> There's an implied assumption that VCC is a stable voltage.
> I've never tried (for example) VCC as a sine wave added to an
> offset so that it was always inside the limits, but varying

If ripple is a problem, stop the Vcc boost immediately before sampling
and there'll be no noise on Vcc. With a ** suitable ** reservoir cap
and low enough Ivcc (SLEEP the PIC for an ADC reading ?), Vcc
should hold up for the acquisition. I'd be quite confident though that
noise can be filtered to an app-dependent ** acceptable ** level
without going to a lot of software/firmware bother

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist