> Joe Dowlen wrote:
>
> >      Floating point arithmetic is a necessity when using the PIC14000
> >      processor. The A/D converter calibration is difficult at best if you
> >      do not use floating point math (to correct for the calibration
> >      constants).
>
> Good point.

Is there any reason why the floating point maths should be done in the PIC
rather than in the PC that programs them?  One of the great things about PICs
after all is their ISP'ability and the ease of writing C to interface with
their programming logic.  I would think a program to read out the calibration
constants, scale them into whatever format the PIC software "really" needs,
and then write them back [elsewhere] would be the most sensible way of handl-
ing PIC14000 calibration.