At 08:59 10/26/2001 +0100, D Lloyd wrote:
>Is anyone aware of methods for "extending" the resolution of an AtoD
>converter in software/some other cleverness? I'm aware that you can take
>several readings in succession then average them to obtain some measure of
>improvement but is there any better solutions?
>
>Also, has anyone implemented a sigma-delta converter in software? Is there
>a rule of thumb for "processing required vs number of bits of resolution" ?
>I believe the Nyquist limit of the sampled signal is 1.6kHz and a 16-bit
>micro (with on board 16x16 multiply) would be available to do the work.

My rule of thumb is to make the hardware adequate, if at all possible. You
don't say what resolution you need, but at 1.6kHz you have quite a choice
of ADCs. How much you have to do in software depends on the hardware you
have and the goal (neither of which is stated... :).

ge

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.