Seven Fuzz Topologies of the 1960s, Pt. 1: The Maestro Fuzz-Tone FZ-1 (and How to Read These Sheets)
This is Pt. 1 of Seven Fuzz Topologies of the 1960s: clean schematics with representative values, and a signal-by-signal picture of how each one clips. Redrawn from the topology sheets in How 1960s Engineers Invented Fuzz. A new topology posts every few days.
How to read these
Every circuit is drawn the way the video draws them: NPN symbols, positive supply at the top, ground at the bottom. That makes the family resemblances easy to see side by side. Five of the seven originals were built with PNP germanium transistors on a positive-ground supply. To build one of those as its ancestors did, use PNP parts, make the rail −9 V (or −3 V) with the positive battery terminal grounded, and flip every electrolytic and diode. To build it as drawn, use NPN silicon (2N3904, BC549, 2N2222) and expect a slightly harder, brighter clip than germanium gives.
Values are starting points, not gospel. Where they come from a documented trace or the original patent the bench notes say so; where they are representative they say that too. Germanium circuits in particular are biased by the transistor’s leakage as much as by the resistors, so socket the transistors and measure collector voltages before judging the sound.
Reference designators are per circuit (R1 in one sheet has nothing to do with R1 in another). The 8th chapter of the video was not in the screenshots, so it is not here.
Clipping diagrams
Blue: the input, or the signal as it would be without clipping (dashed).
Green: the signal after the stage named in the panel title.
Orange dashed marks the level that does the clipping: a supply rail, transistor saturation, or a diode threshold.
Each row reads left to right along the signal path. The dotted grey line, where shown, is the DC bias point the stage sits at with no signal.
The series at a glance
| Topology | Original devices | What does the clipping | |
|---|---|---|---|
| 1 | 3-Transistor Saturation Cascade Maestro Fuzz-Tone FZ-1 topology | PNP germanium ×3 (2N270), 3 V, positive ground | transistor saturation / cut-off, asymmetric, then a differentiator |
| 2 | 2-Transistor Shunt-Series Feedback Fuzz Dallas-Arbiter Fuzz Face topology | PNP germanium ×2 (AC128 / NKT275), 9 V, positive ground | transistor saturation / cut-off, strongly asymmetric, gain-dependent |
| 3 | 3-Transistor Modified Darlington Sola Sound Tone Bender MkII topology | PNP germanium ×3 (OC75 / OC81D), 9 V, positive ground | transistor saturation, two clipping stages in series |
| 4 | Differential Pair / Frequency Doubler Roger Mayer Octavia topology | Silicon NPN/PNP mix (MPSA18, 2N6519 / 2N4401), 9 V | full-wave rectification (frequency doubling) plus stage overdrive |
| 5 | Parallel Phase-Blend (Subtractive Mix) Mosrite Fuzzrite topology | PNP germanium ×2 (2N2613 / 2N408), 9 V, positive ground; later silicon | transistor saturation in each stage, then antiphase cancellation |
| 6 | Cascaded Boost → Clip → Buffer Sam Ash Fuzz Box / Astrotone topology | Silicon NPN ×2 (2N3565), 9 V, negative ground | symmetric diode clipping at ±0.6 V |
| 7 | Capacitive-Coupled Dual-NPN Cascade Jordan Boss Tone "California" topology | Silicon NPN ×2, 9 V, negative ground | transistor saturation ×2, then symmetric diode clipping |
3-Transistor Saturation Cascade
Maestro Fuzz-Tone FZ-1 topology · 1962
Three germanium transistors, no diodes. Q1 buffers, Q2 and Q3 are driven so hard they slam into the supply rail on one half-cycle and into saturation on the other. The ATTACK control shifts Q2’s bias so the clipping is lopsided.
- Devices
- PNP germanium ×3 (2N270), 3 V, positive ground
- Supply as drawn
- +3 V (two AA cells)
- Controls
- ATTACK · VOLUME
- Clipping
- transistor saturation / cut-off, asymmetric, then a differentiator
Stage by stage
- Q1 is an emitter follower: it does no clipping, it just turns the guitar’s high impedance into something Q2 can be driven from. R3 (470 k) is temperature compensation for the germanium part.
- Q2 has almost no collector resistor (1.5 k) and its base is pulled toward the rail through the ATTACK rheostat. With ATTACK low the base current is huge and Q2 sits near saturation; the output wave is a rail-to-rail pulse whose duty cycle depends on the knob. That bias shift is what makes the top and bottom of the wave clip at different levels, giving the "positive half clipped / negative half clipped" picture from the video.
- Q3 repeats the trick with a 10 k load: whatever survived Q2 is squared off again.
- C4/R10 (3.3 nF into 56 k) is a differentiator. It throws away most of the low end of the square wave and keeps the edges, which is why an FZ-1 sounds thin and buzzy rather than fat: the "Satisfaction" sound.
Bench notes & values
- Values are from Gibson’s 1962 patent (US 3,213,181), which is the FZ-1. Production FZ-1A units dropped to 1.5 V and used 1 µF coupling caps instead of 20 µF; the sound leans even more gated.
- Originals are PNP with positive ground. Build it as drawn with three NPN silicon parts (2N3904, hFE 60–120 is plenty) or flip the supply, electrolytics and transistor polarity for PNP germanium.
- The ATTACK pot is drawn as a rheostat (wiper tied to one end), which is how the patent uses it. Socket Q2; leaky, lowish-gain germanium (hFE 80–130) is what makes the knob interesting.
- Bench cost: about $3 in silicon, $15–30 if you chase real 2N270s.