The Maestro Fuzz Tone is the original commercial fuzz. The line up of the Fuzz Tones includes the FZ-1, the FZ-1A, the FZ-1B (Version 1), the FZ-1B (Version 2), and the FZ-1S.
The -1 and -1A are both germanium units, while the later models used silicon. The -1 and -1A have a distinct fuzz effect due to an unusual circuit topology that this device introduced and was heavily copied throughout the 1960s. I call this topology "No Bias" due to the lack of voltage across the base-emitter junction.
"No Bias" stages work best with germanium transistors due to their lower forward bias voltages and large collector leak currents. When the later Fuzz Tones switched to silicon, three completely different fuzz topologies were substituted in a search for the next hit sound. While the FZ-1 was a leader in the early fuzz craze, the -1Bs and the -1S appear to include imitations of other designs, and thus the later pedals do not have near the prestige of the original Fuzz Tone. They are all interesting though, and each is examined below.
Values and voltages taken from the text of the patent. The 470R* resistor may be a mistake, with 470k being the likely correct value. Transistors are germanium with significant collector leak currents.
The original schematic appears to have omitted a resistor (top). The middle one shows the alleged correction. Actually it is supposed to be a 470k resistor, but as there were no 4s or 7s to copy, I just grabbed the 500k from the volume control when making this mockup. FWIW, I'm sure if you used a 500k it would work about the same as a 470k. The circuit usually functions with leaving it out as well. A traced schematic and photos of an actual unit are shown here: http://guitar-mod.com/rg_pedal_maestro_repair.html
I would be curious if it was truly an error, or if really early models indeed omitted the resistor. I will be carefully inspecting every FZ-1 I see from now on!
The bottom image is taken from designer Glen Snoddy's 1965 U.S. patent. This is very similar, but the patent actually gives dc voltage levels too! Q1's base voltage is given as -1.5V, Q2's collector voltage is given as -3V, and Q3's collector voltage is given as -2.5V. The patent also includes waveforms for a 100mV (no indication of rms or p-p) input signal with the ATTACK at max and min.
Note the "missing" resistor has become 470Ω. Whether this is a mistake or not may never be cleared up. It certainly looks incorrect, as such a low resistance from the negative supply to the base would tend to saturate the transistor, bringing the collector voltage closer to 0V. As the dc voltage at the collector is given to be -3V (near cutoff), 470k is probably the intended value. In my own experiments, both 470k and open circuit work (requires a germanium transistor with enough collector leak current), while 470Ω does not, so I believe it is an error in the patent. The text of the patent notes this resistor helps guard against changes in temperature. This is in reference to the fact that we are relying on the collector leak current in all 3 of these germanium transistor stages, which functions as an internal resistor from collector to base. If the temperature drops, the leak current diminishes and the internal resistor appears larger. The 470k resistor is an external biasing resistor much like you see on any conventional transistor amplifier, providing V- to the base for forward biasing.
The FZ-1 introduces "No Bias" fuzz. To modern eyes, this circuit looks outright wrong, as though it cannot even function. Assuming there is a collector leak of 0A (like most modern devices), the first and last stages should have base-emitter voltages of 0V, and thus be cutoff. A strong input signal will forward bias the relatively low germanium diode junctions (approx. 100mV-300mV) and some signal will squeak through, get amplified by the second stage, and then splatter again through the 3rd stage. If you substituted silicon transistors, you'd get these results, but that's actually not how the FZ-1 works.
The interesting thing about germanium transistors is that the collector leak will be >0A, and the base-emitter voltages will shockingly be >0V. The patent gives us a voltage of -1.5V at the base of Q1, which simply makes no sense unless you consider collector leak current. Equally baffling is the given voltage of -2.5V at Q3's collector. This "leak" is due to inherent flaws in germanium transistors; flaws that led to silicon quickly rendering germanium obsolete for most electronics. Because of collector leak current, the first and last stages may be forward active at idle. The degree of bias will be different depending on actual device and ambient temperature. I have a photo from a 1957 electronics manual that illustrates the internal mechanisms of a typical germanium transistor on the page devoted to those devices (see link at beginning of paragraph).
Another interesting note from the patent is that the second and third stages are said to add a minimum of 25% THD to the input signal, with 100% THD being preferred. The patent states that the purpose of this device is to "produce significant tone modification as distinguished from what might be classed as mere distortion." In other words, this is, in Snoddy's mind, a primitive version of what we would end up calling synthesizers. The device is supposed to simulate "trumpets, trombones, and tubas." I also like this note regarding playing multiple strings: "Playing two or more strings at the same time produces unpredictable and unpleasant noise. If two or more tones are desired simultaneously, a complete separate system is required for each tone." Such a "separate system" approach would catch on in the 1970s with hex pickups to drive guitar-to-CV synthesizer systems. I wonder if Snoddy experimented with such a thing for his Fuzz Tone.
The output capacitor is notably small so as to introduce a noticeable bass cut and sharpness to the fuzz. Increasing the size of the cap creates a more square fuzz output, with increased mids and bass at the expense of the sharp signature tone of the FZ-1. The fist Sola Sound Tone Bender (modified FZ-1) increased this to 100nF (0.1µF).
The FZ-1 was a famous flop. While it seems the top recording guitarists all had to have one, the general public snubbed them. Sales tapered off quickly after the initial release. The Beatles were photographed with them in 1963, but the recordings they made were not released. Another British invasion group, the Rolling Stones, would finally make the pedal a hit when it was featured on "Satisfaction" in June 1965. The pedal had already been copied, modified, and released as the Sola Sound Tone Bender, but now the copying and evolving was going to explode. (The Tone Bender was also featured in a June 1965 hit single: "Heart Full of Soul" by the Yardbirds.)
