltage dip (do lights dim?) or a conducted mains noise that somehow feeds into the mixer causing it to motorboat (which sounds like a pretty rare scenario).

If you have a scope look at the waveforms inside the mixer when it is misbehaving... If not perhaps try small experimental changes. Does it motorboat with no input connected or input shorted?

It's always something but usually the last thing you check.

JR

 

[ccaudle]

that noise seems like what is normally refered to as "motorboating" noise

Agree. The traditional way you got motorboating in amplifier circuits was that the power supply sagged when the output had a heavy load and/or high output, and that sag in the power supply affected the input stage, which then sent a signal through to the output stage, which again caused the power supply to sag, etc. You could prevent it by making the power supply regulated or very low impedance, or by making the input stage have a high power supply rejection ratio.

Since this problem begins with a sagging input voltage to the power supply, it could be something similar. Increasing the capacitance might be a brute force way to help, but hard to tell without knowing more about what the circuit power requirements are. I don't think I've ever heard of motorboating in line level gear, but depending on the size of the mixer it might be a special case if there are a lot of channels (i.e. pulling a large amount of power). Regulated power rails are usually a help against that as well, but that circuit shows adjustable regulators, which could in theory be adjusted high enough that they don't have much room for regulation if the input power drops a little. There could also be layout issues in the audio circuitry independent of the power supply that provides a feedback path, and the noise on the power supply lines is just enough to put the circuit over the edge into feedback.

Most of those problems should be visible if you can get an oscilloscope into the circuitry, and even when the problem is not occurring there may be clues such as varying power supply voltages, or varying DC operating points in any discrete circuitry that indicate it is always operating at the edge of oscillation.

 

[smilan]

I asked the owner of the mixer for some pictures from inside and outside.
The mixer contains four channels, based on AD797 mic pres and a LA3600 EQ. There's a direct out to each channel, an AUX send for tuner with selector that select which channel will be sent to the tuner.
It looks pretty straightforward design. I'll get the mixer soon so I'll be able to do some troubleshooting myself and hopefully be able to make it work good.