Hi, a few days ago, all of a sudden my home audio mixing desk went into a spasm. A loud buzz and the some of the meters in clipping and reducing the volume on anything and disconnecting everything did not help. Turning it off did.
Now it's already over 20 years old and the faders are not to good anymore anyway, so i ordered something to replace it with. It's a bit bigger and i will need to modify the cupboard it is sitting on, but what to do, it was the closest match i could find.
Anyway, i thought i'd take the old one apart and see if it could be repaired. Normally i don't do audio much. Not enough information and not enough time to learn. A visual inspection of the PCB turned up nothing strange, a bit of dirt which would easily explain the crackly faders, but nothing obvious like bursted capacitors or visually burned IC's
So my next though was, 'The Op-Amps' and just start removing them until the problem goes away. I should check the PSU input i supposed, but the 18VAC external adapter works fine, and there is power. I will check it with an oscilloscope just in case, but the defect didn't happen during a power cycle (i leave power on most the time anyway) and it was rather sudden. (and loud)
I was wondering if anybody has any other suggestions or ideas on what may have broken.
Difficult to guess from a distance, but if the adapter output is AC, I might suppose that there would be a rectifier bridge inside the unit and that one of the diodes in this rectifier bridge has suddenly failed short and you have AC leaking into the circuit..
It may also be that the Electrolytic capacitor(s) that smooth the rectified DC supply have gone dry and shorted or are electrically leaky. They may not necessarily exhibit any obvious external signs of distress. However, rectifiers don't usually fail unless they have been put under some stress, often by failing capacitors. This is assuming the unit has a linear rather than a switch mode PSU.
Does it sound like mains hum, or more like motorboating?
Perhaps best to avoid Of course this is all mere speculation at this point.
If you start pulling socket-ed OP amps, make sure you know where they came from so you know where to put them back. They may not all be the same!
Check the power supply voltages, I'd expect 3 rails, probably 0V then plus and minus the same voltage, probably 15VDC. Check they are clean with an oscilloscope.
Isolate each channel to see if you can tell which one is causing the problem, or maybe it's after the mixing stage.
Like any other fault finding break the problem into parts and locate what works and what doesn't.
Do you have something that will generate a sinewave you can feed into the inputs to test them?
Hint: The op-amps normally do not fail. The weakest link is the Power Supply and if it is HUM you are getting that would make me about 95% sure it was the main and possibly subsequent capacitors in the power supply. Your channel output level LEDs appear to indicate this.
More information such as make and model would help you get a better answer.
An old tech told me once:
Half of the faults can be analysed with your eyes and your nose.
I agree than investigating the supply would be the first step.
Leo..
Oh thank you, i spend some time looking for them, but kept going around in circles.
That doesn't matter all that much.
didn't find the power supply rails within that document unfortunately.
The closest i can describe it is the moment you plug in a hot-lead into a guitar, and the tip touches the ring, but as a constant sound.
Well it could be. The PSU is 18VAC, and there is a 15v DC rail i think. 48VAC for Phantom and a small 5v rail for the FX and probably SPDIF . How likely is it those diodes in the rectifier bridge actually breaking though ? They used 4x 1N4002, for both the 15v & 5v each. 4x 7815 in a TO-220 package with a big heatsink and a 7805. what can break.
They are soic-8 package 4580d all of them. easy to find and really pulled of 4 just to see how hard it would be and it's really easy enough. The markings on the PCB are clear for rotation. Should be ok.
That is my suspicion now. The faders are analog, so anything should just be reduced when i close the faders.
yeah but the other parts aren't all that likely either. But clearly the Power supply rail (and the 15v+- in particular) are a suspect.
Well it's more like holding guitar lead with your hand, but then really loud, which made me think that 1 of the Op-Amps was receiving a max volume input so to say, almost directly from the power rail.
I don't think i have any space for it, but it does fall into 'general electronics' Knowledgeable people may be able to help.
