Superfuzz _farting_ (but not always)

[rankot]

Just finished Superfuzz few days ago. Sounds great, but I like it much better with Si diodes, or Si/Schottky asymmetrical distortion then with Ge.

When Ge is selected for clipping, I get some very nasty farting on note decay (especially on low notes), unless I add some resistance in series with diodes (Clip pot) which cures it. Very strange. I used this schematic for my build - actually the same as original, except for few added pots. And somehow I've dropped one 10u capacitor after clipping diodes. Could it be the problem? It seems that this farting is produced by excessive clipping of Ge diodes?





HELP!

[Mark Hammer]

1)  Nice graphics.  My compliments.

2) I have come to the conclusion that the post-doubling diodes are not there to "clip" in the sense of producing deliberate distortion.  Rather, they are there to provide a quick-and-dirty limiter, whose purpose is to make the octave more audible by maintaining a relatively steady volume level as the octave emerges from the fog of all those doubled harmonics and "blooms".  Lift the diode connection fro ground and you'll find that the remainder of the circuit does a more than adequate job distortion, but the octave is not quite as audible because the volume appears to die out just as the octave comes out of hiding.  As proof-of-concept, I added a diode pair (+small series resistance to ground) in my Green Ringer and it improved the audibility of the octave considerably.

Remember that the octave we want to hear is the doubling of the note fundamental, NOT the doubling of any other harmonic content.  The most harmonic content the string produces is at the very start of the note, which soon dies away leaving the naked fundamental.  This is also why we generally don't get a pleasing octave-up unless: a) we pick above the 7th fret, or thereabouts, where the string is stiffer (because it's shorter) and provides less harmonic content, b) we use the neck pickup, and c) roll the guitar tone back.

3) One wonders if a small-value cap between collector and base in the typical input stage might help in trimming off top end to yield a more audible octave.

4) Since you seem to be using a variable midscoop, I don't see R32/R33 serving any useful purpose.  Their original role is to match the level between an unfiltered and a midscooped output, when the "Tone" switch is installed...which you don't have.

5) I get the idea of the "Octave" control, but I suspect  little of its range is truly useful.  Once you disturb the balance between the two phase-reversed outputs by even a little, the octave disappears.  In which case, maybe it is simpler to just install a 1k trimmer to set for optimum octave, and use a SPST toggle to insert a 100k-220k fixed resistor in series to cancel the octave; not unless there are unexpectedly interesting (and easily replicable) sounds in range of that control.

6) As far as I can tell, the "missing" 10uf cap is there simply to provide some isolation between the diodes and the filter.  You will note that a number of posted schematics for the Superfuzz and related derivatives (Acetone, Honey Fuzz, et al.) show the pre-diodes and post-diodes caps as polarity-reversed.  That is, the negative side of the first cap is tied to the diodes, and so is the negative side of the 2nd 10uf cap.  Maybe that serves a purpose.

[rankot]

Thanks Mark, lots of good thoughts as always! Glad you like this graphic, but it's not mine just found it on Net searching for Superfuzz. It seems that there's a beer called like that

I left R32/R33 on my schematic just in case I decide to drop tone pot and add tone switch, but I didn't install them anyway. However, what I forgot to mention, is that I use CA3046 transistor array and I didn't measure hfe of its transistors. Maybe I should?

'Octave' control works fine and this range is quite OK; however, it didn't affect farting, because it was present before I added that pot. I'll try what happens with different picking, as you proposed.

In the meantime, I tried to add that missing cap, but no improvement either. Maybe I shall reverse it's polarity? Or maybe C6 and C13 polarity too shall be reversed?

[rankot]

I have just noticed that my R9 (100k) goes before R1, while on all other scematics' it's after it. My C2 is 1n, while it's 2n2 on original. Could it be the cause of this farting???

I didn't measure hfe of my transistors, just set the voltage on both bases to same value using trimmer, too.

Everything else seems to be the same as in original schematic.

[Mark Hammer]

Curious about your use of a CA3046.  I've been wanting to try that myself.  I did make one using a 2SC1583 for the mirror pair that does the doubling, and found the octave far superior to any I'd built before using discrete transistors.  The gain-matching of the two transistors is critical, and the 1583 provides it.  So naturally, I'm curious about what a 3046 does.

[rankot]

It sounds good  to me. However, I didn't measure hfe of the mirrored pair (those two are important, are they?), so now I must desolder the part to do that. I have another CA3046 by RCA at the desk and hfe for those two trannies are 151 and 158 if I recall well. Another CA3046 I have have considerably smaller hfe of around 60 or even less, but I didn't try it in the circuit.

