FOXX Tone Machine VS UNIVOX Superfuzz

[RCZ53]

I am wondering if there is a tweak I can make to my 70's tone machine to make the octave effect more pronounced. I replaced one 2n3565 (i believe it was Q3) and it seemed to help the overall sound. The voltages on all the transistors pretty closely matched a printout I made off the net (i can't remember whose site. duh!) except Q3 which is why i replaced it however the reading was the same.
I am wondering if a new set of diodes would do the trick. It just doesn't have the octave effect that my old Apollo (Shin-ei superfuzz) has. But maybe that is just the nature of the beast. Any suggestions would be much appreciated. thanks all. regards Roger

[MartyB]

Hey Roger,

  My Foxx Tone Machine has a nice octave up.  Like all of the ones I've tried its more pronounced near the 12th of the GBE strings.  I used Japanese Germs in it.  My Tycho Octavia has a stronger octave effect, but there's these less palatable - I'd say -unwanted harmonic things in there.  The FTM IMHO has a much better Hendrix vibe.   Don't know bout the Shin-ei.  Lotsa choices, Green Ringer, Octup etc.  There was a recent thread on this you might search for.

MartyB

[Mark Hammer]

Of the various octave-up boxes I've made (Superfuzz, FTM, Mayer Octavia, Anderton Octave-doubling Fuzz, Elektor Octave Doubler, Green Ringer), the FTM is far and away my favourite and has yielded the "best" octave sound.  I've made two of them and both behave similarly without much attention to component selection required.  Not sure what trannies were used but absolutely certain none were 2N3565 (likely 2SC1815 or 2N3391, or 2N5088).

Ironically, I decided to hunker down and put the last few touches on a Superfuzz last night (it kept *staring* at me Tim; I couldn't help myself!).  It is one sick pedal.  We're talking flesh-eating disease sick, not nagging chest cold sick.  Still, I find the octave up attainable is not nearly as robust as my trusty FTM.

Keep in mind that the octave up in these assorted units comes from essentially splitting the two half-cycles and "folding over" one half cycle such that there are two "blips" in the space formerly occupied by one.   IN the discrete units, the two half cycles come from the phase splitter formed by the transistor with equal value resistors on the emitter and collector (4k7 in the case of the FTM).  Each output of the phase splitter has one half-cycle "trimmed" by means of a diode, and what remains is mixed together.  Mixing the two outputs in equal proportions is essential to this arrangement producing an easily audible octave sound.

There are a bunch of things that play a role yielding "equal proportions".  If you ever took a look at J.C. Maillet's mods to the Green Ringer, they include a "nulling" pot which adjusts the resistance for one of the phase splitter outputs to balance the two signal levels.  In the FTM, there are no series resistors to contend with but there ARE those pesky 100k resistors to ground, which may not be as equal as you assume.  Given the variation between diodes, it is also worth considering selecting the diodes involved such that voltage drops along each phase splitter output are approximately equal.  Finally, in an OOOOOOOOLLLLLLDDDDDD issue of Electronotes, his eminence Bernie Hutchins notes that optimum phase splitter performance may come from the emitter resistor being just a tad smaller than the collector resistor.  I am given to understand that "smaller" means no more than 3-5%, or what amounts to essentially the tolerance of the garden variety resistor.

Taking all these things into account, it *should* be possible to make the two half-cycles as equivalent in amplitude as is humanly (and electronically) possible.  As for the rest of the circuit, I gather it is all pretty much noncritical.

I might point out also that the tone control on the FTM is essentially a notch filter, such as found on the Superfuzz, with a variable tap-point on the 2nd resistor (which in this instance is made up of the 4k7 resistor and 50k pot).  Taking the Superfuzz/FTM comparison into ac@#$%, I will simply note that I found the octave effect more robust in the Superfuzz in the non-notch position (two-resistor divider to equate levels and nothing more).  Perhaps you might consider replacing the tone control of the FTM (which is pretty damn useless if you ask me) with the two-position thing on the Superfuzz to provide more midrange squawk.  Alternatively, consider upping the 3n3 cap in the tone circuit with a 4n7 cap for more mids.

