Slight FUZZ - either Si or Ge - is fizzy

[seedlings]

I love, love, love the dynamics of a fuzz.  Si or Ge, doesn't matter, when full bore fuzz shred, they both sound great.  I have a couple of circuits on the breadboard now, one with 2N3904 transistors and one with Russian Ge units.  I'm looking for a 'slight' fuzz or 'light' fuzz.  A clean-ish, loose fuzz with not a lot of hair.  I've been playing with R values and biasing and in/out/bypass caps and feedback resistors, and B-C caps, but there always seems to be an unnatural fizzy sound at some point in the decay.

What could you recommend for a loose, clean-ish and/or jangly fuzz?  I tried series resistor before the input cap to simulate a 'rolled - back guitar volume,' which is the sound I'm looking for at full guitar volume.

CHAD

[pinkjimiphoton]

 put trimmers to your c's, and bias HOT, not cold. the standard "4.5 volts" will always produce a fuzz, but if ya want more dynamics, you may prefer it around 5-7 volts. then if ya run the knobs about half way up on the fuzz, you should be able to dial in whatever ya need.
if ya need it cleaner, like you said, turn down your guitar a little. you can use a fixed resistor or a trim pot, but it introduces some filtering and the tone definitely will change slightly... i prefer my guitar volume, so i can just tweak it in to wherever i need it. ymmv.

[digi2t]

You might want to look at the Axis Face. Nice smooth Si fuzz, not too fussy for transistors, has a nice warm feel.

http://fuzzcentral.ssguitar.com/axisface.php

[seedlings]

Cool tips.  I hoped to go Si for better reliability with heat.  The axisface notes suggest a lower hfe units (facepalm), so I'll have to dig around to see what's in the transistor bin.  I remember reading somewhere that the lower gain silicone were prone to radio noise. But that may have been a specific build.

I hear what you're talking about Jimi with the higher voltage for V2 collector.

Went up to a 4k7 emitter resistor for Q2.  Also went up to 100k into the base of Q1, and a 47n input and 10n output, and it's closer.

CHAD