Note that the picture is NOT a FZ-1A, but a FZ-1S with toggle switch and 3 knobs.
Above is an FZ-1A schematic. The FZ-1A was the 1965 redesign of the unit. The redesign appears to have occurred after the Rolling Stones' "Satisfaction" created a new demand for Fuzz Tones.
The basic idea is the same as the FZ-1, but instead of the FZ-1's two AA cells, the FZ-1A is powered by a single cell AA cell, reducing the total supply from 3V to 1.5V.
It's unclear why Maestro needed the redesign, and outright baffling why they would lower the supply voltage while others (notable the Tone Bender) were increasing the power to 9V. A possible guess is that it still functioned at 1.5V, so why force the customer to include the extra battery cell?
The 2nd stage coupling capacitors have been reduced to a much more normal 1µF from the monstrous 20µF specified in the older schematic.
Q2's collector resistor has increased from 1.5k to 10k, possibly increasing the effective output from that stage, although the extremely low supply voltage limits the overall gain, so the net effect may be in shaping of the resulting clipped signal.
The volume control has consolidated the large value control with shunt resistor into a single medium value control.
Note that the picture in the schematic above is of a FZ-1S, not a FZ-1A. The FZ-1A enclosures look pretty identical to the FZ-1 enclosures.
The FZ-1B (Version 1) changed the Fuzz Tone dramatically. Gone are the No Bias stages. A fairly tame shunt feedback voltage amplifier (Q2) gets passively clipped by D1. An effect cancel style bypass via SW1 merely mutes the fuzz tone when not desired. Q1's biasing is very unusual. It is a common collector amplifier that is biases only R2's direct connection to V+. The drawing may fool you into thinking R2 supplies shunt feedback - it does not. The +9V is low impedance and does not supply the base with negative feedback.
This version does not have a strong fuzz tone like the previous and following designs. If looking for that signature "Fuzz Tone sound," do not look to the FZ-1B. The sound is distinctly more overdrive, but the forced mix of clean (and 180 deg out-of-phase) signal via R4/C3 really confuses things.
Simply snipping the clean path out is advisable, or reconnecting/replacing the BALANCE control with a BLEND control (clean at CCW terminal, fuzz at CW terminal). See how the BALANCE control is implemented in the FZ-1S for yet another option.
It's unclear what year the FZ-1B replaced the FZ-1A. The 1966 Astro Amp Astrotone includes a passive diode clipper and the idea seems to catch on after that among several designs. The FZ-1B(V1) may have been influenced by this trend.
For reasons undeclared by Maestro, the FZ-1B received a dramatic makeover while retaining the exact same model number. The two can be differentiated by "Version 1" and "Version 2."
The schematic above has some nice labels to guide us around. The input buffer (Q1, "Impedance Matcher") again uses a questionable biasing scheme of just sticking a high impedance resistor to base, like Version 1. The fuzz preamp (Q2) has a high gain shunt feedback amplifier. The 100k collector resistor is a bit too high and surely causes some extra fuzz due to early saturation. Our "BALANCE" control mixes this first stage with the output of the second stage. Say, that's a 2 Stage Fuzz Blend topology! The "Fuzz Amp" (Q3) is analyzed on the Gated Fuzz topology page.
The bypass scheme returned to the more conventional switch-the-output jack.
Version 1 is notably more overdrive (a not a very good overdrive) than fuzz. It also marked the Fuzz Tone's first silicon design, which doesn't work as well with No Bias stages. My guess is that Version 1's lack of eerie "fuzz" that is so key to the -1 and -1A sounds quickly gave Maestro negative feedback from dealers and end users. The Q4 "squelcher"/gate circuit may be an attempt to impart the No Bias effect into silicon devices.
The 2 Stage Fuzz Blend was quite popular from the Mosrite Fuzzrite and its many imitations. The Fuzzrite may have been an influence on this design.
The "Squelcher" transistor, while a very well know gain control idea, appears to be novel among vintage fuzz effects. The FZ-1B(V2) is notable for introducing this new topology of Gated Fuzz.
The original schematic contains some obvious errors I have corrected in the above schematic.
(Original errors: R7 connected to +9, R1B connected to R14/R9/R5 node, R1 not labelled A/B, several parts all labelled the same)
The FZ-1S changes things completely yet again! Q1 recalls the -1B input buffers, but with stable biasing via R4/R5. C3 and C2 offer split paths for the signal, and the twin R7/R23 33k resistors recombine things before output amplifier Q2.
Q3/Q4 is are two cascaded shunt feedback amplifiers. Q5/Q6 are diode connected transistors that provide passive shunt clipping. R22/C8/R11 is a passive filter network. Fuzz Sw. toggles between the straight or filtered tones. The fuzz path is notably similar to the Univox Superfuzz and Fender Blender.
R1 is a dual gang pot that turns up fuzz and turns down clean as it rotates clockwise. Sort of proto-Klon in that respesct. R2 provides a master output control.
The FZ-1S shows an interesting evolution from the earlier fuzz topologies. It seems to follow the general trend of 1960s fuzz evolving into 1970s distortion/overdrive.
U.S. Patent 3,213,181, G.T. Snoddy Et Al "Tone Modifier For Electrically Amplified Electro-Mechanically Produced Musical Tones" Filed May 3, 1962, Granted Oct 19, 1965