No Bias
The No Bias fuzz topology is a bipolar transistor stage that has no forward bias applied to the base terminal. Oddly, the transistor will be forward biased in most designs. I almost wanted to call this a variant of "self bias" except I didn't want to confuse this with the operation of tube self bias circuits. The transistor will be a germanium type and the circuit takes advantage of the otherwise undesirable "collector leak" feature of germanium transistors. The above schematic is shown in old fashioned "positive ground" orientation to match typical fuzz schematics. A modern drawing would be upside down.
Basic transistor theory tells us that this transistor should be cutoff. The dc voltage at both the base and emitter should be 0V. When the base-emitter voltage (with respect to each other, not ground) is less than the forward bias voltage (usually 100mV to 300mV for germanium), the transistor should be cutoff. Strangely, when using leaking germanium transistors, a small voltage may develop at the base terminal. This is due to internal leakage and has a positive feedback effect with heat. I was only able to find a textbook explanation of this in a 1950s transistor manual. This effect still exists in bipolar junction transistors, but is incredible small and unnoticeable with modern devices. The leakage is represented in the textbook as a resistor connecting collector and base, with a current source labelled "thermal" connected in parallel to the leakage resistor.
Diagram from "Handbook of Semiconductor Electronics" Edited by Hunter, McGraw-Hill 1956
The voltage developed due to collector-base leakage and temperature can either allow the transistor to "self-bias" so to speak, or be just on the brink of turning on. If just at the brink, the guitar signal voltage will be enough to forward bias the base-emitter junction. The resultant tone may be noticeably "gated" or "choppy" if the transistor is mostly cutoff.
The internal leakage is not usually known unless you test a batch of transistors or stick them in a circuit. Even then, the stability is low and temperature will have a very noticeable effect on the circuit operation. For these reasons, No Bias fuzz circuits tend to vary from unit to unit and situation to situation.
The Fuzz Tone FZ-1 is the earliest example of this configuration, and it is found in the 3rd transistor stage. The relatively high value collector resistor of 10kΩ and lack of emitter resistor makes for a very distorted output. The FZ-1 patent claims the output of the final stage should have 100% THD, with a minimum of 25% THD. It also notes that, due to collector leakage, the collector voltage should be about -2.5V with a -3.1V supply voltage.
The original Sola Sound Tone Bender (retroactively labelled the "Mk I") was a modified FZ-1, retaining the basic topology of all three stages, but making significant changes in power supply voltage (-9V) and component values to increase the sustain and bottom end.
The Sola Sound Tone Bender Mk II uses the No Bias fuzz in its first stage, effectively moving the 3rd stage of the FZ/Mk I up to the front. This drives a circuit I call the 2 Transistor English Fuzz, which appears in the so-called Tone Bender Mk I.V or 1.5, sort of the "missing link" version.
For non-fuzz effect purposes, this is generally called a "class C" amplifier, and it has practical purposes. Class C circuits often switch things on and off. A simple example would be a "signal present" LED.
The second stage of the Maestro Fuzz Tone allows for variable bias adjustment as a front panel control labelled "Attack." This is really where you can set the bias to "no bias." The above diagram is a simplified version of that control. The actual FZ-1 control has a stop resistor to set the a limit on the lowest setting, so that complete cut off cannot be achieved if a leaky germanium transistor is used. When the bias is set near cut off, the envelope of the guitar signal has a fast decay due to the cutoff or near cutoff condition of the transistor. The transistor is cut off when the variable resistance is small, and forward active when the resistance is made large.
Unlike the other stages of the FZ-1, this stage has a resistor from V- to base. This resistor provides a concrete forward bias path independent of collector leakage, so this topology can be used for low leakage silicon devices. From the text of the FZ-1 patent, it is possible that the designer did not include this resistor initially (matching the design of the other stages) but decided it was necessary to provide reliable operation. Without enough collector leak current, the "attack" control could mute the output when set to the minimum resistance, which would be confusing for the end user. Because cold temperatures lower collector leak currents, playing in a cold environment could also cause complete cut off. The additional resistor, and the (not shown) stop resistor provide a reliable enough source of bias for the old germanium devices, regardless of the temperature (assuming the room is within normal human comfort limitations, I suppose).
The No Bias Emitter Follower
Oldest know version is the Maestro Fuzztone FZ-1. This configuration does not voltage amplify. This stage requires subsequent amplification (2 additional stages, both common emitter voltage amplifiers, follow in the FZ-1). The 1965 revision FZ-1A also contains the No Bias circuit. The Tone Bender Mk I, being a modified FZ-1, also contains the common collector version.
This configuration actually makes for a surprisingly clean emitter follower using a germanium transistor, contrary to what you'd expect to get from a silicon device. Even at the FZ-1 and FZ-1A low V- values (-3V and -1.5V) the swing at the emitter is usually enough for a clean transfer from input to output.
So this stage does not actually fuzz (with germanium devices), but is included here for its role in the Fuzz Tones and Tone Benders.
Late 60's "Diode Variation"
At some point a common emitter variation with a diode appeared and was subsequently used in several designs. The diode replaces the resistor in the first diagram. The diode is oriented as to have the opposite polarity of the base-emitter diode of the transistor.
Silicon Versions?
Maestro switched to silicon in the late 1960s, and may have tried to emulate the "no bias" sound with the FZ-1B (V2) and its distinct gated fuzz topology.
Circa 1966, the Sunn amplifier company came out with a 1.5V powered silicon device called the Sunn Buzz SB-1G, and this appears to be setup in a way to simulate "no bias" fuzz with silicon devices, but I have no way of confirming if that was the designer's intention or not.
Examples of the No Bias:
Hornby Skewes Zonk Machine
Amplifier Corp. of America Fuzz King
LRE Fuzz Sound
Examples of the No Bias "Diode Variation":
Baldwin-Burns Buzzaround
UMI Buzz Tone and Volume Expander
Apollo / Conrad / Kimberly/ Etc Buzz Box and Volume Expander
Example of a silicon imitation: