Chuck D. Bones
Circuit Wizard
The first time I breadboarded this, I was not impressed. Then I listened to Andy Martin demo it.
Nothing like Andy playing Black Sabbath to make a fuzz pedal sound good.
So I thought I'd give it another go. The Repro is a Fuzz Face with a Sziklai pair doing the job of the 2nd transistor. After that is a Fender tone stack and a buffer. The bottom-end is very heavy and that's because the low impedance of R1 and C1 allow the guitar to beat Q1 to death. I used a Blue Top MP38A for Q1. I noticed that the circuit is very temperature sensitive, no big surprise there because Q1's HFE, Vbe and leakage all strongly influence the bias point for Q1-Q3. I connected a DMM to Q3-E, put my finger on Q1 and watched the bias run away. First thing I did was swap Q1 & Q2, installed a 1.5K resistor between E & B on Q3 to absorb some of Q2's leakage and increased R1 & R4 about 7x. It was still pretty bottom-heavy, so I reduced C1 to 15nF. There is still plenty of bottom-end. C1 is a good place to install a FAT switch. Now it's sounding more to my liking. I also reduced R8 to 3.3K because I wanted to bias Q2 & Q3 a little hotter. I used a C100K pot with a 4.7K in series for the Bias trim. At the low-resistance end (full CW), the sound is very gated and sputtery. Down the dial a bit, there is a strong 2nd harmonic presence. Below that there is a sweet spot where the mid presence increases and clipping is very symmetric. This thing has plenty of gain, around 75dB at full tilt. BASS & TREBLE were very touchy at the bottom end of rotation, so I changed them to A250K. With a silicon transistor for Q1, the bias point is very stable. We still get plenty of germanium tone (whatever that is) from Q2. Q3 can be any silicon PNP. I tried 2N5087 and 2N3906, they sound the same. I ended up increasing C4 to 220pF to move the mid notch down a little; one could go even larger. I bumped R10 up to 6.8K because that's what Fender uses in the Deluxe Reverb tone stack. That 500K VOLUME control hurts my brain. What's the point of putting a buffer after the tone stack if you're going to make the output impedance so high with a 500K volume control? I'm running A100K presently, but anything from A10K on up would work. I increased C8 to 1uF. I wanted less loading on the tone stack, so I made Q4 a MOSFET (BS170) and increased R11 to 3.3M and R12 to 1M. A 2N7000, JFET or Darlington would also work, just adjust R11 to get 4.5V to 5V on the top of R13.
L to R: Volume, Bass, Treble, Bias, Gain. Output jack is between Treble & Bass, power jack is hiding under the Gain control. Lots of spare transistors scattered around at the bottom of the pic.
So I thought I'd give it another go. The Repro is a Fuzz Face with a Sziklai pair doing the job of the 2nd transistor. After that is a Fender tone stack and a buffer. The bottom-end is very heavy and that's because the low impedance of R1 and C1 allow the guitar to beat Q1 to death. I used a Blue Top MP38A for Q1. I noticed that the circuit is very temperature sensitive, no big surprise there because Q1's HFE, Vbe and leakage all strongly influence the bias point for Q1-Q3. I connected a DMM to Q3-E, put my finger on Q1 and watched the bias run away. First thing I did was swap Q1 & Q2, installed a 1.5K resistor between E & B on Q3 to absorb some of Q2's leakage and increased R1 & R4 about 7x. It was still pretty bottom-heavy, so I reduced C1 to 15nF. There is still plenty of bottom-end. C1 is a good place to install a FAT switch. Now it's sounding more to my liking. I also reduced R8 to 3.3K because I wanted to bias Q2 & Q3 a little hotter. I used a C100K pot with a 4.7K in series for the Bias trim. At the low-resistance end (full CW), the sound is very gated and sputtery. Down the dial a bit, there is a strong 2nd harmonic presence. Below that there is a sweet spot where the mid presence increases and clipping is very symmetric. This thing has plenty of gain, around 75dB at full tilt. BASS & TREBLE were very touchy at the bottom end of rotation, so I changed them to A250K. With a silicon transistor for Q1, the bias point is very stable. We still get plenty of germanium tone (whatever that is) from Q2. Q3 can be any silicon PNP. I tried 2N5087 and 2N3906, they sound the same. I ended up increasing C4 to 220pF to move the mid notch down a little; one could go even larger. I bumped R10 up to 6.8K because that's what Fender uses in the Deluxe Reverb tone stack. That 500K VOLUME control hurts my brain. What's the point of putting a buffer after the tone stack if you're going to make the output impedance so high with a 500K volume control? I'm running A100K presently, but anything from A10K on up would work. I increased C8 to 1uF. I wanted less loading on the tone stack, so I made Q4 a MOSFET (BS170) and increased R11 to 3.3M and R12 to 1M. A 2N7000, JFET or Darlington would also work, just adjust R11 to get 4.5V to 5V on the top of R13.
L to R: Volume, Bass, Treble, Bias, Gain. Output jack is between Treble & Bass, power jack is hiding under the Gain control. Lots of spare transistors scattered around at the bottom of the pic.
Last edited: