Univox Square Wave fuzz build...

[LucifersTrip]

I finally got around to building this one and I wasn't able to find any build report for years....And that old build also did away with the unique "distortion" control, which I didn't want to do...so here goes...

Original:


The first thing I did was try a bunch of different JFETs. The classic J201 won out. All others gave me a more gated, broken sound.

There were 2 problems. It was way below unity and the fuzz was still mildly gated with too much pick percussion (clickiness) evident.
It turned out to much easier fixes than expected. I built two versions:

#1
--------

Volume problem:
To increase volume, I slowly increased the vol pot. Even at 1M, it was still below unity. So, I added diodes in parallel. Finally it was above unity, but now the distortion (really a tone control...closer to 0K the more bassy) didn't give me a good sweep.  So, I simply increased the 27K cutoff to 100K and subbed a 1M distortion pot instead of the 250K. Beautiful!  That gave me a nice sweep from just barely too bassy to a nice thin fuzz.

Mild Gating Problem:
Anytime I see an odd small value from an emitter, bass or collector to a rail, I always suspect it to be very sensitive, and it was...I slowly increased from 0K and hit a perfect spot around 2.6-2.8K. The mild gating and pick percussion were gone, resulting in a killer smooth fuzz with super long sustain.  It was nearly perfect, but a bit too harsh on the highs. A simple tweaking of the 680 gave me a sweet spot right around 1.8-2.2K. If you raise it much higher, you'll wind up with less power...and around 4-5K will start to gate again. Further, I increased the 2M to 3M, which gave a bit more richness to the higher strings. I tried up to 7M...still good (thicker on the high strings)  but I didn't want to take a chance it would change the tone too much.

Here's the final with voltages...It's a pretty loud (will feedback at full) fuzz with a good sweep from fat to thin and killer sustain. It really doesn't sound exactly like many of the other fuzzes. The only thing it slightly reminded me of was a Carlsbro Suzz (v2):




#2 (cheapout version)
--------

Volume problem:

Well, there's no cheaper way to increase the volume than remove the diodes...and in this circuit, the diodes don't do nearly as much as others...just a bit more fizz.

I removed the diodes, but realized that if I left one in place it gave more richness to the high strings and killed a little harshness. With the increase in volume, came an increase in bass & boomy-ness. After trying numerous values, I decreased the vol pot to 20K which solved the problem. Like in version 1, I lost a good distortion sweep, so I increased the cutoff to 100K and the pot to 500K. In this case 1M thinned it out too much.

Though I could've probably tweaked the 3.9K and 680, I didn't since there was absolutely no gating or pick percussion with the diodes removed. Also, I built this version first, so I didn't think about doing that, since it was unnecessary.



have a blast

[Mark Hammer]

I made myself one a while back and also dickered around with it.  It is, fundamentally, a Fuzz Face, with some bells and whistles.

Which ones did *I* add?

1) I installed a 3-position SPSDT toggle to select between a pair of GE diodes to ground, SI diodes to ground and no diodes.  Big volume increase with the latter, but a nice wooly warm overdrive tone.

2) The SW comes with a fixed gain.  The heck with that!  I installed a 2k or 5k pot between the 33uf cap and the ground end of Q2, to vary the resistance in series with the cap.  Works like a charm.

3) The existing tone control was "different" for its time, and is basically a bass-cut control, ranging from woofy to nasal mosquito tones.  FEH!!  I replaced it with a 10k fixed resistor just ahead of the volume pot and a second 3-position toggle to select between treble-cutting caps.

It is now a VERY nice sounding pedal, with tones ranging from a pleasing warm growl (max treble cut, lower gain, no diodes), to a full-tilt sizzle.

There is a lot to work with in this circuit.  Glad someone else found it rewarding too.

[jrod]

2) The SW comes with a fixed gain.  The heck with that!  I installed a 2k or 5k pot between the 33uf cap and the ground end of Q2, to vary the resistance in series with the cap.  Works like a charm.

