[Re: Re[2]: [PolyModular]]

[Harry Bissell]

Harry Bissell: The "spread spectrum" technique is better at helping a cpu pass
FCC (or similar in your area) RFI emissions tests, where the noise must be "so
many" DB down... The fast rising clock edges will probably still break through
in the analog circuits. Controlling the rise/fall times might help. There are
some studies that indicate that random noise added to a D/A converter can help
mask the 10101010 dithering that humans find so annoying, esp. when there is
other noise like power supply hum that makes the noise 1111000011110000. 
There are other ways....


"Tony Allgood" <oakley at techrepairs.freeserve.co.uk> wrote:
A nice trick to avoid HF interference on time division muliplexing is to
use a single bit output of a pseudo-random noise generator instead of a
fixed frequency. Or you could scramble it with an ex-or gate. No good
for commercial designs though, its still under patent protection.

Regards,

Tony Allgood, Cumbria, UK

Rack mounted moog filter and the TB3030 SuperBassline projects:

http://aupe.phys.andrews.edu/diy_archive/schematics/oakley/  






____________________________________________________________________
Get your own FREE, personal Netscape WebMail account today at http://webmail.netscape.com.