What bothers me is the fact that when I simulate this in ltspice, I get quite obvious voltage differences in the circuit that show that most of my 10u caps are reversed, so I will have to pull the PCB out of the box and reverse them. Then I'll maybe try to desolder the IC itself, too. I don't know if that capacitor orientation can cause this farting, but I'll try that.

[duck_arse]

I recently finished a superfuzz build, and I matched my rectifying transistors on the dmm, at least. when I fired it up, I applied a test signal and "trimmed" for best octave equal on the cro. however, when I plugged in some audio and listened, it was horrible rising hissss as signal died. so I instead retrimmed for lowest hiss, and called it finished.

have you tried retrimming the pair? it might be a similar/the same problem, but differing symptoms.

[rankot]

I'll try. But my first hind was the capacitor polarity, since I've build it with few caps reversed, so I presume that when input signal is strong enough, it causes dc passing through caps?

I set the trim pot so I have the same voltage on both bases, but maybe I shall really try to play with that!

[rankot]

I have just reversed all the wrong oriented caps (also put 1u block caps on some places instead of 10u to cut some bass), and it seems that pedal is working much better, although there's still some "farting" or oscillating on very low notes (B/C/D/E on low tuned guitar). I didn't try it with bass, but noticed that farting is gone when Expand pot is not turned completely CW.

So if anyone has an idea what to do to get rid of it, please help, or I will have to live with this

[rankot]

OK, I decided to bite the bullet and desolder CA3046. Now I tried to measure transistors in it, and they are all around 100 hFE, but the second one (pins 3-4-5) is reported as a diode from pins 4-5, so it seems that it has an internal problem. The strange thing is that it worked well inside pedal, so probably I've burned it during desolder process, or there's some strange mojo going on.

I have another IC with hFE ranging around 150 and another around 60, so I will try them both. Now I've soldered a socket, just in case. And I will try with some regular transistors, too.

[duck_arse]

now we getting somewhere!

[rankot]

Tested both 60 and 150 hFE transistor arrays, and they both oscillate as hell (some kind of LFO), producing what I described as "farting" even more. Then put back the first one array (one with hFE around 100 and with second transistor reported as non-working), and it works great, with very little farting (if any). I'm puzzled.

I'm gonna pack this one into enclosure and let it be as it is.

[aelling]

I recently finished a superfuzz build, and I matched my rectifying transistors on the dmm, at least. when I fired it up, I applied a test signal and "trimmed" for best octave equal on the cro. however, when I plugged in some audio and listened, it was horrible rising hissss as signal died. so I instead retrimmed for lowest hiss, and called it finished.

have you tried retrimming the pair? it might be a similar/the same problem, but differing symptoms.

I finished two superfuzz builds recently and had the same problem, like a wave of hiss if you mute or as signal dies out. If you matched the rectifying transistors, try removing the trimpot and connect the two 22K Base resistors on those Q's straight to ground (like they do on this schematic).

That got rid it for one of my builds, the other built it was the trimpot and some dodgy (or rather fake) 1N34a "germanium" diodes from Tayda, I took the trimpot out and put some Russian D9K diodes in and that got rid of the wave of hiss.

[aelling]

I recently finished a superfuzz build, and I matched my rectifying transistors on the dmm, at least. when I fired it up, I applied a test signal and "trimmed" for best octave equal on the cro. however, when I plugged in some audio and listened, it was horrible rising hissss as signal died. so I instead retrimmed for lowest hiss, and called it finished.

have you tried retrimming the pair? it might be a similar/the same problem, but differing symptoms.

I finished two superfuzz builds recently and had the same problem, like a wave of hiss if you mute or as signal dies out. If you matched the rectifying transistors, try removing the trimpot and connect the two 22K Base resistors on those Q's straight to ground (like they do on this schematic).

That got rid it for one of my builds, the other built it was the trimpot and some dodgy (or rather fake) 1N34a "germanium" diodes from Tayda, I took the trimpot out and put some Russian D9K diodes in and that got rid of the wave of hiss.

Oh wow, my dumbass didn't notice I bumped the volume knob on my guitar down a bit when I was comparing the clones to an original superfuzz. So if you have that wave of hiss when you mute the strings or whatever, roll the volume knob on your guitar down a tiny hair until it disappears. I would add though the octave I got after removing the trimpot sounds way better than with the trimpot, I dunno if it's like a placebo effect or something but at least to my ears the octave sounds better.