> sure what's causing it or how to cancel it's effect.   The
> thought crossed my mind to take like, 10 measurements for every
> display update, and average them out.  Haven't gotten around to
> doing this yet though, because I don't think it will help the
> drift problem. Only the noise. If anyone has an algorithim that's
> already working I'd love to see it :)
>
> I'm also investigating the posibility of using a PIC with an a/d
> on it and doing it that way. I wonder if the a/d will have the
> same problems though?

I did this on a commodore PET about ten years ago. The most
critical thing is the logic level at which your input triggers.

If you are having these kind of problems I would GUESS putting
in a buffer of some sort will help... Maybe just a voltage
comparator.

I used an OP-AMP with the all the parts set really close to
each other. Helps with temp drift aswell. With care and
lots of averaging you can get really good results.

Nic