Univox Square Wave

[Mark Hammer]

I threw one of these together some time ago and was not all that thrilled with it.  But, the other night I was going through things that were in an almost-done state and came across it.  The chassis was prepared for it and legended, so I figured I'd see if I could turn it into something decent.  You'll find the schematic here: http://www.univox.org/pics/schematics/squarewave.jpg

Stare at it long enough, and you'll begin to see that it is essentially a Fuzz Face with a JFET in the 2nd transistor position, and a diode clipping pair after to add more sizzle.  You'll also see that it has a fixed distortion setting and merely allows for tone adjustment and output level.

The tonal adjustment is essentially a bass trim control, so that the tone goes from full bandwidth to thin and nasal.  Different for its time, I'm sure, but not inspiring.

I did a few things to it in hopes of improving it.  One was to use silicon diodes for the diode pair.  A second was to stick a 5k pot between the 33uf cap and ground to adjust gain. I imagine a 2k is probably better, but 5k was not any annoyance.  This gave an adjustment from a pleasant grit right up to full bore squarishness.

I stuck a 2k7 resistor between the diodes and the 4u7 cap before them, and a .01uf cap to ground in parallel with the diodes, just to tame the fizz a bit.

The tone control had to go.  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.  But, I added another .01uf cap to ground, just before the volume pot, which improved matters.  Skip the midscoop, stick a 4k7k resistor in place of the existing tone control network, and a .01uf cap to ground where that resistor meets the volume pot, and I think you'll be pleased.  A nice "HEY!!  I'm talkin' to YOU!!" kind of tone.

The result is a decent-sounding fuzz with some tonal variation and the usual sort of cleans-up-as-you-turn-down performance.  Not as loud as a Fuzz Face, but I think that can be improved by increasing the value of the volume pot.  All in all, much better than the original.

[Mark Hammer]

A little more tinkering last night.  From the diodes, it goes to a 5k6 fixed resistor, and then to the stock 50k volume pot.  With a 2k7 and 5k6 resistor leading up to the input of the volume, the output is substantial.  Much louder than my AMZ MosFace fuzz.  To tame the top end, I placed a 6800pf cap to ground from the junction of the 5k6 resistor and volume pot.  probably a 4700pf is better.  But VERY meaty and authoritative-sounding.  Lots of good solid bass that lets you know who's boss.  I also installed a switch that lifts the ground side of the diode pair, so that you have a choice of basic transistor overdrive (which is warm and familiar), and "extra-clipped" sound.  Without the diodes, this sucker is VERY loud.

I used to hate it, but now I like it.

[John Lyons]

Good stuff Mark.
I like when you can turn around a circuit that is either ho hum or not that great and make it into something that works for you.
Example: Funny how many people pass off the fuzz face when it can have so many sounds and sound unrecognizable compared to a typical fuzz face. Just change a couple parts and it's a different animal altogether.

[Radamus]

Any chance of a schematic of your changes? You make it sound so good.

[Mark Hammer]

Here you go.  If you build it, tell me what you think.

[Radamus]

Thanks for the quick reply. The parts are going in my order, can't say when I'll get to this one, but I'll let you know.

Thanks again.

[Radamus]

Hello,

It took a me a while to get the parts for this pedal, but I put one together over the last couple days. I plugged it in with a battery at around 8.8v. I was pleased initially because my LED worked (first millennium), but when I turned my amp on, I got no sound when the effect was engaged.

Voltages:
2n5088
C: 2.3
B: .71
E: .68

2n5458
D:3.00
G:2.23
S:2.23

I find the 5458 voltages a little odd, but when I audio probed this unit, I lose sound immediately after the base of the 2n5088. The collector is silent, except for some intermittent fuzz. Mostly silent.

I checked my work and it looks like everything's soldered correctly. Doesn't mean that it is (and it probably isn't), but do you have any idea where I should start looking? By the way, both pots are maxed. I think the clipping switch was off.

Any ideas?

[soggybag]

Reminds me of some of the Tim Escobedo fuzzes. Like the Jinx and Punch in the face.

[Radamus]

I'm really at a loss here. I've been tinkering with it for a couple hours. I can't figure out why that damn collector is completely silent. I think I'm done for the day, but I hate leaving these things undone.

[Mark Hammer]

The only help I can offer at this point is to simply repeat the mantra: "Check your FET pinouts"

[Radamus]

You have the numbers listed, rather than the names. Is 1 the Drain? I socketed both transistors and I tried them both in different directions.

