Is pseudo-random switching possible?

[disantlor]

My friend has an old MXR Digital Delay rack and it has an external hold input that just repeats whatever is in its buffer as long as that input is shorted.  We thought itd be cool to make a little device that randomly closes and opens the switch while you play/sing.  I'm not sure how to generate such a signal, though a little searching seems to indicate that you can do it with white noise somehow.  Only idea I've come up with so far is to take 2 or 3 slow oscillators with different frequencies constructively/destructively summing at the gate of a FET switch.  Seems like it might give the impression of randomness.  What do you think?

[R.G.]

As with so many other questions, read GEO - in this case "Pseudo Random LFO" for some PR approaches.

In today's world, the best PR source is probably a 12F508 PIC microcontroller, programmed to put out a PR sequence that would repeat every few years or so. About $1.50 for the chip.

[RaceDriver205]

You know anything about digital circuits, disantlor? Ive got an switching idea that uses discrete logic ICs you can get from standard electronic stores.
It would use a decade counter, an OR gate, a timer and a flip-flop etc. You ever seen an electronic dice circuit? Very simple circuit to get a random number.
Interested?

[The Tone God]

If you are looking for a binary result, meaning on or off, and the need for near true randomness is not that high then a uC would be overkill IMHO. Granted it maybe be one of the better options in terms of near true random output but in this case I don't think it is that important. Besides while the cost of the uC maybe be $1.50 you would also need to get a programmer and either find or write the software for it. Its more then just the $1.50 uC for someone who is not prepared and does not know how to use a uC.

I think the easiest/cheapest ways to do this is to modify one of the many "coin tosser" circuits out there which is similar to the "dice" circuits just with a different number of outputs. Most work by using a high speed clock stepping the output and is enabled/gated by a secondary source like a button or another clock. Hmm...maybe you can use two clocks, one high speed and one the gate speed, feed the high speed into the data side of the flip flop and feed the flip flop clock input with the gate clock. Two standard ICs. Simple enough and might even be cheaper then a uC.

Andrew

[RaceDriver205]

Yeah, pretty much the idea I was thinking of. Damn straight about the programming, my JDM prog still doesn't work.