Hey Mark. Did you set up the pot just like a Fuzz Face "Fuzz" pot replacing the 3K9 resistor to ground from Q2 emitter?




LT - Thanks for the build report. I always enjoy reading your fuzz posts!

[Mark Hammer]

Hey Mark. Did you set up the pot just like a Fuzz Face "Fuzz" pot replacing the 3K9 resistor to ground from Q2 emitter?

Nah, just a variable resistance between the cap and transistor.  The "classic" use of a 1k or 2k pot with the cap connected to the pot wiper is simply a way of providing a fixed and variable path with fewer parts (i.e., one less resistor).  Using a 5k variable resistance between cap and FET effectively removes the AC bypass, such that gain is dictated by the drain and source resistors.

Electrons don't really care.  They're just looking for an easy route to ground! 

[jrod]

Oh ok, I see! Thanks for the info Mark!

[alex_spaceman]

Very interesting, I got my eyes on the SW and I think it won't be too long before I give it a go...

About the v2, did it make any difference whether you removed one or the other diode?

[Mark Hammer]

It never matters which of the two diodes one uses.

[LucifersTrip]

I made myself one a while back and also dickered around with it.  It is, fundamentally, a Fuzz Face, with some bells and whistles.

1) I installed a 3-position SPSDT toggle to select between a pair of GE diodes to ground, SI diodes to ground and no diodes.  Big volume increase with the latter, but a nice wooly warm overdrive tone.

2) The SW comes with a fixed gain.  The heck with that!  I installed a 2k or 5k pot between the 33uf cap and the ground end of Q2, to vary the resistance in series with the cap.  Works like a charm.

3) The existing tone control was "different" for its time, and is basically a bass-cut control, ranging from woofy to nasal mosquito tones.  FEH!!  I replaced it with a 10k fixed resistor just ahead of the volume pot and a second 3-position toggle to select between treble-cutting caps.

There is a lot to work with in this circuit.  Glad someone else found it rewarding too.

yes, exactly...yours was actually the build I was referring to...It was cool, but for me, it honestly changed the unique character of the circuit turning it into a more standard controllable fuzz....which I guess might be better for modern standards.  you also actually noted that it may have not been the right choice: "I wired up a Superfuzz-style midscoop filter, in place of the 250k+27k/.001uf network, and while it wasn't awful, in retrospect, I don't think it necessarily suits this thing."

With the diodes removed and no vol pot change, it's way too different & too boomy...and didn't have enough volume/power for me in the diode position.

when I first plugged in the original...my first thought was, "Oh no another Shin-E FY-2 silicon. This is gonna need a boost stage."  But I never tried that. That may be the best way to keep the original character of the circuit and get some decent volume. I would definitely recommend adjusting the 680 & 3.9k. I'd find it hard to believe that those exact spots would give the best sound right off....

[Mark Hammer]

One way to make it less boomy is to simply reduce the value of the 33uf cap so that the gain is applied less to the low end.  On strategy I've tried in past with some success is to tie the wiper of a pot (and 5k would be just fine here) to the ground side of the transistor, whether bipolar or FET, and run two different values of cap to ground, one from each outside lug of the pot.  At midpoint, you get no additioal gain.  Rotate in either direction you get two different kinds of gain.  Towards the 33uf side you get gain applied across the full spectrum.  Towards the other side you get more gain applied to mids and highs, depending on the value chosen (consider 3u3 to start).  I gather the net effect is similar to the variable bass bypass that Joe Gagan so elegantly incorporated into his various designs, except this packs bass emphasis and gain into one bidirectional control.  Take out the diodes to increase signal level and stick in a BMP tone control and you have a lot of flexibility.