[RCZ53]

Thank you both for your help. I guess the diodes in question would be D1 and D2 looking at JD Sleep's schemo. there is another pair but I am thinking they are for distortion. I will check voltages of D1 and D2 and resistance of the 100k resistors for starters. Thanks again I plan to look for the recent thread on octaves as well. best regards, Roger

[jmusser]

Mark, is the "nulling out pot" the same thing that RG referred in his excellent thread on the Super Fuzz? It was a 10k trimmer that went on the two 22K biasing resistors of the octave generator section, that RG said you use as an octave control on your Super Fuzz. I've also heard that more octave can be had by adding a second full wave rectifier to the 1st one. I guess this would be a bridge circuit, that would give you more "perceived" octave. I noticed John Hollis does this on his Titan Octave, and ZVex mentioned it the other day regarding his Johnny Octave. I didn't know how much the word "perceived" meant, until I read RGs reply on my post "How Octaves Work".

[Johnny Guitar]

I hope this isn't too contradictory: I really like the octave up on my FTM, but I have to admit I really only think of it as producing two distinct tones which are useful (at least with my Strat).

1) Back PU (bridge) with the guitar tone all the way up especially on the two unwound strings (high E and B). Lots of nasty overtones. I believe Belew used this a lot in the early 80s.

2) Front PU (neck) with the guitar tone all the way down. This works best at the 12th fret (as was mentioned above), but the effect is not what you might expect (it's not like that Cream era Clapton "woman tone"); a very bright clear piercing tone.

Sorry, I can't answer your question as to tweaks. I had thought about adding a foot switch to roll the tone off (to simulate #2), but the effect isn't nearly as good when using the back PU for that one.

John

[bwanasonic]

1) Back PU (bridge) with the guitar tone all the way up especially on the two unwound strings (high E and B). Lots of nasty overtones. I believe Belew used this a lot in the early 80s.

I also remember that Mr. Belew mentioned that he was only satisfied with the FTM direct to the board and that he couldn't seem to get a satisfying tone with FTM>amp. In general the most pronounced octave effect is usually from neck pickup w/ tone rolled back.

Kerry M

[tomtom]

As far as I know (not far !), the FTM sounds great with low gain transistors (2222 with 200 hfe for me), 1% resitors on the phase splitter and matched Ge diodes help achieving a good octave.
The Superfuzz can't do a good fuzz alone (I mean I have not find a way to cut the octave in a good way soundwise, the fuzz is kind of gated) but the octave on this one is quite good.

 I think the FTM needs a better tonestack, like AMZ presence control or a superfuzz notch made adjustable. Keep in mind that the notch  seems to give a more detailed octave.

 I wonder what's happened in the first two stages of the FTM.

 I have found that a small serie resistor inline with the lower diode in the wavefolder seems to give the octave a more crazy sound, nice.
 A small cap to ground on the input (only in octave mode) seems to cut some highs (for a more natural octave sound)

 Love the FTM but I keep my Apollo when I need big big sound (great with a phasor !)

[RCZ53]

Can you tell me how to match the diodes? I have some NOS Ge ones from smallbear that should work. thanks
Roger

[Mark Hammer]

Though not matching for *everything*, if you have a DMM with a diode tester range (it's actually part of the resistance-testing settings on mine), essentially what you do is measure the forward voltage drop of the diode.  Black/ground lead goes to stripe, and red/pos lead goes to the other end.  Germanium diodes should show you a reading of anywhere from around 190mv to 280mv, though 230-250 is most common.  Silicon diodes can give you a reading of anywhere from 450mv or so up to 600mv or more sometimes, though 510-560mv is typical in my experience.  You don't need to match them to 1% or anything anal-retentive like that, but if you had, say, three Ge diodes measuring 193mv, 212mv and 221mv, choosing the latter two as your pair after the phase splitter ought to, in theory at least, assist in getting a better octave balance.  Though, you can gather from my previous post that they will certainly not be the only factors that do that.  Every little thing helps though.  If you are fortunate enough to have a pair that are even better matched than that, go with them.  The sole caveat I would offer is that those with LOWER readings (e.g., 193mv + 202mv vs 212mv + 221mv) will be preferable for other reasons, namely the fact that they will eat up less signal, yielding a little more "push" when it comes to distortion further along in the circuit, and a slightly better S/N ratio.  On my Superfuzz, I installed a 3-way switch to select between no diodes, Ge diodes and Si diodes.  A useful mod IMHO.

