Superfuzz: comments/concerns ... Y'alls' two cents, please?

[Brossman]

Hey guys,

I've been reading through lots of the old post concerning the Unicord Superfuzz to gather/implement new ideas.  Let me CLEARLY STATE that MY PURPOSE for this project is to take an already very amazing and powerful pedal and give it more tonal variety.  First, I have some definitive points of inclusion in this project, then some questions at the end.  For reference, here's the schemo...



I've decided upon 2n3904's for a comp between "good tone" and noisy circuitry... I heard too many "sterile" comments regarding 5088/9's, so even though I have a boatload, it's a no-go on these guys.

---Adding a 1-2Mohm pulldown resistor at input...is this necessary?---

I'm implementing a 4p5t rotary to have clipping diodes in the following setup:
1 -> 3-Si, asymmetric
2 -> 2-Si, symmetric
3 -> NONE
4 -> 2-Ge, symmetric
5 -> 3-Ge, asymmetric

----1n34a's are a sure thing for the Ge diodes...any advice for the Si clippers?----

I'm including the 10k-trimmer at Q4,5 as an external control, so one may actively "dial-in" the octave (or not, respectively).

Controlled through a stomp, the tone switch is cool, but a 50kB-pot is said to give a wider tonal range... thoughts?

----On this schemo, every cap is shown as NP...is this kosher? or should I use electrolytics for any particular reason?----

Has anybody redone a schematic for modern Si-trannys with the different pinout?  I've seen some schemos, but seems wrong, swapping in 2n2222 for 2sc828 pin for pin.  I know this is wrong.  In any case, I volunteer to attempt it, if somebody offers to verify...?

----Credit to R.G., Mark Hammer, and MarieG, amongst many other unnamed for their inspiration.----

[petemoore]

---Adding a 1-2Mohm pulldown resistor at input...is this necessary?---
  If it pops when using bypass switch..a good measure to just put the pulldowns in there.
  I'm implementing a 4p5t rotary to have clipping diodes in the following setup:
-> 3-Si, asymmetric
2 -> 2-Si, symmetric
3 -> NONE
4 -> 2-Ge, symmetric
5 -> 3-Ge, asymmetric

----1n34a's are a sure thing for the Ge diodes...any advice for the Si clippers?----
  They need to be in there, put some in there or try varying the threshold with some other threshold diodes.
I'm including the 10k-trimmer at Q4,5 as an external control, so one may actively "dial-in" the octave (or not, respectively).
Controlled through a stomp, the tone switch is cool, but a 50kB-pot is said to give a wider tonal range... thoughts?
  Check my memory, Mark suggests a 50k pot to replace the switch makes the notch adjustable.
  ----On this schemo, every cap is shown as NP...is this kosher? or should I use electrolytics for any particular reason?----
  Nope, two polarized electrolytics can make a non-polarized El-cap:
  <>...series with no polarity showing [ie -'s on the outside, or +'s on the outside
  or
  paralled...showing no polarity with a + and - twisted together on each side.
  This is how non-polar eletrolytics are made...not that there's anything wrong with it .
  These positions probably see inverted polarity, hence the call for non-polarized component there.

[Brossman]

These positions probably see inverted polarity, hence the call for non-polarized component there.

My bad, I forgot electrolytics can be NP, as well... In any case, I'm gonna go out on a limb and guess that because this is Unicord, these were not your specialty Hi-grade audio caps...? I mean, were ANY of the originals radials? or were they cheapo ceramics?

On this note, anybody got any gut-shots (original or clone)?

Regarding the diode, Mark suggests that w/o any diodes, Superfuzz produces a very warm distortion tone...

[tubelectron]

Hi Brossman,

Here is the schematic of the SuperFuzz I built for my daughter somer times ago.

What's different from the std Superfuzz :

1 - increased gain transistors for Q1 (input), Q4-Q5 (octave), probably a bit more recent than the 2SC1828.
2 - "More" switch, which boosts the gain of the 1st stage on request, giving more fuzz density and increasing the scoop effect.
3 - footswitch for mid-scoop tone instead of slide switch, for instant shift.

To add REAL more tonal variety, I would have worked on the Tee-Bridge tone-filter values, and maybe switch them via a rotating selector. But my daughter found useless to go further, so...



