> Ok, so it is good. I am surprised you can pass tests with such
> a simple device.
>
> Peter

Ah, but I haven't proved that with ent.exe yet ;-)  I stand by
my claim that a random number of pulses with random
timing between them is all you need. I'd feel pretty confident
going head-to-head with a s/w PRNG. The research I did
beforehand proved to me that an unpredictable external
event was the only way to go

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.