Simple FET compressors (Orange squeezer) - hum and hiss

[scintillation]

I built an Orange squeezer compressor recently, to be more precise, the Madbean cupcake version.

Searching through this forum, a common compression question seems to be "it is noisy", a common response is "that's because it's a compressor".

The net result, someone building a compressor could just assume, if their unit is noisy, it's meant to be like that.

From my own observations, adding a high frequency cap in parallel with the gain setting resistor on the op amp (I had a 300 pF on the bench so just put that in) removed the hiss. More of note though, a 1000 uF on the power supply line removed all the LF noise, 220 uF was an improvement, but still a noticeable improvement increasing the value to 1000 uF. I use a linear power supply with individually isolated outputs for my pedals. I am also using an NE5532, but haven't done any comparisons with others. Now, I can't really tell when my compressor is on or off (when I'm not playing anything).

I thought I would add this, in the hope that someone doesn't settle for a noisy compressor because the internet told them it would be noisy.

[Mark Hammer]

It's not so much that compressors themselves are noisey.  Rather, they can't tell the difference between input noise and simple pianissimo.  They apply gain to residual noise at the input as if it were a low-level signal.  If you feed a compressor with a quiet signal, then throwing 20db of gain at that signal won't make the result objectionable.  The problem occurs when the input signal has a basic 60db S/N ratio and then you throw 20db gain (or more) at that noise.

Stated another way, the noise does not come from the compressor itself, but what the compressor does with any noise you feed it.  This is why I'm such a fan of the SSM2166 chip.  It includes downward expansion, such that residual noise is simply not amplified.

[scintillation]

I was reticent to over simplify my "common question and response", but inevitably I did. Thanks for making point I tried to make much clearer.

[Mark Hammer]

No problem.

And at the risk of repeating myself a little too much, this is precisely why it is ill-advised to stick a compressor *after* a high-gain device.  If your "brown sound" pedal throws a gain of 250x at the input (and any residual noise accompanying it), and you feed that to a compressor, when you stop playing, that now-amplified residual noise will get even more prominence.

The secret/trick to noise-free compression is to do whatever it takes to provide the compressor with the quietest possible signal you can, whether by virtue of where you insert the pedal in your path, or what you do to any devices that precede it.  Decent cable probably makes a bit of difference.

[PRR]

> adding a high frequency cap in parallel with the gain setting resistor on the op amp (...300 pF...) removed the hiss.

300pfd against 220K is 2.5KHz high-cut. Yes, that will remove a lot of hiss. It also damps the high end of guitar (we often want overtones to 5KHz to get through the speaker).

300pFd (against 220K) is probably a good compromise for many people. May not be best (too mellow) for ear-burning zing-scream sounds.

> 1000 uF on the power supply line removed all the LF noise

If the power supply were perfect, that would not be needed; but we often use imperfect power supplies. IMHO 1,000uFd is not too large for a 1-opamp thing if it does the job (easier than beefing-up supply filtering).

Considering the low current demand, R12 10 Ohms could probably be upped to 100 Ohms. This gives very little extra voltage drop, but gives C8 ten-times the leverage aganst crap coming in from +9V. (Stock 10 Ohms against 47uFd low-passes for 350Hz, which hardly seems ample in a world of 60Hz and 180Hz disturbances.) (100r and 1,000u works down to 2hz, so has a lot of action in audio bass range.)

Which comes around to what we are saying. If you "use" a compressor, soft sounds get louder, and you have to be extra-careful about unwanted soft sounds.

In the days when tubes did most pro audio, often most of the tubes could use AC heat without a hum problem, even mike-amps; but compressors/limiters often had to use DC heat to keep hum away.