The nulling trimpot shown on SOME (but not all) Superfuzz variants serves the same purpose as the one J.C. has on the Green Ringer, but does not do so in the same way.

My own experience (though admitedly, "ideal" tone control settings depend a LOT on pickups characteristics, picking style and amp-speaker properties) is that introducing a notch tends not to help in making octaves seem more noticeable.  In part this would seem to be because the notch is often in the zone of the octave fundamentals.  The harmonics of the octave get lost amidst all the other buzz arising from the distortion.  One of the reasons why many find octaves more noticeable when using the neck pickup with the tone rolled off.  I found it ironic that when switching from no notch to notch on the Superfuzz I finished the other night, it seemed to switch from octave unit to "merely" a splatty fuzz.  That may or may not have been because of how the trimpot was adjusted, but it was fairly clear to me that attenuating the fundamentals, and shining the spotlight on the upper harmonics made the otaves all but disappear.

[jmusser]

Thanks Mark, good information as usual. The up octaves are just something that really intrigues me, and I tend to like their bite over conventional face type fuzzes. Sometimes I like that silky sound, but in general I really like the nasty edge of the up octaves. You can darken them with your amp, and really get some growl. I'm still trying to figure out if there is a way to reference at a higher level than half the fundemental voltage, so you could get a true 2nd octave up, so that signal can be mixed in with the "standard" octave up. Of course, after reading RGs post on octaves, I know that has a lot to do with perception as to what your ear percieves, and what frequency you're actually at. What I "think" I want is a circuit that starts you out on the nut of the guitar, at the same frequency you would normally get at the 12th fret with an up octave pedal. I have no idea if this is even possible. Then, be able to blend the "normal" octave up frequency with that. Just like an MOC-1 does down octave.

[Mark Hammer]

Well here is where R.G. or maybe Ton or JC would need to barge in and explain what it is about the various phase-splitter-based octave up boxes that makes something like the Green Ringer sound more like a ring modulator than an octave box, and the FTM deliver a "purer" octave.

Clearly there are ways of splitting and recombining half-cycles (derived from a discrete phase-splitter and a pair of diodes to extract relevant half-cycles from each phase splitter output), and then there are ways....

When JC Maillet was visiting this region a few summers back, and dropped by my place, we spent some time talking about octave-up boxes and the point at and conditions under which notes start to generated sums and differences rather than *just* octaves.  Part of the fartiness that sometimes accompanies octave-up boxes is the manner in which the RM characteristics start to "bleed in".  A light went on in JC's head regarding the math behind it, but I have to say that the Greek letters did not come so freely to my own.  I think this is a matter that deserves discussion, and a comparison of the assorted ways that designers have implemented "split-'n-separate" strategies for octave effects, and their audible impact.

Personally, I don't see either aspect (pure octaves vs RM "fartiness") as the ideal or goal.  Rather, both are musically useful and valid, and being able to get the balance you want WHEN you want is the ideal.  If I have an ideal, it's a guy like Jeff Beck who seems to have the art of selective intermodulation effects mastered.  Dialog about JB can be located elsewhere, but I raise his name just to give a sense of the extent to which shifting the balance between octaves and RM on the fly is a tool of musical expression in the right hands.

[jmusser]

I hope RG does get hold of this topic. There are several up octaves as you have mentioned, and they all have a different amount of clarity, brightness, fuzz, background mayhem, sypathetic strings, clang and unwanted harmonics. So far the one that I've built that sounds different from the rest, has been Tim's Octup Blender. I'm not sure exactly what type of harmonics, and octaves are going on in there because they are so well blended that you can't really differentiate between them. But, it has the eerieness that the double down octave of the Arion MOC-1 when the fundemental signal is stripped off. It has a very thick up octave tone to it, that I have not heard on any effects that I have built, or any other sound samples of up octaves. The fuzz is especially thick and synth sounding.

