>> I don't believe that putting
>> Vdd on an analogue input pin causes any physical harm. A25 is not
>> very well explained is it, but I can't say I'm totally au fait with every
>> aspect of the A/D, just helping you kick a ball around
>
>It's good you did, cause I hadn't looked at the analog i/p model until you
>mentioned it.  I'm sure they mention the max limit on the input signal with
>reference to Vref+ for a good reason.  I'm sure the answer will pop up
>sometime, even if I have to ask Microchip's tech support directly.
>
>In the meanwhile, I'm designing with a zener added to the input.  However,
>seeing the input model fresh in my mind, I have an alternate idea --
perhaps
>I can add my own protection diode (non-zener) to the input and going to my
>+2.5V ref instead of Vdd (as the internal upper input-protection diode
does).
>I've given up trying to derive a formula for the sensor for various
reasons,
>so I'm using a table to derive the measured data based on the a/d
conversion
>readings.  With this system, I can calibrate around any effects presented
by
>any diode(s) I add to the input pin.

How many analogue inputs are you actually using? If only one then I do not
think you have a problem.

However if using more than one, and one analogue input goes outside the A25
limit then IIRC all analogue inputs will show either full scale or 0
depending on which way you are outside the limit (Vref+ + 0.3V or Vref- -
0.3V), even though you may be inside the Vss and Vdd limits (remember that
you are allowed to have Vref- more positive than Vss).

I seem to remember this being discussed in the past. Have not hit the
problem myself though, as I have used only one input so far.

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