Window comparator "crossover" distortion idea

[earthtonesaudio]

You could use a window comparator to detect when your guitar goes "near zero" and use the comparator's output to attenuate the audio path. Depending on the size of the window and how much you attenuate, many different shades of crossover distortion could be obtained.

Probably would be a good idea to buffer/split the input signal so the attenuated signal doesn't feed back into the window comparator... or don't and see what madness ensues!

[PRR]

All the time I have spent reducing crossover, and you WANT it?

Anyway it takes an exceptional player to make crossover "musical".

> use the comparator's output to attenuate the audio

This "feature" is a natural flaw. Your way seems like the hard way. Maybe you need to get that fancy to get good control; maybe not. Whomp this together and try:



Note that the output must approach a Volt. You probably need some attenuation after this, if you are going into a guitar amp.

Yes, the output gets smaller as crossover is increased. You are slicing-out the middle, there's less left. You may want to "compensate" that; it won't be easy.

[PRR]

Reaching way-back in old theory:

What you want is a "Dead Band" circuit. This is older than electronics. A steam engine has a governor to control speed. Say 100RPM. Slop in the linkage means the governor may not act between 95RMP and 105RPM. This may be useful: your home thermostat comes on at 65 and off at 68, but is no-action between those points because the fire is on/off and you don't want to be cycling every few seconds.

I bet there's a Burr Brown paper on opamp dead-band design but today it may only exist as a brownish scan.

[earthtonesaudio]

Okay, built this with TL074 and a JFET.  Gain of 10 input stage drives window comp and second audio stage wired as a high gain inverting amplifier with clipping diodes.  In between the two stages is a J113 wired as a switch.  The window comp is wired so the output goes low as the signal passes through the window, this signal drives the JFET through a 500k pot with a 3n3 cap from wiper to GND.  This slows down the signal going to the JFET by a variable amount.  The window voltages are set by two 1M resistors with a 100k pot between them, so the window is also variable.

Across the JFET's source and drain terminals, I've been placing random components for fun... a couple antiparallel diodes there makes some interesting distortion, but a smallish cap makes a highpass filter when the JFET is off, so certain settings give trebly clean attack followed by fuzz, other settings are Slow Gear-ish.  Some settings are incredibly sensitive to input dynamics.

Schematic soon, I'm up past my bedtime 'cause I've been having so much fun.

Your way seems like the hard way. Maybe you need to get that fancy to get good control

I think this sums it up nicely.

[Gus]

Look for the boss HM2 schematic note the ge diodes.  Then there is the Machine.

http://www.effectsdatabase.com/model/boss/compact/hm2

[earthtonesaudio]

[earthtonesaudio]

Corrected some of the connections at the window comparator, cleaned up the schem a bit and re-sized:

[The Tone God]

I've played quite a bit with crossover distortion so here is one idea:

http://www.thetonegod.com/tech/blade/blade.html

And more crossover madness:

http://www.thetonegod.com/tech/finish_line/finish_line.html

Andrew

[Gurner]

Perhaps i'm missing what you're trying to achieve, but your running this on a single supply rail, yet rather than biasing to half the supply, your opamps (IC1C &IC1D)  are referenced to the signal ground?

[earthtonesaudio]

Perhaps i'm missing what you're trying to achieve, but your running this on a single supply rail, yet rather than biasing to half the supply, your opamps (IC1C &IC1D)  are referenced to the signal ground?

I just used the AGND symbol to indicate the 1/2 supply voltage reference (4.5VDC).  Sorry for the confusion.

I've played quite a bit with crossover distortion so here is one idea:

http://www.thetonegod.com/tech/blade/blade.html

And more crossover madness:

http://www.thetonegod.com/tech/finish_line/finish_line.html

Andrew

Thanks for sharing those.  I really like your "Blade" because it's like a crossover distortion laboratory, in a way.
I think the effect I've drawn up here counts as "crossover distortion" because that's where the action of it occurs, but with C6 in there it becomes something slightly different.  I also tried putting a pair of antiparallel diodes in its place, and that gave a slightly different sound.  With the diodes in there, it's like having one crossover distortion within another crossover distortion... weird.

[Gurner]

Perhaps i'm missing what you're trying to achieve, but your running this on a single supply rail, yet rather than biasing to half the supply, your opamps (IC1C &IC1D)  are referenced to the signal ground?

I just used the AGND symbol to indicate the 1/2 supply voltage reference (4.5VDC).  Sorry for the confusion.

Aaah...now theat you say that, I can see you've used two different gnd symbols - Since you use Eagle, if you take a look under their library 'supply1', they provide a symbol for half supply - it's called VCC/2.

Cheers.

[earthtonesaudio]

Aaah...now theat you say that, I can see you've used two different gnd symbols - Since you use Eagle, if you take a look under their library 'supply1', they provide a symbol for half supply - it's called VCC/2.
Cheers.

Haha, I actually considered using that symbol, but decided against it because the AGND symbol is more like the symbol I use when I draw schems by hand.  



Here's a clip of some of the sounds this can make.