A+!

[Electron Tornado]

Fuzz guts:   http://members.fortunecity.com/uzzfay/home.html

The Super Fuzz is a cool pedal. I got to play with an original orange and blue example several years ago that someone loaned me.

It looks like the tone switch selects between a fixed high pass filter and what I kind of looks like a notch filter. I looked through a book on filters, but didn't see an example quite like that. Does anyone have a reference with a good explanation of that filter?

[Brossman]

@Bruno - Thankyou!  Your layout is one I have been studying extensively for some time now.  I'm particularly interested in your boost feature... I love fuzz, why not admit it?

@Tornado - Wow, what a great site! they even had shots of the Bee-Baa (found that schemo, albeit a different one, on a google search 3 yrs ago...never knew exactly what it was, until now!)

actually, here http://members.fortunecity.com/uzzfay/beebaa/af100.html is the BB schemo... I noticed that it has a similar character to the SF.  Can anybody tell me what this is/does?

Here's what I'm looking at...some kind of notch filter, perhaps?  First pic is BB, second pic is SF...

[tubelectron]

@Electron Tornado & Brossman,

Yes, these are notch filters called "T-bridge". Here is the (very) theorical formulation for the tuning frequency :

F = 1/2.PI.R.C , with C = SQR Root (C serial . C to GND) if the 2 R are equal.

And if C to GND = 10.C serial, then :

F = 1/2.PI.2.R.C to GND

A similar notch filter exists on many fuzzboxes, amon others : Shin-Ei FY-2, Uni-Fuzz, Ace Tone FuzzMaster, Roland BeeBaa AF100, Ibanez Std Wau-Fuzz... The purpose of these filters are to produce a mid-scoop in the frequency response, like we can often hear it on Metal guitar tunes.

Here is the schematic of my Sbrodj Fuzz : an FY-2 with additional output recovery stage to compensate the loss due to the original mid-scoop notch filter. It gives a psychedelic broken sound...



A+!

[Brossman]

Okay, so referencing the Unicord schemo for the T-filter, if I were to implement Mark Hammer's "50kB-pot instead of a switch" idea, would lug 1 and 3 go on either side of the 0.001uF cap, then connect wiper to the 0.1uF cap to ground?

[tubelectron]

Well, I think that the best would be to make a trial and see if it's a conclusive idea, to get a reliable answer. I wouldn't be able to tell if that will alter the tuning frequency or the notch amplitude, or may be both...

A+!

[Brossman]

How about this schematic?  Can anybody verify the pinout on this?  Compared to other SF-schemos, it seems like the 2n2222's are just swapped in...

[slacker]

if I were to implement Mark Hammer's "50kB-pot instead of a switch" idea,

To do this the pot goes between the bottom of the 0.1uf cap and ground, the wiper (lug 2) and one of the outer lugs to the cap and the other outer lug to ground. With the pot turned one way the cap is connected to ground so you get same sound as the scoop setting of the tone switch. With the pot the other way you get essentially the same sound as the unscooped setting. In between you get a nice range of tones. 

Not sure what you mean about the pinouts, whatever transistors you use the base, collector and emitter connect to the same places in the circuit, so all the schematics will look the same. The pinout only matter on the PCB layouts, because different transistors might have the base, collector and emitter leads in a different order.

[Derringer]

I honestly don't think that having all of those different diode configurations on tap will make too big of a difference in the overall sound. There's nothing subtle sounding about the superfuzz circuit to begin with. I'd recommend trying the different combos out on a breadboard first to see if they do make much of a difference or to just simply see which configuration sounds the best and go with it.


Otherwise, the superfuzz that's on my pedal board right now follows this schem


and this vero



I added some sockets between the base and collector of the first and last transistor stages for treble-canceling caps. I don't think it makes too much of a difference in the first stage since there's already so much treble roll-off in that feedback network but I really like having a 390 pf cap between base and collector of the last gain stage.

I used a 0.056uf cap in the T-Notch filter instead of the 0.1 because I didn't like how far the 0.1 uf cap scooped the mids. A 10K linear pot was all I needed here between that cap and ground to make the mid scoop variable.

