This fuzz looks like a joke. But it's not.

[phaeton]

Well, i've tried a number of different transistors in both positions, and the differences are subtle.  Biggest difference from the 2N222As are 2N5089s- just a little bit more 'crackle' on the top.  Unless you stick it in backwards like I did the first time   
Once again, I don't really see much difference in the sound if Q2 gets pulled out of it.  Anyone else experiencing this?

My verdict is that the 2N222As sound the best.  My preference is to use .1uF caps instead of the 2.2uFs, but that's me.  Brett just discovered another one of those great-sounding, simple circuits.

[brett]

Hi
My big amp is up and running again, and it demonstrates some of the weaknesses in the circuit. 

I've decided that the gain is too low (if the gain is increased, the sustain should improve a lot).  Also, I'm thinking that it sounds too "vintage" and could do with some extra treble.  Both of these things can probably be achieved by reducing that 100k collector resistor.  I'll try 47k, 33k, 22k.. until I find a sweet spot.  I guess the 33k used in the fuzzface is an obvious clue to what will sound good.

cheers

[Dragonfly]

this is a fuzz i designed a long time back, which is based on the electra distortion...its also very similar to the "no joke fuzz", except it uses asymmetrical diode clipping (as well as tranny clipping), instead of clipping with a second transistor...

you might try "subbing" some of these values into the "no joke", as the "old school fuzz" really sounds quite good....

AC
http://aronnelson.com/gallery/album17/DRAGONFLY_OLD_SCHOOL_FUZZ

[phaeton]

Hi
My big amp is up and running again, and it demonstrates some of the weaknesses in the circuit. 

I've decided that the gain is too low (if the gain is increased, the sustain should improve a lot).  Also, I'm thinking that it sounds too "vintage" and could do with some extra treble.  Both of these things can probably be achieved by reducing that 100k collector resistor.  I'll try 47k, 33k, 22k.. until I find a sweet spot.  I guess the 33k used in the fuzzface is an obvious clue to what will sound good.

cheers

Reducing the collector resistor will mainly give it more gain, but not do much in the way of increasing treble, right?

[brett]

Hi
Reducing the collector resistor won't affect the gain, coz the transistor is running flat out anyway.  Until it gets to about 5k or less.

Reducing the collector resistor increases the voltage on the collector (V=IR, so halving the resistor halves the voltage drop across it).
But there is a side-effect:  the 1M bias resistor gets its source voltage from the collector.  Raising the voltage on the collector increases the bias current, which will increase the input impedance and "buffer" the input better, increasing treble.  For such a simple circuit, there are still lots of interactions between components (like the fuzzface).

Thanks for the Old School info Dragonfly.  It looks cool.  But in this circuit I was really trying to have one side of the signal cutoff (by the transistor), and the other side clipped (by a diode).  I'm guessing that its the cutoff that gives the fuzzface lots of its mojo (it causes the asymetrics, at least).
cheers