Wait, no. I think I realized the problem. Apparently on the 5458 it goes D-S-G. Thanks for that catch. I thought the worst case would be that the transistor was backwards, not completely rearranged. Well, there's my initiation into FETs.

I'll give that a shot, but that will probably do it. Thanks.

[Radamus]

Okay. I fixed the pinout of the 5458. I have the same exact problem still. The voltages on the 5088 seem just fine (c2.28, b.71, e.67) but it's not amplifying.

5458
D 3.00
S 2.28
G 2.29

The collector of the 5088 isn't grounded anywhere, or there wouldn't be voltage on the gate or the collector. There should be some sound, right?

I don't know if I know how to diagnose this myself, so if there's anything I can provide that will help, let me know.

Thanks

[Radamus]

Hello,

It's been a couple days and I've been a little bit afraid to work on this pedal. I just audio probed it again and I have the same exact problem. The 5088 transistor has no sound at the collector. the 5458 is in the right way now, as it was for my last post. The voltages are the same too. I've stared at the underside looking for solder bridges for hours and no luck. I've tried measuring resistances as much as possible, but have had little luck actually getting an answer from my DMM.

Is there any way you can point me in the right direction here? Can I answer any more questions? I've done as much as I know how to do and I've provided all the information I know to provide.

Please help me out.

Edit: I replaced the 2M resistor in the circuit with a 2.2M. The place I ordered from didn't have any 2M. I'll try putting two 1M resistors in series and see if that helps any.

Edit 2: I lied before. The voltages are not the same. New voltages:
5088
C 3.14
B .78
E .59

5458
D 2.62
G 3.14
S 2.51

I think those voltages look a little better than they did before, but still no sound at C (or G).

[Radamus]

I read the articles over on Geofex again to refresh my memory about debugging. What I found is that my transistor voltages on that first guy were not different enough from B to E to allow conduction. I put the two 1M resistors in series to see if that would fix anything. It didn't. I tried a number of different values in place of the 680Ω resistor. I tried a lot of low value resistors in series with the 680 and I got none of them to work properly. Then I grounded E and got loads of sound. The thing is a monster. It's a really thick fuzz, maybe a little too thick for my taste, but it's really powerful nonetheless. It's not what I would expect a square wave to sound like, not harsh enough, but definitely a cool fuzz. I'm playing it on bass and it gets pretty muddy at the lower end of the E string, but it can get some pretty great sounds higher up, so I guess it's much better for guitar.

Just to be safe, the 5088 is now at c~.9, B=.47, e=0. Is that going to cause me any problems? If so, how should I go about biasing this thing a little better?

I'm not sure where our two builds differ that I needed to change the emitter resistor. Any thoughts?

Anyway, thanks for experimenting with this. It's a cool project, and it'll probably get better when I put it in a box.

[SpencerPedals]

Any reason why a J201 wouldn't work for this?  I have very few FETS since I have stayed away from them, but I'd like to breadboard this.  I checked the data sheets and they're both n-channel so I assume so, but I don't know a damn thing about FETS, quite honestly.

[Mark Hammer]

It might work ok.  Give it a shot.  Just be careful to verify pinouts.

[snk]

A second was to stick a 5k pot between the 33uf cap and ground to adjust gain. I imagine a 2k is probably better, but 5k was not any annoyance.  This gave an adjustment from a pleasant grit right up to full bore squarishness.

Hello,
I'm resurrecting this old thread : I am building the Univox SquareWave, and I am willing to feature this mod.

Am I right assuming that the lug would b wired like that?
- lug 1 to ground
- lug 2 to lug 3
- lug 3 to negative side of the 33µF cap

[antonis]

I'm resurrecting this old thread : I am building the Univox SquareWave, and I am willing to feature this mod.
Am I right assuming that the lug would b wired like that?
- lug 1 to ground
- lug 2 to lug 3
- lug 3 to negative side of the 33µF cap

Yes..!! 

Pot is wired as variable resistor providing FET variable gain and BJT variable negative feedback..

[Mark Hammer]

Note that, since the original used fixed gain, much like the EHX Muff Fuzz, that also used fixed gain, the diode pair is to provide some clipping over and above what the modest gain provides.  The variable gain mod will extract a lot more sizzle that doesn't require diodes to produce it.  But be forewarned: without the diodes, or with them lifted, this thing is LOUD - much louder than with them.  That's not bad or good, just something to anticipate.  Putting diode-lift on a footswitch would likely get you thrown out of the band or club!

[snk]

Thank you, Antonis and Mark!