I wouldn't bother with an external control to dial in the octave. I had that on an earlier design and it was pretty useless. The sounds you got even at the extreme throws of the pot didn't sound all that unique. It sounded best with the octave dialed in. So I just left that as an internal trimpot.

The last mod I did was to install that 10K trim between the germanium diodes and ground. It limits how hard the diodes clip the signal so that they don't totally clamp down on the sound.  To my ears, it warmed things up nicely. I have the pot turned so there's about 5K resistance there.

[Brossman]

Wow, thanks for such a comprehensive explanation! Much abliged!

[Electron Tornado]

@Electron Tornado & Brossman,

Yes, these are notch filters called "T-bridge". Here is the (very) theorical formulation for the tuning frequency :

F = 1/2.PI.R.C , with C = SQR Root (C serial . C to GND) if the 2 R are equal.

And if C to GND = 10.C serial, then :

F = 1/2.PI.2.R.C to GND

A similar notch filter exists on many fuzzboxes, amon others : Shin-Ei FY-2, Uni-Fuzz, Ace Tone FuzzMaster, Roland BeeBaa AF100, Ibanez Std Wau-Fuzz... The purpose of these filters are to produce a mid-scoop in the frequency response, like we can often hear it on Metal guitar tunes.

Here is the schematic of my Sbrodj Fuzz : an FY-2 with additional output recovery stage to compensate the loss due to the original mid-scoop notch filter. It gives a psychedelic broken sound...



A+!

Merci' Bruno. It looks very similar to the twin-T, but without the resistor to ground.

[Mark Hammer]

Check out Madbean's clone project of the Z-Vex Octane.  It's a sort of über-Superfuzz, with a variable mid-scoop, instead of two switch positions.  Nice sounding pedal.

[wookie]

One mod I've always wanted to attempt on a super-fuzz but never got around to bread boarding was an option to bypass the Q4/Q5 pair and go straight to the clipping diodes.  A SPDT switch inserted in the Q4.C/Q5.C/10uF net could be used that would let you either connect that 10uF to the Q4.C/Q5.C net as usual, or pickup the signal from the base of Q3.  Being able to cut-out the octave would really increase the versatility of an already great fuzz.....if it sounds anything like I think it would, that is. 

[Mark Hammer]

If you want to cancel the doubling, just do what the Foxx TM does and lift one of the two complementary signals, whether the one coming from the emitter or the collector of Q3.  It takes both of them to make the octave.

[tubelectron]

@Derringer : good job, indeed... & interesting ideas notably about the notch filter. When changing the setting of your 10KB TONE pot, do you notice a variation in tuning of the filter ? Combined with a variation in amplitude or not ? I should experiment more about that...

If you want to cancel the doubling, just do what the Foxx TM does

Mark is right : on the Foxx Tone Machine, there is a switch to de-activate the octave function (which is by the way more drastic and aggressive than on the SuperFuzz).
When the octave function is off, the sound becomes similar to Big Muff. It's what I have on my FoXX TM pseudo-clone, built from the GGG PCB, with a "More" gain switch and footswitchable octave operation :




A+!

[Mark Hammer]

On my Superfuzz, I have a 3-position toggle that implements either Ge diodes, Si diodes, or no diodes.  Even without any clipping diodes, there is still plenty of distortion left.

Note that the FTM places a diode in series with each half signal coming off the phase splitter (Q3).  This results in some crossover distortion, both in the octave and non-octave positions.  I've installed a 3-way toggle that  provides octave, non-octave with X-over diode, and non-octave without X-over didoe.  Makes an audible difference, but again there is still some distortion left.

[Derringer]

@Derringer : good job, indeed... & interesting ideas notably about the notch filter. When changing the setting of your 10KB TONE pot, do you notice a variation in tuning of the filter ? Combined with a variation in amplitude or not ? I should experiment more about that...

thank you!

there isn't much volume loss at all really
I'd say only a perceived volume loss because the sound just won't cut through as well without some midrange
I practically never run it full scooped

having the mids variable on this pedal is imo the best mod to do

to learn more of the ins and outs of that filter look here:
http://www.diystompboxes.com/smfforum/index.php?topic=25788.0