High Freq Switching as an Option for Mixing On-Board Effect Settings

[Jdansti]

This is an academic question at this time. I've been thinking about the use of high frequency switching (HFS) to blend two or more effect settings within a given effect, potentially producing a different (maybe better) effect/sound.

For example, say I've got a fuzz/distortion that has two different sets of clipping diodes, and each set produces a very different sounding effect. So on my hypothetical pedal, I could have a manual switch to select between the two sets of clippers.  I could also use a small on-board mixer to blend the outputs from the two clipping circuits (does anyone know if this has been done?).  A third option would be to mix the two signals using some sort of HFS.  Has does anyone know of effects using this?

My hypothetical pedal is a fuzz, but maybe HFS would sound good on other types of effects.

My initial questions:

1) Would mixing or HFS even be desirable?

2) If so, would HFS result in a different sound than a mixing circuit?

3) In the case of diode clipping, would it be necessary to switch between two complete amplification/clipping circuits, or just between sets of diodes?

4) What range of frequencies should the switching be?  > audio?

5) Would a HFS box be a good idea to blend two or more pedals?

6) Should I avoid making silly posts at 4AM?

Any thoughts would be appreciated.

[defaced]

Sixth example down. Seems to me you can swap the resistors and diodes to allow amplitude or crossovers distortion configurations, and mixes thereof. http://www.muzique.com/lab/sat2.htm

[samhay]

Possibly a very cool idea. I would imagine that for this to be musical, the switching frequency would have to be lower than the frequency of the note you are clipping, so you are in familiar LFO territory.
I started a post a few days ago about panning between two sources using a LFO or similar. The same approaches that were suggested could/should be useful here if you use a square wave LFO.

Edit - Oh, but then I guess it would not be HIGH freq switching.

[Jdansti]

Thanks Mike. I'll definitely have to try that!  Looks pretty simple.

I'm still interested if anyone knows how blending with a potentiometer would compare with high frequency switching.

Edit:
Sam-sorry I posted right after you did and didn't see yours. Good point about the frequency I'm not sure whether the switching freq would have to be above or below the frequency of the note that is clipped.

If we took a pure sine wave and switched at a frequency that is higher than the frequency of the wave, then the wave would change shape before it completed a single cycle. The converse would be true if we switched at a frequency that is lower than the frequency of the sine wave. You could potentially have multiple cycles go through one set of clippers before the signal would be switched to the other set of clippers.

[artifus]

sounds like a job for fuzzy logic - cmos style.  http://en.wikipedia.org/wiki/Logic_gate

[TOPLEL]

You mean a high speed square LFO clocking a switching IC? (i've forgot its name but i've seen it here frequently, i think last time in the arr!fx arpeggiator)
I think the clocking frequency should be twice of our hearing range (you only hear each sound half the time)

So when the clock square goes 1 the IC let's through one sound and @ 0 it lets through the other?

Maybe with PWM (on the clock signal) the mixing balance could be changed (perfect 50%/50% square would make the 50-50 mix, 30%-70% PWM "modified"  square could make a 30-70 mix etc. Maybe.)

[Jdansti]

Sorry for my ambiguous language!  I'm new to this area of electronics.  I'll try to describe my idea in a different way.

Imagine a fuzz (as an example) that has two sets of clipping diodes you can choose via a switch. Let's say that when the switch is in the up position (A), you get Ge diodes, and in the down position (B) you get LEDs (this is just an example, so don't think too much about the difference between Ge and LED clipping).

So I get a certain sound in A mode and another in B mode. Now I want to combine the A and B modes to produce a C mode/sound. One way would be to blend the A and B modes with a pot, and that might be a desirable thing to do.  I'm wondering what happens if I quickly and constantly switch back and forth between A and B.  I'll call that combination C'. I have a "stompbox monkey" that will quickly flip the A/B section switch back and forth while I play (he's a fast little monkey).  How does the sound change?

So now I have a C' mode that should sound different than A, B, or C. So I don't get reported to the SPCA, I replace the monkey with a circuit to do the A/B switching for him. The new switching circuit frequency is adjustable via a pot and it doesn't fling poop. Assuming that the switching technique would produce a different and desirable sound, what do you all think would be the range of useful switching frequency (i.e., close to guitar frequencies, above, below, or all of the above)?

If this seems worth doing, what type of circuit would emulate my monkey flipping the switch?

I hope this makes sense.

[amptramp]

The switching frequency has to be above by at least a factor of two and preferably by more than a decade above the highest audio frequency you want to produce.  After this is done, you need a lowpass filter to eliminate switching hash.

The Paul Nelson PWM phaser works this way.  Instead of having to match FET's to get the phaser to work properly, CD4066 switches are driven by a PWM signal derived from the LFO.  Switching is at 40 KHz.  The schematic is here:

http://www.diystompboxes.com/smfforum/index.php?topic=76747.0

There is a similar one here:

http://www.uni-bonn.de/~uzs159/phaser.html

If you have triangle wave PWM signals, you will get linear interpolation and not logarithmic (db) mixing, but this may be best anyway.

[samhay]

If the 'monkey' is switching a lot faster - say an order of magnitude or more - than audio frequencies, then I'm pretty  sure the end result will sound very similar to if you just mixed the two signals with e.g. a pot (assuming you filter the hash out).
It might get interesting if you switch at a similar frequency as the note currently being played, but that will not be without its challenges.

[Jdansti]

Thanks for the info and links, Ron.

Sam-I was also thinking that certain frequencies might be similar to using a blending pot, but other frequencies might produce some nice harmonics.

One thing that high freq switching can do is alter the duty cycle for each circuit being switched. Probably the only way to know for sure how this would really sound would be to experiment and see what happens.  Before I go all out and build a big switching circuit, I'll probably breadboard a 555 timer and a solid state relay chip just to see if its worth going further.

Thanks for the input, guys!