Boss CS 3 Troubleshooting

[DIY Bass]

I have a noisy CS 3 and am trying to take the noise. Noise is only there with the effect engaged, not in bypass. Using the schematic here https://www.diystompboxes.com/smfforum/index.php?topic=63909.0

The noise seems to start at pin 1 of IC2. When pin 1 gets shorted to pin 2 the noise disappeared. Based on that I replaced C5 and R6. The noise is still there unless pin 1 is shorted to pin 2. Any ideas about what to look for?

[Rob Strand]

The noise seems to start at pin 1 of IC2. When pin 1 gets shorted to pin 2 the noise disappeared. Based on that I replaced C5 and R6. The noise is still there unless pin 1 is shorted to pin 2. Any ideas about what to look for?

Shorting pin 1 to pin 2 forces the gain to zero so it basically kills the signal and the noise upto that point.

The short answer is it normal for those to compressors to have some degree of noise - excluding excessive noise due to a faulty part.   The noise actually comes from IC3 in the form of a noise current out of IC3 pin 8.   That noise current flows through the 30k resistor between pins 2 and 1 of IC2 to produce a noise voltage (Vnoise = Inoise * 30k).   Now you might think to just reduce the 30k to get less noise  that will reduce the noise but it will also reduce the signal.   For the signal there is a signal *current* coming of IC3 pin 8 and in the same way Vsig = Isig * 30k.  So by reducing the 30k you reduce the signal and the noise but you don't change the signal to noise.

The way to reduce the noise is to use a pre-emphasis/de-emphasis scheme.   This is what the MXR Dynacomp does.     

One way: You boost the highs at say IC1a then cut them at the output of the level wiper.  Unfortunately that's not so easy to get right as the impedance of the level control isn't constant.

Another way is to place a series RC network across the 33k on pin 1 IC3.  That boosts the highs coming in.  Then place an RC network across pins 1 and 2 of IC 2a.     This scheme is more or less like the MXR dynacomp scheme.

See if you can find the LM-2 and LM-2B schematics.  Both these pedal have pre-emphasis/de-emphasis.  However the way it is done isn't as clean as the Dynacomp method (partly because of the way they do the level control.)

[DIY Bass]

So what you are saying is along the lines I was thinking.  This is my son's compressor and I had noticed that it was pretty noisy - almost unusably so.  My suggestion was that he turn down the sustain and level controls to something less than a roar and that would also turn down the noise.  I know that with high compression you can get noise.  I am not used to quite so much, but it sounds like what might be going on here.  I may still go with my solution and just get him to turn it down a bit :-)

[Rob Strand]

You can remove some of the noise when there is no signal by limiting the maximum gain of the VCA (when there is no signal).   There's an old trick where the compressor has a fix gain below a certain input level.   Most limiters tend to do this by nature since they don't start compressing until the signal is above the threshold.

FWIW, I've got a Behringer Limiter and it's noisy because it has a heap of TL064 opamps in the signal path. A different problem to yours.  It's a limiter, has pre-emphasis and de-emphasis but in the wrong spots!  It has the potential of being good but, as is, it's rubbish!.


[EDIT:  Actually for the trick in the first paragraph, if I'm reading the schematic correctly the Sustain control does exactly that.  So adjust the Sustain control to gate out the noise.]

[DIY Bass]

OK, I think I may have worked out what was happening.   This is back on the pedal board and has so far been a lot quieter and better behaved (which is nice).  When I was poking around inside it I noticed that one of the audio jacks was a bit loose and tightened it up.  I didn't even pay much attention to it - just "Oh, that's a bit loose, may as well tighten it while I'm here".  Thing is that there is no ground wire to that jack - it grounds through the case.  A slightly dodgy ground connection could certainly explain what we were hearing with it in use, and now why the problem is reduced to standard hiss that you can get from a compressor.