HOMO ELECTROMAGNETICUS
Cloning and Modifying a Mosrite Fuzzrite
Part I: Cloning
Mosrite created and sold the Fuzzrite in the 1960s. It was an effect
for the guitar that made the instrument sound like an angry hornet.
It's not a sound that everyone likes, but it holds a special place
in my heart for its use in some Spaghetti Western film scores.
If one looks up this circuit online, one will quickly find a schematic
that looks like Figure 1. Note that Q1 and Q2 are given as the common
2N2222 BJT simply because I could not find conclusive information
about the silicon BJTs used in the original. Basically, this circuit
works by cascading two simple BJT gain stages together and crudely
mixing their respective outputs with a potentiometer. Stick another
potentiometer on as the output volume control and it's done.
Most schematics for this fuzz omit the 22k resistor shown in Figure 2.
This is unfortunate, as it is a key to this effect's unique sound! In
the original effect, the 22k resistor was found not on the PCB, but
rather behind it and soldered directly to the one of the "fuzz"
potentiometer's leads. Given that, it's easy to see why this resistor
is often forgotten. To see how this resistor effects the circuit, I
ran a SPICE simulation while changing the value of the 22k resistor.
One of the simulations ran includes the case where R10 = 10MOhm, which
is roughly equivalent to removing it from the circuit entirely.
It can be seen that omitting R10 would give the effect a totally
different sound than the original design! Without R10, the fuzz
effect would have a lot more low frequency content and sound similar
to other popular fuzz effects of the same era. The mechanism is simple
- R10 and C3 form a simple high-pass filter together that is "before"
the "fuzz" potentiometer.
If one intends to get the famously deadly buzz-saw tone this pedal is
known for, it's best not to forget the 22k resistor!
Part II: Modifying
I play a baritone guitar tuned a fourth lower than a standard guitar,
so the frequency response of this effect is not quite right for my own
use in its stock form. Simply, the lower register notes on the
baritone guitar are very quiet when using the effect. We can't have
that!
The above simulation of the effect of R10 in the Fuzzrite circuit
showed a frequency peak around 120Hz. Specifically, this peak is
centered at the B that's two frets above the A string on a normal
guitar. To modify this circuit to be more appropriate for a
baritone guitar, the obvious choice would be to move this peak down to
the F# that's two frets above the E string (~92Hz).
The location of the aforementioned peak in the low-frequency response
of the Fuzzrite is largely controlled by C1 and C4. They are the input
coupling capacitors for the two BJT gain stages. Below are the results
of a simulation where the values of these capacitors are changed in
tandem.
Not only does increasing C1 and C4 in the circuit move the
low-frequency peak closer to 0Hz, it also decreases the magnitude of that
peak. Choosing C1 and C4 to be 68nF to place the peak at ~92Hz, the peak
will be a few dB down from the original. Recalling Figure 3, one can
increase R10 to get back some low-frequency amplitude. A small adjustment
in the location of the peak required only a small adjustment in R10 to
recover the lost amplitude. 33k was used in the modified circuit instead
of the original 22k. This way, the circuit will sound as close as
possible to the original, while at the same time be more suited for a
baritone guitar.
In addition to adjusting the low-frequency peak of the effect, I thought
it may also be prudent to add a simple tone control to tame the high
frequency response. I personally like the buzzy nature of the pedal, but
it's not hard to imagine the high frequency content being a bit too much
in some situations. To create this tone control, I simply copied the
designs people use in tone knobs on guitars and stuck a 100k pot with a
capacitor to ground on the circuit's output. A simple technique like this
is not without side effects, but fuzz pedals are not expected to be
well-behaved. As is standard, a logarithmic potentiometer is used for
tone.
Finally, the circuit needed some other small modifications to accommodate
modern parts. The "fuzz" potentiometer in the original circuit was valued
at 350k, but one can't get such potentiometers anymore. With a resistor
across lugs 1 and 3 of a more common potentiometer, one can approximate
more esoteric values. In this case, I used a 1M resistor across a 500k
potentiometer to approximate the original 350k. The volume potentiometer
in the original circuit was valued at 35k, but for this one I grew lazy
and simply used a 50k model. Coupling capacitors C2 and C3 were
originally specified as 2nF, but 2.2nF versions are close enough and
much easier to come by.
The modified circuit as-built is presented below.
As a parting note, I'd like to add a statement about transistor
selection in circuits that use the "self-biasing" technique employed in
the Mosrite Fuzzrite. Good biasing technique makes a given transistor
circuit's behavior depend on anything but the Beta or hfe of a
transistor. This is typically done because of the wide spread of Beta
/ hfe values that exist among transistors of the same part number. It's
not hard to imagine why designing such variance out of a circuit is
desirable. The "self-biasing" or "DC-feedback" biasing technique used in
this circuit is a good example of such a biasing technique and makes the
circuit fairly immune to Beta / hfe variation.
What does this mean for the budding circuit guru? It means that
swapping out different transistors in this circuit, hoping to achieve a
different sound, is a fruitless activity. For a wide range of transistor
Betas / hfe values, each stage has about 44dB of gain and each collector
will sit at a voltage of about 0.6V, give or take 50mV.
Some people will of course claim that they can hear a difference between
two transistors with an hfe of 150 and 200 in this circuit, but I'll
leave any "golden ears" discussion to the audiophile crowd.