Dino has put a lot of work into this project over the years and I can't begin to imagine how tantalising it must have felt to have the parts side of the PCB barely out of reach! One of these pedals has somehow landed into my stash, and so I'm happy to be able to share a 'true' schematic with the community here.
I have to say that I'm actually somewhat disappointed with the sound of this pedal though. It's basically a misbiased Fuzz Face kind of effect, and there's no excuse for professional pedals sounding like that in the 1970s! With the multimeter I found a suspiciously low Q2 collector voltage and so I was convinced that something was wrong with the pedal. Clipping a trimpot onto R6 allowed me to increase the supply voltage going to the transistor, and bringing the collector voltage up to 6v made a drastic improvement to the sound — fatter and more sustain.
I measured all the relevant resistors in-circuit and they were all more or less still in spec, so I consulted a large language model and described the schematic and the parts values etc to it. If you look beyond generated pictures of shrimp Jesus, this technology is fucking amazing! The computer agreed that Q2C was low, and suggested various more things to troubleshoot. Eventually it suggested that I simulate the circuit with LTSpice, which is something I've never even thought about attempting before. This LLM just straight up taught me like I'm a complete idiot to draw up and test the circuit in software, and it even wrote out a model 2N3565 transistor in the format for me to import into LTSpice and use for the test. What a time to be alive!
The so-called AI plotted my readings with the simulated ones out in a handy table for all of us to nerd over, and it's led me to believe that this Select-a-fuzz circuit might just be flawed. I'd be interested to hear y'all's thoughts as well though. Apparently 2N3565's come in a very wide range of gains and so the LLM and I have agreed that Q2 might just be too low for the circuit and that replacing it might improve the sound. With it being a collectible vintage pedal, however, I'm thinking of just piggybacking the bias resistor instead.
Code: Select all
Transistor Node Measured (V) Simulated (V) Difference (V)
Q1 Base (B) 0.62 0.63 -0.01
Collector (C) 1.37 1.38 -0.01
Emitter (E) 0.00 0.00 0.00
Q2 Base (B) 1.37 1.38 -0.01
Collector (C) 2.41 3.48 -1.07
Emitter (E) 0.71 0.68 +0.03
Q3 Base (B) 0.54 0.52 +0.02
Collector (C) 7.51 7.97 -0.46
Emitter (E) 0.00 0.00 0.00
(NB: Q3's bias varies depending on where the tone pot is set. My own measurements were taken with the tone pot set halfway, and in LTSpice I laid it out such that Q3B connects to C10 and R8 via a pair of 5k ohm resistors.)
I'd say, stick to Dino's schematic for now
