Random interval switch

Sure, as long as size and power does not matter.

Good luck

hamilton

Maybe you can adapt one of these:

h

Yes, with an FPGA.

For random events the 'most likely' time between two of them is zero.

(It seems a bit wierd)

George H.

Is that because random events (without defined boundaries) occur at an unbounded (infinite) frequency?

If that is not the reason, I would appreciate an education concerning this "wierd" aspect of probability.

Thanks, john

Given your criteria below, the "best" approach *is* a cheap MCU.

Presumably, you mean "random" in the colloquial sense and not strictly literally?

When you claim "2-20/hour", do you mean *that* literally?

For example, if you set it to "2" and start a stopwatch the instant it "rings" the FIRST time... and, BY CHANCE, the *second* time happens to be exactly 3 minutes later (satisfying your "no closer than

3 minutes" criteria), can the observer rest assured that it will *not* ring again for at least 57 minutes?

(i.e., what is the shape of your probability distribution)

Sure! Micro is just the cheapest, smallest, easiest, lowest power way to do it!

But, hey, you can always be costlier, larger, harder or hotter ;-)

Of course. Just a string of shift registers arranged in a PSRG (pseudo- random sequence generator) and a bunch of random "glue" logic (gates, data selectors, magnitude comparators, PROMs, etc.) to get the effect you want.

Sorry I have no clue how to make a specific circuit. (well, I could come up with one, but certainly not for free!) ;-)

And, of course, you've already disallowed the "best design approach" by saying, "without a micro."

Good Luck! Rich

And, if you enforce that rule over a longer period, what should that period be?

'They' do constrained random events for gaming machines, control the odds, so it's sort of random, but not really?

Depends on application? What is the application?

Agree, can't imagine another cheaper method -> thinking of slow, filtered noisy oscillator with the desired event distribution, that would take more effort than programming a PIC or whatever.

I suppose somebody here can do a DSP chip for this?

Grant.

A standard noise generator shift register + feedbacks and a counter for the minimum time may do the job. This could be implemented in a CPLD like the Xilinx 9500 series.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

So, how many of these do you want to build ?

What does the buzzer look like ? (voltage, current, sound level )

When do you what it done by ?

hamilton

Hi John, I guess only weird in that it took me a while to understand it. I use to have this idea that the decay time should be the most likely time.

Sorry for confusing anyone.

George H.