1994-06-22 - Re: Hardware generators

Header Data

From: Phil Karn <karn@qualcomm.com>
To: wcs@anchor.ho.att.com
Message Hash: e353664234e331fb5358febcbe11a081dcd7dcadfb1c5dc449ccc6a45a92fad8
Message ID: <199406222135.OAA06068@servo.qualcomm.com>
Reply To: <9406200816.AA03886@anchor.ho.att.com>
UTC Datetime: 1994-06-22 21:38:10 UTC
Raw Date: Wed, 22 Jun 94 14:38:10 PDT

Raw message

From: Phil Karn <karn@qualcomm.com>
Date: Wed, 22 Jun 94 14:38:10 PDT
To: wcs@anchor.ho.att.com
Subject: Re: Hardware generators
In-Reply-To: <9406200816.AA03886@anchor.ho.att.com>
Message-ID: <199406222135.OAA06068@servo.qualcomm.com>
MIME-Version: 1.0
Content-Type: text/plain

Is this trip really necessary?

Seems to me you can get strong random numbers without having to build
hardware from scratch. Just get a sound interface (widely available
for both ISA and parallel interfaces) and run the noise from an FM
receiver into the A/D converter. Hash the output with MD5 and you have
some pretty strong random numbers.

I wouldn't be surprised if with some 16-bit boards you could do away
with the FM receiver and just sample the noise from the analog preamp
and mixer stages. Many of those boards are rather noisy.

It is always better to buy than to build when you can make effective
use of some mass-market PC peripheral. Not only do you save a lot of
time, but it'll be much easier for others to replicate too.