Pleasing tweak of a Foxx Tone Machine

[Mark Hammer]

The stock Foxx Tone Machine is an excellent octave-up fuzz.  Normally, it uses a pair of germanium diodes for extracting the octave-up, and second pair of GE diodes just ahead of the tone-circuit for adding a bit more clipping.  Using the Madbean "Wolfshirt" layout, I subbed a pair of 1N5817 Schottky diodes in place of the octave-doubling set, and a pair of Si 1N4148 diodes in place of the second Ge pair used for clipping.

The result is that, by lowering the required forward voltage about 100mv, via subbing the Schottkys for the GEs, octaving is easier to get.  As well, by raising the required forward voltage for additional clipping, using 1N4148 instead of 1N34 or similar, it is possible to get some grunty sounds, with very high output levels, even at low drive..

[Mark Hammer]

It never fails.  I think I've done something innovative and find out someone has done it long before.  Paul Nelson posted an octave-up fuzz here, based on a hex inverter, but also using 1N5817s for rectifying.

http://www.diystompboxes.com/smfforum/index.php?topic=69308.msg555208#msg555208

[bool]

You should try some BAT-type schottkys for octave-up. (bat41, bat43..)

[Mark Hammer]

Not to be impudent, but why?  In this application, there is nothing special about the 5817, other than being readily available and having a lower forward voltage than a germanium type.  It there were diodes that could do the rectifying and had a Vf of, say, 100mv, that would make a difference.

[bool]

Eh ... datasheets will reveal a couple of electrical merits that can be had with these small schottkys (hint: parasitics and rev. leak), but I will reveal the two most important ones here:

- the small ones usually cost less (if you shop around)
- the small ones look better

... so there: no nonsense; plain and simple

[Mark Hammer]

Well there's some Boolean logic for ya!   

I may actually have some of those in my "diodes" drawer, but the 5817s are as big and obvious as my 1N400x types, so I found them first.

[Plexi]

Finishing it as my late night building... and making some research about the diodes on it I found that great advise.

Thanks Mark!

[Mark Hammer]

After some discussion with Dino/digi2t at the Montreal pedal show on Saturday, here are some further tweaks to perfect a Tone Machine.

The diode pair to ground are not a requirement for distortion.  Lift them and you'll find that the circuit is actually still pretty distorted, whether in the octave or non-octave mode.  I suspect that their function in most octave-up units of a similar sort of design is to simply clamp the output level, and impose a bit of compression. so that the octave will appear to bloom.  Normally, the signal amplitude would drop after initial pick attack.  However, the initial pick attack has lots of harmonic content.  When that content is doubled, it obscures/masks the octave.  After a brief period, the harmonic content dies away, and brings the fundamental into prominence, which the circuit doubles to produce an audible octave.  But the level would normally drop just as the octave becomes more prominent.  The diodes to ground are to hold the signal level relatively steady, so that once all the noise and obscuring harmonics are peeled away, the octave can seem to hold its place and level.  That's how it "blooms".

BUT...the diode pair introduces clipping as it clamps the level, which also produces additional harmonic content. We want the octave to stick out, without anything else covering it up.  Dino has a great solution he noted to me.  He sticks a small resistance between the diode pair and ground (let's say a 2k trimmer, for argument's sake).  This softens the clipping to reduce some of the unwanted harmonic content, but still works towards keeping a steady level.

On top of that, I tried placing a small-ish cap value in parallel with the diodes (and now trimmer).  Something in the range of 10-47nf will shave off even more of the unwanted harmonic content to round off the tone and make the octave stick out more.

[Plexi]

Thanks Mark

Oh boy... You'll make me use dpdt switch with 2 pairs of diodes, with sockets on it to test.
I'll try 1N60 and 1N4148. 
And I'll use on/off/on with two pairs of diodes in output diodes (to ground).

It seems that the diodes are an important factor here, more the octave pair, and if they're matched.

About the small resistor and cap, in parallel with the pair to ground:I was thinking in some potentiometer (100k?), middle lug to ground, and outer lugs with both pairs to output. That way you can pan between both ways.
Is that a good idea?


AND I have there a SHO waiting if is true that this one have small output volume... I'll let my ears to decide.

[Mark Hammer]

The diode pair that takes the signal off the emitter and collector of the phase splitter should be matched for forward voltage.  A pretty simple kind of matching, and generally pretty easy to accomplish; moreso than matching transistors.  If you have them within 10mv of each other (e.g., 307mv and 298mv), you're in good shape.

The clippper-softening resistance doesn't have to be very big.  Really, 5k is probably overkill.

[Plexi]

You got this Mark...

There's no simple way, due the complicated position in the schematic, of switch the diodes with dpdt switch.

I'll have to try switching one, or maybe try in sockets.

[Mark Hammer]

Just to follow up, I just fired up my modded Green Ringer clone.  The GR lacks the diode pair to ground that the Tone Machine, and Superfuzz have.  It uses the same approach to generating the doubling that the FTM does, though.  And the octave is fleeting, with some "bloom", but because the level dies away quickly, you don't really get to hear the bloom.

UPDATE:  Added a germanium pair in series with a 1k resistor to ground on the output of my GR and the octave comes through better with the inclusion of the crude compression that the diode pair provide.  Recommended.

Again, empirical support for the hypothesis that the diode pair is NOT for producing distortion, but for maintaining level.

[Plexi]

Great Mark, I'll add that 1k res. and see how it sounds

My personal update: tried a few diodes, and the best octave comes with a 1N60 (226) and 1N4148 (584).
I used KSP13 (darlington) in Q1 to put more 'spicy' in the thing.
I will be extremist, and I will put darlingtons everywhere.

Anyway: the "fuzz" pot (50k) acts strange.
It clean abruptly in the last 5%.

[Mark Hammer]

When you say you tried a few diodes, do you mean added on at the output as I described, or do you mean the diodes normally sed to generate the octaving?

[Plexi]

When you say you tried a few diodes, do you mean added on at the output as I described, or do you mean the diodes normally sed to generate the octaving?

Sended to produce the octave.
As you said, I matched the pair (found both 1N60 with 226) in the drawer.
Tried a lot of diodes, just for fun (inquisitive me? neh... )

I left the output pair for later 
At the moment, there's there 1N60: GREAT punch and bass, even with octave on.

[bool]

1N60 germanium or schottky?

[Plexi]

1N60 germanium or schottky?

Schottky ones.
But IMHO, newer silicon schottky diodes performs well as older germanium ones.
As long I get less than 300R, I think there's fine 

[Plexi]

Resolved the Sustain knob abrupt cleaning: placing a 5k/10k resistor instead the 220R to ground.
OR just connect lug 1 direct to ground: it'll mute the signal all the way closed.

[aelling]

A neat mod to this pedal I just figured out by probing around is you take a 100k pot, connect lug2&3 to ground and connect lug1 to lug3 of the tone pot. It will make the tone control more useful, fully cw it will be close to the original tone machine (with or without octave engaged), as you go ccw it will scoop the mids much like a big muff without octave and a superfuzz with octave.

[Plexi]

A neat mod to this pedal I just figured out by probing around is you take a 100k pot, connect lug2&3 to ground and connect lug1 to lug3 of the tone pot. It will make the tone control more useful, fully cw it will be close to the original tone machine (with or without octave engaged), as you go ccw it will scoop the mids much like a big muff without octave and a superfuzz with octave.

Nice!
ll try it.. basically you're turning the tone stack to a BM one, adding resistance to ground in the high pass filter.