Saw (was: square) wave FUZZ box ?

[A.S.P.]

why does everybody make square-wave fuzzes?
wouldn`t a saw-wave offer more sound options?
are there any out there?

[toneman]

Hey ASP,
Show me how U'd do it 
I'd love 2 know how   
Synths don't count, BTW   
Tone

[Pete-Galati]

This collection of Univox schematics that I stumbled on has a fuzz called "Squarewave"

http://www.univox.org/schematics.html

Don't know what you'd get if you built it, but I don't imagine that technically it'd be square waves.

I've got an old Syquential Circuits synth tucked into the closet, and I imagine that would get you closer to a square wave with that.  But squarewaves are not a subject I know much about.

Pete

[A.S.P.]

hey tone, you`re trying to read between the lines... 
that doesn`t count either.

edited the topic title & initial post...

[gez]

wouldn`t a saw-wave offer more sound options?
are there any out there?

I did a true saw-wave octave up once (ramps up then plummets to earth).  OK, but I prefer the sound of square wave (saw was a little harsh sounding).

[jmusser]

Tim Escobedo's Simple Squarewave Shaper has pots on there, that adjusts the wave form from square to trapozoid, to kind of a sawtooth type wave. In sawtooth mode, it makes the fuzz sound sort of like a trombone.

[R.G.]

There aren't saw or triangle waveform fuzzes because there's no straightforward way to do it in analog circuits.

A square wave is easy - whatever you have, you just amplify it and clip it. If you amplify and clip enough, you get a square wave.  A saw wave requires a verical side, easy enough, but then the sloped side has to slope just enough to reach the next vertical side in one waveform period, and has to do that for all audio signals. There's no dirt-cheap simple way to make the slope adjust so that happens. If you have a constant slope like simple circuits do, the amplitude of the waveform decreases linearly with frequency.

There's stuff you can do for that. You could try a frequency sensitive slope with a frequency sensing circuit and adjustable integration. You could do a constant slope then use a compressor to restore the amplitude to relatively level with frequency. These all add considerable complexity.

There are some generalities that underly stompboxes, generalities that have to do with available technologies and commercially practical means. That's really the line of thinking that started me down the road that resulted in the "Wrappers" artical at GEO.

Underlying it all is economics. No one will build a stompbox for the mass market that cannot be built and sold for a profit. That means that you have to sell a few for huge prices (that's the Howard Dumble/Cornish model) or a zillion for cheap (the Boss/Ibanez/Dano/everybody else model). Guitarists in general won't buy stompboxes that are too onerous to haul into a small bar gig. They won't buy one that has strange power requirements, usually a 9V battery or a wall wart. 120Vac power supply pedals exist, but they're rare. There aren't any powered by D cells, or car batteries. The pedal has to have a sturdy enough enclosure, has to have a few controls that are quick to adjust on stage, and many others. These all together form a shape that defines stompbox in our minds.

Taking all that and the economics of electronics together, you come up with a rough set of guidelines. Step over the line to too much circuitry required to do the processing and it won't fit in the box or won't be affordable. The astonishing rise of cheap digital processing power was a step sideways out of the analog strictures and that alone was responsible for the digital multi-effects. It changed the rules about what was cheap.

[H S]

There are things you can do with a phase locked loop to get a saw, but it's the kind of circuit that has to "track" your input and it only works on one note at a time.

[A.S.P.]

thank you, R.G., for that in deep explanation;
that was what I was afraid to hear;
mr. Toneman: of course my idea was inspired by mighty
brassy/stringy synth-sounds, that are nice basics
to be filtered, and offer a different sound than the usual "fuss".

[Mark Hammer]

One of these days (right after I finish those 300 odd projects hanging around from 10-20 years ago), I'm going to have to reverse engineer my old Korg PME Waveshaper.  It is a fuzz of sorts but has a continuously variable waveshape control that goes between saw and square.  I haven't looked at it closely in a while, but there are a bunch of SIP chips in there (a lot actually), with a 13600 or two among them.  As boards go, this one is sorta like a DIM-C in terms of how much stuff is packed in there for so few controls.  Could be a relevant addition to this thread, though.

