SynthModules

--- In SynthModules@yahoogroups.com, "grantrichter2001"
<grichter@a...> wrote:
> Did you notice the CALCRAND subroutine?
> 
> We are free running timer W at 16 MHz and then sampling the 
> counter "sawtooth" in a non-sychronized fashion. That number is 
> then used as the seed for the RANDOM function.

Yes, I've seen CALCRAND. It's been useful in some of my other
programs. And I do appreciate the work you and Mike put into working
that out. In this case, though, the seed I'm using is the integer part
of a mixed binary number; the lower 12 bits are the fractional part
and the remaining bits represent the integer. RANDOM doesn't perform
so well here. It's not so much the cycle length as it is the "quality"
of the numbers. There are repeating patterns shorter than the cycle
length. I'll re-think the problem and see if there's a different
approach I can take. In any case, the program does what I wanted. I'm
happy. :)

For the non-programming types that may be reading this, I feel I
should point out that this is not a problem with Brice's design or
anyone else involved in the design of the PSIM-1. This has to do with
the language provided with the AtomPro. And it's not really a problem,
just a quirk to be aware of.

Woody