[RCZ53]

Just wanted to thank everyone for their comments. I replaced the diodes with a matched pair of NOS germanium diodes and I am getting a much better octave effect. The readings I got on the old pair with my DMM was 220 and 402 (not sure if that is Ma or what)  Then using the same DMM i sorted through about 15 diodes to find a couple reading 421 and 428 I haven't messed with the 100k resistors yet but I am curious to see if that would fine tune the effect even more. Again thanks and special props to Mark Hammer. regards Roger

[Mark Hammer]

Bigger props to you for your persistance and willingness to experiment.  

Yep, if you take two identical twins and make one of them stand beside and 40 feet behind the other, they aren't going to look all THAT identical.  That's what happens when the two diodes are that discrepant.  And having the two "twins" (the positive and inverted negative half-cycles) stand shoulder to shoulder in equal amounts is what makes octaves come through.

You may be able to improve the octave a little bit more, but my guess is that you've gone a great deal of the critical distance already.  Maybe now's the time to kick back and enjoy it, though if there IS any more noticeable improvement to be made by matching of other components, I think owners/builders of octave-up fuzzes would be very interested to knwo your findings.

A big thanks back in your direction for the feedback and followup.  Completing the circle is what makes postings here so valuable to others.

And to all those reading this, don't you think it's time we all got our heads together and made a combined octave up/down box with 3-way mixing?  The Super Octaver at my site is one thing, but certainly we can do better, can't we?

[jmusser]

I wish I had a lot of technical stuff to add to help, but my ignorance abounds! One thing I noticed on Tim's Octup Blender, was that it seemed to be hooked up in parallel. The original signal was in parallel with the up octave, so that there was a clean transistion between the two. It looks like that would be one of the big things in a circuit like you're describing. If the different otaves were all in parallel, then they could be mixed separately anyway you want, as long as they go in common and out on a common output. I have done this with my MOC-1, and Tim's Octup Blender. You talk about eerie :shock:  Now, if I can only find a tech type to make me a true second octave up to mix in. Yea Baby! Oh yea, why is it that most up octave circuits use germanium diodes? Is it because of the lower clipping voltage and how it effects the wave form?

[Mark Hammer]

why is it that most up octave circuits use germanium diodes? Is it because of the lower clipping voltage and how it effects the wave form?

Excellent question.  My own guess is that:

a) a low of these effects (especially all Shin-Ei derivatives) are actually of a much older vintage (late 60's) when Ge diodes where much more available and much more common.

b) Ge diodes DO eat up less signal, requiring less gain recovery and providing less hiss overall.

Note, however, that in some instances, the Ge diodes are not at all necessary for the octave aspect, only for the distortion/fuzz aspect.  For example, I got a Superfuzz up and running last week, and installed a 3-way toggle to either lift the post-octave diode pair completely, use a pair of Ge diodes or a pair of Si diodes instead.  The switch changes the nature of the distortion achieved (and output level), but not the "octaveness".

[petemoore]

Series a couple Si's for each Ge...Edit: Doh...Nevermind, captain Jack Splayed the fan.

[puretube]

information lost...

The Flageolator.


btw: to a 50/60Hz sinewave, all (perfect) 2way rectifiers are created equal.

[jmusser]

I like that idea a lot, but not being a tech type, I don't have much to offer. I'd just like to get hold of it after you come up with it!  :shock: On Tim's Octup Blender, it has very good separation like I said. From what I can make out, the circuit is in parallel. It seems like that it what you would want to do if you were planning to mix and match octave, so one wouldn't be driving the other one, to make trash out of it. The MOC-1 that I keep talking about (and praising) does the same thing with down octave, where they are completelty separated from one another. I've never seen the schematic, but I'm guessing that's how it would be. Am I out in left field with my thinking here? Now the tricky part, making a true second octave up that is also selective, in the same way the two octave downs are in the MOC-1. i have no idea what type of gymnastics I'm talking about here, but it looks like all these features have to be separate inteties that are re joined at the output.