[Nasse]

I found one old lost circuit last week or something, about the time there was talk about "wave folding" or something. It was just a crude two transistor fuzz, but with a waveform adjustment pot. Something like out of phase signal taken somewhere, mixed on squared signal trough a cap filtering lows. And there were some waveform pics and it did go from rounded square to something that looked more like a (badly) distorted saw wave than square. I´ll try to post it some day perhaps. And maybe Joe Gagan and Z Vex have done something like this, of course this is not what was asked but which fuzzes have any kind of control for waveform dunno me not

[gez]

This is a cheap and cheerful way of doing it…well, sort of (only the octave up bit is saw tooth).  This is the ‘Mark I’ and is a little messy, what I ended up with is a lot more efficient (sorry, no schematic for that one folks) but the basic principal is the same.  I’d do things very differently now, but then that’s a good sign I suppose…



The amplitude problem RG mentioned was got round by using passive integrators.  With lower frequencies the waveform gets compressed due to the longer time constants.  This does result in distortion of the waveform, but like I said it’s cheap and cheerful.  There is a slight taper in amplitude as frequency increases though (done deliberately as this sounds more natural to my ears).

Disclaimer: This version was only bread boarded and the schematic has been residing in my ‘Works In Progress’ file for a while now so I can’t guarantee values are correct etc.  More circuitry would be needed to get something more professional sounding, but this is stompboxable.  Breadboard first!

Thank you Mr Keen for the mixer idea.

[JimRayden]

I scoped the ROG Whisker Biscuit and by twiddling the knobs, I'd go from normal distortion to sawtooth to a square wave lookalike. If you're not too picky about your shape purity, I think it'll do fine.

-----------
Jimbo

[Tim Escobedo]

If you've ever heard raw saw waves from a old analog synth, you may notice thay can be very harsh indeed. I guess thay have some technical advantage by exhibiting odd and even harmonics compared to the odd-only harmonic spectra of a 50% duty cycle square wave. Unfiltered, I would evaluate the square wave as probably "smoother" sounding. IMO, of course.

Another big stumbling block, as mentioned, is the difficulty of shaping a guitar signal into a nice saw wave in a manner that does not involve digital modelling. The Simple Square Wave Shaper is a hack of a simple circuit offering some saw wave shaping. Dodgy on/off dynamics and no filtering, the result I describe as "brassy". In the lower registers, it makes me think of a tuba. Through a distortion box. 

With a tracking filter and VCA to restore dynamics, it might get really interesting. By that time, you can pretty much call it a synthesizer.

[Paul Perry (Frostwave)]

Given a genuine square wave (not easy, if you are starting from a guitar) it should be simple to make a saw, just by turning on an integrator when the square wave goes positive, and discharging it quickly & starting over on the next positive edge.
Now, the amplitude of the triangle will vary with the frequency of the square wave.... but at least you will have a saw.

[H S]

Given a genuine square wave (not easy, if you are starting from a guitar) it should be simple to make a saw, just by turning on an integrator when the square wave goes positive, and discharging it quickly & starting over on the next positive edge.
Now, the amplitude of the triangle will vary with the frequency of the square wave.... but at least you will have a saw.

If that square is swinging from -1 to +1, wouldn't you get a triangle?  If you reset at every zero crossing, positive-going or negative-going, and then full-rectified, then you'd have a saw, up an octave. 

[H S]

Given a genuine square wave (not easy, if you are starting from a guitar) it should be simple to make a saw, just by turning on an integrator when the square wave goes positive, and discharging it quickly & starting over on the next positive edge.
Now, the amplitude of the triangle will vary with the frequency of the square wave.... but at least you will have a saw.

If that square is swinging from -1 to +1, wouldn't you get a triangle?  If you reset at every zero crossing, positive-going or negative-going, and then full-rectified, then you'd have a saw, up an octave. 

Oh, hey!  Then subtract that octave-up saw from the square and get a same-octave saw!

[gez]

Given a genuine square wave (not easy, if you are starting from a guitar) it should be simple to make a saw, just by turning on an integrator when the square wave goes positive, and discharging it quickly & starting over on the next positive edge.

If you discharge it when the square goes negative then you don't end up with a saw tooth - you get tooth, space, tooth, space etc.  You need two integrators - one of which is switched from an inversion of the square so you get space, tooth, space etc - along the lines of what I did above, then both sides get 'zipped' together with a mixer. 

In order to maintain original pitch you'd need to use a flip flop to divide down the original.  Then you'd probably need a fundamental extractor as I'll wager you'll get some odd stuff happening on the first few frets of the sixth string (this is why I kept to octave up).  The parts count starts to increase...

[A.S.P.]

spaced saw - even more bite! (or: harsher...).

guys you scare me - maybe I should withdraw my quest?

[puretube]

Mark: talking `bout K*rg: " clean" down

maybe fits better in the other thread...