More Noise With Pre-Gain Control??

[brian wenz]

Hello Hello-
    Anybody replaced the "fuzz" control on a Fuzz Face with a pre-gain control?? [like Joe Gagan's]  I seem to get more noise in the circuit with the pre-gain control....anybody else notice this??
Brian.

[petemoore]

I'd like to be able to say "Yes"...and probably just should but...
 I did a FF, with 5 knobs, and it was weak and noisy...I kind of pulled it from the box and used the parts for other stuff...
 It had the pregain and cap blend knobs.
 Pregain I now use guitars vol, incaps I just pick one, or two and put a switch on one.

[Mark Hammer]

Keep in mind that when a circuit has a fixed gain, it also increases the amplitude of whatever noise is inherent to the circuit/components.  If you reduce the gain of the circuit to reduce clipping amount, then less gain is applied to those non-instrument components of the signal.  If you reduce the input signal to reduce clipping (while holding gain constant) then you have sort of the purest case of shifting signal-to-noise ratio in favour of noise.

This is frequently made evident in some compressors, and it seems like someone posts here about how annoyingly noisy their compressor is at least once a month.  If the compressor design functions by applying a fixed gain and automatically turning down the input signal going to the gain stage, then you end up with the identical scenario Brian describes, except that the input is being turned down by the circuit, not the user.  The fixed gain stage downstream simply boosts whatever comes into it, and if what comes into it is more noise than signal, so be it.  The Orange Squeezer is a perfect instance of this.

In circumstances like these, attention to input signal quality is important, as is assuring minimal noise accumulation up to the fixed gain stage.  So, shielded input cable, low-noise resistors, etc., are all helpful.  I don't think one can or should issue any sort of across-the-board "don't use fixed-gain/variable-input-attenuation designs" advisory, since many such designs can offer lots of user control and decent sound for little investment or effort.  Still, be mindful of what needs to be in place to make such designs work pleasingly.

[brian wenz]

Hello Hello--
   Thanks, Mark!   So any circuit that has a pre-gain type control [like a Bosstone] would be a little noisier then the same circuit WITHOUT the pre-gain control??
Brian.

[Eric H]

Hello Hello--
   Thanks, Mark!   So any circuit that has a pre-gain type control [like a Bosstone] would be a little noisier then the same circuit WITHOUT the pre-gain control??
Brian.

Not at all, Brian --the self produced noise of the circuit remains the same, but if you turn the input signal down there is less signal to overcome the noise --so it appears noisier (signal to noise ratio). There is a slight signal loss with the pot all the way up, as well, so, in a manner of speaking your statement is correct --but the noise level of the circuit itself doesn't change --the signal to noise ratio becomes slightly worse. I prefer Mark's more elegant response, but perhaps this will help.

-Eric

[brian wenz]

Hello Hello--
   So, basically, putting a pre-gain pot in place of the normal fuzz control [in a Fuzz Face] would not make the circuit noisier??  [everything else  being equal...trannys, caps, etc.]  It SEEMS to be a bit noisier when I do it.....
Thanks!
Brian.

[Mark Hammer]

No, it does not *make* the circuit noisier.  It just stops hiding how noisy the circuit actually is.

I've pondered suggesting to people to stick a small cap between the wiper and ground on output volume pots.  That cap, plus the input-to-wiper leg of the volume pot, would form a lowpass filter whose corner frequency would go lower in frequency as the volume is turned down.  This would also work (maybe even more effectively) in such instances as the "sustain" pot in a Big Muff or any other interstage control where you set how much signal from fixed-gain stage A goes to fixed-gain stage B rather than adjusting the gain of either stage.  The basic notion is that when risk of a degraded S/N ratio feeding a fixed-gain stage increases, treble is cut more, and when "true" signal level is hiked and improves S/N ratio, treble is restored.  E.g., with a 470pf cap between wiper and ground, on a 100k BMP sustain/distortion control, treble is rolled off around 3.7khz when the control is turned down most of the way (wiper set at 90k/10k resistance).

[brian wenz]

Hello Hello-
   Bingo!   Instead of the pot adding racket  it's just making the noise that's already in the circuit more obvious.   I noticed that the noise level increased when I'd stick a fixed resistor in the front of the circuit, too.
Thank you for the cap idea!  I'll mess around with some different values and see what happens.
Brian.