Ok...I finished breadboarding and tweaking this thing and it turn out it is (of course) set up similarly to the FZ-1A or Fuzz-a-Tort (same fuzz control)
I won't go through all the details of my tweaking, but simply:
Q1: If you use something leaky or with low gain, the fuzz will be dirtier and lower volume. I actually wound up with a much higher gain one than I would've expected. The lower gain ones simply didn't give me the smoother, saturated fuzz I got from the 2SA221 (hfe 200, leakage 175uA)
Q2: this was the hardest one to choose. If you use a leaky one here, it basically nullifies the fuzz control. I threw a leaky one in here and it was like the fuzz was full all the time. the fuzz control had virtually no effect. I tried all different gains & leakages until I got a good sweep on the fuzz control. I wound up with a 2SB165 (hfe 100, leakage 100uA), but I could not choose Q2 until I had a correct Q3.
Q3: I started with a transistor with leakage ~ 100-200uA. the collector voltage was -1.55v and the fuzz control was useless. I got a semi-gated dirty fuzz and the the fuzz control was only useful in the last 10%. the first 90% was very low volume and dull and the high E was super gated. it wasn't until I threw in the 2SB376 (yes, also my savior for Q3 on the Burns Buzzaround...hfe 70, leakage 300+ uA) that this came to life. Immediately, Q3's collector voltage dropped to -.7, the fuzz was loud, smooth, killer with good sustain & decay and the fuzz control had a good sweep, but was even better after taking the time to choose Q2.
A couple notes...according to Small Bear, I'm right on the money for Q1's emitter voltage, -.6.
SB recommends -1v on Q2's collector. I could go as high as 1.15v, but the best fuzz was somewhere in the -.75v to -.90v range. I set it to -.85v by changing the 10K off of Q2's collector to 5.1K
Once Q2's collector was set to -.85v, I set Q3's collector to ~ 1/2 supply. It actually sounds decent all the way from -.3v to -1.1v, but indeed 1/2 supply was the best.
Everything is killer, except for one flaw. The circuit only works well with a leaky Q3, but with the leakage comes the typical hiss ("shhhhhhhhhhhhhhhhhhhh...") of a leaky transistor. I can cut down on some shhhhhhhhh... with a 200uF cap across Q3 B-C and completely remove it with a 500uF cap, but there is sacrifice to the fuzz quality.
Is it a catch 22 or is there a solution?
final version (so far):
