[sdiy] Vote for your favorite noise source

[René Schmitz]

At 17:39 13.05.02 +0200, Sowa Roman wrote:
>Each would make different FIR filter response, and
>not limited to only white, pink, red, brown, etc (black? :>)
>The only problem - 31 resistors for each response, or
>rather trimpots as their values for sure won't hit E24.

I think one could make a reasonable approximation. I really 
wouldn't care if it isn't exactly gaussian, even less for pink 
or azure. 

>And 31 may be too low to have good response at low
>frequencies.

The nice thing about this is you can vary the clocking frequency,
the corner frequency of the digital filtering is proportional to 
the clocking rate. I think it could be made relatively low without 
affecting the spectrum of the noise relative to fclk. But in absolute 
frequencies the lower the clock gets the denser the spectrum becomes. 

Cheers,
 René


-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159