[LucifersTrip]

One way to make it less boomy is to simply reduce the value of the 33uf cap so that the gain is applied less to the low end.  On strategy I've tried in past with some success is to tie the wiper of a pot (and 5k would be just fine here) to the ground side of the transistor, whether bipolar or FET, and run two different values of cap to ground, one from each outside lug of the pot.  At midpoint, you get no additioal gain.  Rotate in either direction you get two different kinds of gain.  Towards the 33uf side you get gain applied across the full spectrum.  Towards the other side you get more gain applied to mids and highs, depending on the value chosen (consider 3u3 to start).  I gather the net effect is similar to the variable bass bypass that Joe Gagan so elegantly incorporated into his various designs, except this packs bass emphasis and gain into one bidirectional control.  Take out the diodes to increase signal level and stick in a BMP tone control and you have a lot of flexibility.

yes, definitely numerous ways that you can achieve similar goals on this one....I wonder if reducing the 33uF would be good enough so no altering of the vol pot/cutoff would be necessary.  Though, I would guess there would be more change to the original tone than changing the vol pot/cutoff?

funny, I was just editing my above post to add a quote from you regarding the "distortion" control...which is one of the other reasons I did it differently:

"I wired up a Superfuzz-style midscoop filter, in place of the 250k+27k/.001uf network, and while it wasn't awful, in retrospect, I don't think it necessarily suits this thing."

[Mark Hammer]

Well maybe this thing is what you want:  http://www.diystompboxes.com/smfforum/index.php?topic=95425.0

[Earthscum]

Dude, awesome... I've been eyeballing this circuit for a couple months now, just never plug it together. My cross-referencing led me to some BC transistors that should work in place of the 2SC1312G (which I will be pulling from a recycle board, BC547C, or BC337-040) and 2n5457's seem to be a near perfect spec match to the 2SK34C. The 5458's really aren't too far off, IMO. If I get one plugged together, I'll pop on here and let y'all know how it went with the BC's.

How's this as far as letting bass through? Is it nice and thick, try as is, or should I expect it to be a bit thin and maybe up some HPF'ing cap values? It's hard for me to tell from the clips I've came across on the 'net.

[LucifersTrip]

Dude, awesome... I've been eyeballing this circuit for a couple months now, just never plug it together. My cross-referencing led me to some BC transistors that should work in place of the 2SC1312G (which I will be pulling from a recycle board, BC547C, or BC337-040) and 2n5457's seem to be a near perfect spec match to the 2SK34C. The 5458's really aren't too far off, IMO. If I get one plugged together, I'll pop on here and let y'all know how it went with the BC's.

definitely try all the jfets you have...there was definitely a variety in sound. The 2n5457's were the worst of the ones I tried.

How's this as far as letting bass through? Is it nice and thick, try as is, or should I expect it to be a bit thin and maybe up some HPF'ing cap values? It's hard for me to tell from the clips I've came across on the 'net.

you don't have to worry about bass. with the "distortion" full counterclockwise there's probably too much bass, which is why I increased the 27K cutoff. If it's not enough for you, just do the opposite and reduce that 27K

lookin forward to your build

[Mark Hammer]

definitely try all the jfets you have...there was definitely a variety in sound. The 2n5457's were the worst of the ones I tried.

That's exactly what I have in there!

[Earthscum]

It's late tonight, but I got it on the BB finally. If I could get it to stop oscillating, it appears that it will most definitely put out a square wave.

I used all original values, with the exception of the 2M resistor, which I subbed 2.2M for, and the 33uF cap, which I used 47uF. I have BC 547 and 2n5457 on board right now. It's late, so I'll screw around after burgersNBeers monday at Lucky Joe's. I tried 5089, lessened the oscillation, meh... tried MPF102, and the osc stopped... and I had mud. I pilled the diodes (4148's) and the osc got super crazy.

I do believe this pedal does squares. I think there's a combo to be found, though, for modern (disappearing modern) equivalents. I'm almost thinking of swapping out the 3.9k with a 10k pot and a parallel biasing resistor, and taking the 47uF off the lug for a variable gain.

Quickie voltages show:
Q1
E 0.011V
B 5.1V
C 0.47V

Q2
D 4.7V
S 1.66V
G (0.47V)

And, on a last note before I hit "Enter", I just decided to drop a cap across the diodes... .1u Ceramic got rid of the oscillation. I'll get back tomorrow night and let y'all know a bit more. I'm going to bed. It's midnight here, lol.

