Seven Fuzz Topologies of the 1960s, Pt. 4: The Octavia

Part 4 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.

4

Differential Pair / Frequency Doubler

Roger Mayer Octavia topology · 1967

A phase splitter makes two copies of the signal 180° apart; a diode on each copy passes only its positive half. Summed, the two halves reconstruct the absolute value of the signal, a full-wave rectifier, which contains a strong component at twice the input frequency: the octave up.

Devices
Silicon NPN/PNP mix (MPSA18, 2N6519 / 2N4401), 9 V
Supply as drawn
+9 V
Controls
BOOST (gain) · VOLUME
Clipping
full-wave rectification (frequency doubling) plus stage overdrive
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 (after Q1 buffer / Q2 gain)2 Q3 phase splitter: two copies, 180° apartcollectoremitternegative half is folded up(full-wave rectification)3 D1 + D2: each passes one halftwo peaks per input cycle= one octave up4 Output (AC-coupled): 2× frequency

Stage by stage

  • Q1 is an emitter-follower buffer so the guitar is not loaded down. Q2 is a common-emitter gain stage; its emitter bypass C3 gives it enough gain that the rectifier is driven by a large, already fuzzy signal.
  • Q3 is the heart: a phase splitter with equal collector and emitter resistors (R9 = R10 = 2.2 k). The collector gives an inverted copy, the emitter a non-inverted copy, same amplitude.
  • D1 and D2 (germanium, low forward drop) each conduct on their own positive half-cycle into the shared load R13. Positive half from one, folded-up negative half from the other: the fundamental cancels and what is left has two peaks per input cycle. Play a note at the 12th fret and you hear the octave clearly; play chords and it turns into intermodulation mush, and that’s expected.
  • Q4 buffers the rectified signal to the volume pot. Lifting one diode (the "octave off" switch on many builds) leaves a plain half-wave fuzz.

Bench notes & values

  • Roger Mayer’s original and the Tycobrahe reissue use a small centre-tapped transformer (1.5 k CT : 600 Ω) as the phase splitter, with the diodes on the secondary. The transistor phase splitter drawn here is the "transformerless" equivalent and behaves the same; C5/C6 and R11/R12 are needed because the collector and emitter sit at different DC voltages.
  • Values are a working starting point for a transformerless build (Q3 collector ~5.6 V, emitter ~3.4 V), not a trace of a specific unit. Matched diodes (forward drops within ~20 mV) give a cleaner octave.
  • The Univox Super Fuzz and Fender Blender get the same doubling from a true differential pair (two transistors with a shared emitter resistor, bases driven in antiphase): same family, different implementation.
  • Bench cost: about $4 in parts; a 42TM022 transformer adds ~$4 if you want the original coupling.
Same topologyRoger Mayer OctaviaTycobrahe OctaviaOctavia Evo 1Ace Tone Fuzz Master FM-2Fender BlenderUnicord Super FuzzUnivox Super Fuzz

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. 5: The Mosrite Fuzzrite

Next
Next

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