[sdiy] Simplest random source

[Paul Perry]

>little with Scott's suggestion.  I found this circuit using a 4006 and 3
XOR gates:
>http://www.analog-synth.de/synths/mod2/trigdiv/dig_noise.gif
>The only drawback is lots of CPU cycles.  The shift register "clock" must
cycle 16 times to get a 16bit random >number.  Ok, so the DAC is only 14bit,
and probably 12bit would be enough, but still thats a lot of cycles (I
>guess at least 120 instructions @ ~10 per cycle).
>
I don't know if there are any mathematicians here, but I imagine there is some 
more efficient way, taking advantage of the available bit length in the
processor.
Maybe one of those systems that start off with a seed & multiply out.

paul perry Melb Australia (wondering who he gave his Knouth vol 2 to..)