Which is mains hum picked up by your body and transferred to the input, so if it's really like that I'd suspect the PSU first.
This moderator doesn't have a problem with non-Arduino electronics questions in General Electronics, especially from people who have contributed plenty to the forum.
Is the hum at a specific frequency? Line noise pickup is generally at line frequency (50Hz/60Hz depending on location), power supply hum is often at twice the line frequency because of the bridge rectifier.
I would start by looking at the voltages at the input of the voltage regulators with an oscilloscope, to see if there was excessive hum. A bad capacitor, or bad connection to a capacitor, in the power supply section can cause that.
A bad ground connection somewhere can also cause hum because anything plugged into the mixer will act as an antenna to pick up line frequency.
If the problem is intermittent, does flexing the unit or moving it around cause it to re-appear? That might indicate a cracked solder connection or corrosion on a plug-in connector inside the unit.
Ok so once i found some time and charged the battery pack i use for the oscilloscope the results are such that i think the PSU at least for the 15v +- DC are not the issue.
There is a slight tremor, which can easily be attributed to the big capacitors having worn out a little over time, but nothing that would explain the loud noise and clipping that occurred all of a sudden.
So somehow i don't think it's caused by something from the PSU.
Checked the 5vDC for the effects unit, and found a cleaner signal of course, but stepping down from 18VAC that was to be expected.
If anyone thinks i should replace the Big 1000uF on the input to the 7815 & 7915, that is probably better. As well as changing the 47uF on the output of those, but i think that before that, i should identify the cause.
Of course measurements were done without load, adding a load may change things a little.
electronics tech since 1964. my first thought is that the biggest capacitor in the power supply failed. this is most often the case in these situations. my philosophy is that if it has not failed it will soon, and just pop a new one in.
Well there is 2 really big ones one for 15v- and one for 15v+. Still aren't they just losing capacity over time rather than instant failure ? I am actually going to have to order some i don't have the 35v spec (only 25v which will most likely fail rather soon)
were you measuring across GND and +15, GND and -15 or -15V and +15V?
obviously this has been done with essentially a 1:1 probe. was the oscilloscope input set to 1:1 as well?
With those flat lines at the top, the trace doesn't look like a typical smoothing nor ripple curve. Nor does it look like a typical SMPSU output. It looks more like an inverted (negative going) and somewhat distorted half wave rectification trace. The inversion, of course might be explained by (looking at the photo) the connections to the PCB being reversed?
One rather other odd thing (as mentioned by @david_2018) is that the frequency appears to be 100Hz* which, assuming a 50Hz mains, one might expect from a full bridge rectifier, but not a half wave rectifier. On the other hand half wave rectification might make sense for a -15V-0+15V supply. Something doesn't quite make sense here.
At a setting of 20V/Div, those peaks are at approximately 4V, and if that were to be an AC coupled measurement of ripple, then it is significant and not just a "slight tremmor". One would normally expect ripple to be less than 100mv. On the other hand, if its a DC coupled measurement of the output from the rectifier, then it looks more like it might be one half of the 5V supply rather than a 15V line.
Is there a bridge rectifier or are there diodes inside the power adapter unit? I don't see any that PCB, only what looks like a regulator.
*6 pulses per 60 ms (5ms x 12 graticule squares) equals 100 pulses per second or 100Hz.
I leaned an important lesson a long time ago: fix the obvious problem first. "That can't be the problem so I'll look elsewhere" is pretty much guaranteed to waste a lot of time. The output from the 2 regulators should be dead flat, not all bumpy like that.
Change the capacitors, check the diodes and replace if faulty.
No, don't bother testing the capacitors, at that age they have either failed or will fail soon.
I can see 4 diodes on the PCB near the capacitors, slightly left of center.
Not sure about this unit, the service manual that was posted for the USB version of the mixer shows a power input that is 18V AC each side of a center-tap, with full-wave rectifiers for the + and - side of the power supply.
