If I wanted random numbers that were not generated by a high quality algorithm (which can generate randomness far better than most mechanical devices), I'd go with the traditional radioactive source and a detector approach (as used by NBS to do their famous tables in the handbooks)
Aware Electronics makes a little Geiger-Muller peripheral (9V DC in, pulses out) and that, in combination with a low activity source (say from United Nuclear) would work quite nicely. http://www.aw-el.com/ http://www.unitednuclear.com/ (although, with a bit of thought, you'll be able to think of an easy way to get a 1 microcurie Americium-241 source.. nice alpha emitter, about 27,000 disintegrations per second, and much longer half life than the Po210 alpha source in the antistatic brush) I've had a RM-60 for almost 20 years now, and it still works nicely. It's fun to set up a logging program and just collect data, and see how precisely it matches the theoretical Poisson distribution. > -----Original Message----- > From: [email protected] [mailto:[email protected]] On > Behalf Of Scott Burris > Sent: Wednesday, December 23, 2009 8:07 AM > To: Discussion of precise time and frequency measurement > Subject: [time-nuts] are any time-nuts also random-nuts? > > I saw this USB connected hourglass for producing random numbers: > > http://home.comcast.net/%7Ehourglass/ > > Anyone pursuing perfect randomness in the same way this group pursues > time and frequency? Maybe cryptologists. > > I'm tempted to build an ethernet connected variant of this. Then of course > we need a distribution mechanism. How about RNDP, the Random Number > Distribution Protocol? A la NTP, clients could select for the server with the > most randomness :-) > > Scott > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