[LucifersTrip]

I'd have to guess that you have something wired wrong since you're so far off what I got on Q1 C.  It put out such a low vol in stock formation, I would've found it hard to believe it'd oscillate...but of course, oscillation isn't all about volume.

check the wiring. I doubt Q1 C should be .5 (Did you see my voltages on the schematic?)

good luck

[Earthscum]

yeah, I tried a 5089 (not 88) and had voltages a bit closer to what you posted. After doing a couple FF's, I'm not surprised by the voltages. Remember that the Fet's gate is supposed to be lower than the Source. I ran across something like this before when I was playing around with different transistors in Q2, ala "Many Faces of Fuzz".

I'll get on it again tonight though... piqued my interest, lol.  I am absolutely sure I have everything plugged in correctly. The BC's pinout threw me off a bit, initially. I took awhile looking things up and triple-checking everything after I turned it on and heard a nice steady whistle coming from my headphones, lol.

[Earthscum]

Well, I'm done with this one. I've yanked everything, plugged it all in again, measured resistors and tried a 3rd time. Still oscillates and sounds fantastic, or sounds like utter crap if it isn't into the oscillation territory. I'm also getting approximately the same voltages as I noted above, every time, with the 5457 fet and ANY transistor in Q1. So, I can only assume I keep doing something wrong and I'm shelving it. 

[LucifersTrip]

Well, I'm done with this one. I've yanked everything, plugged it all in again, measured resistors and tried a 3rd time. Still oscillates and sounds fantastic, or sounds like utter crap if it isn't into the oscillation territory. I'm also getting approximately the same voltages as I noted above, every time, with the 5457 fet and ANY transistor in Q1. So, I can only assume I keep doing something wrong and I'm shelving it.  

That's funny...I built another of my V1 mod just to see if it would come out the same, or if I could cause it to oscillate (I couldn't)....and it did come out almost identical.  The only difference was that the sweet spots on the E/S resistors were 1.75K & 3.1K respectively. All voltages readings were exactly as noted in the V1 schematic, except Q2 D was 3.88v instead of 3.77v.

I cannot stress enough tweaking the E/S resistors. I tried the 5457 again and it was clicky and mildly gated, but a little adjustment of those resistors put it right in focus, though still not as sharp or powerful as with the J201.

I'd try again, especially since it has a bit of uniqueness to it...



Edit:

I'd love to know how you got 5v on Q1 base. I tried and couldn't. Even if I shorted the 560K on the collector it only hit 2.59v

[Earthscum]

Edit:

I'd love to know how you got 5v on Q1 base. I tried and couldn't. Even if I shorted the 560K on the collector it only hit 2.59v

HAH! No wonder... that was supposed to be .51V. I was looking at your voltages thinking "Uh, mine really aren't that far off".

So, before I ripped it out I tried swapping out resistors. Yeah, sound changes. Just for the hell of it I used E:(180R) and 1.8k, C:180k. With the 1.8k it biased up closer to the voltages you posted, and no oscillation. BUT, fiddling with stuff revealed the same thing... either super sweet and unstable sounding, or muddy and blech. The thing sounds fine at higher gain settings, to me, but I think it would be a sound better chased through at least one more stage of boost rather than cramming all the gain into few components.

BTW, I did figure out a neat little "tone" control trick with this. I used a 10k at the end, with a 20k pot. At the junction of the 2 I dropped a .01uF cap to ground, LP T-filter, kinda. It gave a whole new dimension to the "distortion" control. It also got a bit muddier(I did it to quell the osc), but I think it's worth playing around with if you managed to get a tone you liked. With the 500k/100k in your third, it could add some interesting control with a value as small as 470p or 560p  

I about threw the breadboard out the front door... I hate whistling and jingling... and wet socks. It just kept hitting a tone that about drove me mad, so I have a slightly unfounded dislike of this circuit now. I should've tried a BC in Q1, it at least was toned down enough that I didn't want to jam a pencil through my temple, lol.