Seven Fuzz Topologies of the 1960s, Pt. 2: The Fuzz Face

Part 2 of Seven Fuzz Topologies of the 1960s, a bench-reference series. New here? Pt. 1 covers the drawing conventions and how to read the clipping diagrams.

2

2-Transistor Shunt-Series Feedback Fuzz

Dallas-Arbiter Fuzz Face topology · 1966

Two direct-coupled gain stages wrapped in a 100 k feedback loop from Q2’s emitter back to Q1’s base. The loop sets the bias and the input impedance, which is why the circuit responds so strongly to the guitar’s volume knob.

Devices
PNP germanium ×2 (AC128 / NKT275), 9 V, positive ground
Supply as drawn
+9 V
Controls
FUZZ · VOLUME
Clipping
transistor saturation / cut-off, strongly asymmetric, gain-dependent
Schematic drawn NPN / positive rail for comparison; see conventions above
How it clips signal at each point in the chain (dashed = what the wave would be without clipping)
1 Input (guitar volume sets the drive)+V0 VQ2 collector biasonly the top rounds off → gritty, "cleans up"2 Guitar volume ≈ 4: soft, asymmetric+V0 VQ2 collector biasQ2 slams the +V rail on one half,starves toward 0 V on the other3 Guitar volume 10: hard, asymmetric

Stage by stage

  • Q1’s collector drives Q2’s base directly, with no coupling cap, so the two transistors share one DC operating point, set by R4 (100 k) feeding Q2’s emitter voltage back to Q1’s base.
  • Q2 has a tiny collector resistor (470 Ω) and a bigger emitter resistor (the 1 k FUZZ pot), so its collector sits around 40 % of the supply. On one half-cycle Q2 is driven into the +V rail; on the other it starves toward 0 V, but the two halves are clipped at different levels and with different shapes. That asymmetry is the Fuzz Face’s even-harmonic warmth.
  • The FUZZ pot works by bypassing the emitter resistor with C2 (20 µF). Wiper down: the emitter resistor is unbypassed, gain is low, the clip is soft. Wiper up: the emitter is bypassed for AC, gain is maximum and the wave squares off.
  • Input impedance is only a few kilohms, so rolling the guitar volume back changes the drive and the loading at the same time, which is the classic clean-up.

Bench notes & values

  • These are the standard Arbiter germanium values. Q1 around hFE 70 and Q2 around 120 is the well-known sweet spot; Q2 collector should measure roughly 4–5 V below the rail with FUZZ at maximum.
  • For silicon (BC108/BC183/2N3904) most builders drop R1 to 47–100 k and R2 to 330 Ω, and add 470 pF–1 nF from Q1 collector to base to stop it squealing.
  • Positive-ground originals will not share a power supply with your negative-ground pedals; a charge-pump inverter or an NPN build avoids that.
  • Bench cost: under $3 in silicon; a matched pair of tested germanium PNPs is $15–40.
Same topologyArbiter Fuzz FaceDallas Arbiter Fuzz FaceColorsound Fuzz BoxEHX Muff Fuzz / Little Muff PiGoya FuryGoya PantherJennings FuzzMarshall Supa FuzzPark Fuzz SoundRangemaster FuzzbugRoger Mayer Axis FuzzRotosound Fuzz BoxSola Sound Tone Bender MK1.5Vox V816 Distortion BoosterVox V8161 / V8162

Sources for values: Gibson patent US 3,213,181 (FZ-1); Arbiter Fuzz Face germanium schematic; Sola Sound Tone Bender Professional MkII (pedalparts.co.uk BOM, vero-p2p); Aion FX Orpheus (Fuzzrite) and Astra (Astrotone) documentation; fuzzcentral.ssguitar.com Octavia parts lists. Octavia and Boss Tone stage values are representative working designs, not traces.

Drawn with schemdraw and matplotlib. Made for Jason’s bench at Illicit Apothecary.

Previous
Previous

Seven Fuzz Topologies of the 1960s, Pt. 3: The Tone Bender MkII

Next
Next

Seven Fuzz Topologies of the 1960s, Pt. 1: The Maestro Fuzz-Tone FZ-1 (and How to Read These Sheets)