Foxx Tone Machine

[Christoper]

Good evening,

I recently acquired a Warm Audio clone of the Foxx Tone Machine and had a few questions about the circuit



Firstly, what is the tone control doing and why? I mean, I understand that it's a HPF and LPF filter that intersect to make a seesaw filter, but why did they implement it the way they did?

Second, what's the fun little RC network between Q2 and Q1 for?

Third, are there any good mods for the octave that aren't just replacing the switch with either a foot switch or a pot?

Fourth, are there any notable known differences between the Warm Audio Foxy Tone Box and an original Foxx Tone Machine?

[FleshOnGear]

1) I don't think there's a way to know why the designers chose to build the tone control this way. Because the 0.003 cap doesn't have a resistor to ground following it, the tone control won't see saw nearly as much as a BM tone control. This control mostly affects just the treble, and scoops a lot of mids.

2) It appears to be a negative feedback loop around Q1 that removes mids. So it should focus the gain of Q1 in the mids.

3) What octave mod would you have in mind? The Tone Machine is one of the most effective upper octave fuzzes I've ever heard. I'm not aware of any famous mods for it, but maybe someone else does.

4) I don't know about this. My guess is that Warm Audio probably tried to make it pretty faithful to the original. Perhaps others know more.

Enjoy the pedal! The Tone Machine circuit is a lot of fun.

[antonis]

1. You can transfer 4k7 resistor (and short its open pads) between 3nF / Tone pot upper lug and GND..
(does now look more familiar to you..??)
What is "special" here is the part of Tone pot contributing in HPF corner frequency variation (togheter with Q4 input impedance..)



3. Current octave configuration is the simplest it can be..

By opening SW1, you turn Q2 into a unity gain CE amp ala Vulcan style..

[Mark Hammer]

The octaving is produced by the two diodes "chopping off" half the wave coming off the inverting and non-inverting outputs of Q2.  However, because they are diodes, they ALSO chop off a bit of the "other" half of the wave.  So, for example, if the signal hitting the input to each of those diodes is 1V, peak-to-peak, and the diodes are standard silicon (1N914/1N4148 type), they will "subtract" about 600mv from the result.

That, in itself, is no big deal, because there are other gain stages later on in the signal path.  BUT, cancelling the octave effect does not result in either a clean sound or a Fuzz Face kind of distortion, because that one remaining diode still lops off a little more than half the wave.  The effect of that is not quite so noticeable if one is using diodes with a lower forward voltage, like germanium or Shottky.  But if one uses silicon diodes, the difference between forcing the entire signal to pass through that diode...and not, IS audible.

So, what I like to do is use a 3-position on-off-on toggle (SPDT is fine, or DPDT if you feel the need for mode-indicator LEDs), to yield octaving, non-octave with "wave-chopping" or fuzz with NO wave-chopping.

The common of the switch goes to the junction of the 100k/10uf pair.  In one side position, it connects to the second diode.  In the middle position, it lifts the connection to the "octaving" diode, to yield the standard Tone Machine fuzz sound.  But in the "other" side position, it bridges the default diode and connects directly to the 10uf cap coming off the collector of Q2.  Technically, that's not a mod of the octave, but a mod of the non-octave tone.  As noted, the audibility of that difference will depend on the type of diodes used.

Finally, the diode pair going to ground is NOT responsible for the fuzz.  Break their ground connection, and the sound will be about 95% as fuzzy as before.  They are there to provide a quick and dirty peak limiter.  The little feedback network between Q2 and Q1 helps out by emphasizing the note fundamentals a little more, but pretty much any analog octave fuzz will obscure the octave initially in a "harmonic haze" of everything getting doubled at the pick attack.  By the time the string has settled down a bit, to lose the clutter of unwanted harmonics to reveal the octave, the volume has dropped.  The diode pair to ground is there to maintain an approximate equal volume between initial pick attack, and when the octave "blooms".  Of course, those diodes will clip a little (which is where the other 5% of fuzziness comes from), but they are not necessary for any fuzz.  Stick a 100R resistor between them and ground, and strap a small-ish cap (e.g., 470-1000pf) in parallel with the diodes, and whatever fizz they might add will be reduced, without impacting their peak-limiter role.

[Christoper]

This is all great information. Fiddling around with the pedal a little more makes me realize that it's just about perfect as is.

It is a very neat little circuit

[antonis]

What Mark said about signal full-wave rectification: 
(not exactly but you get the idea..)