In-Reply-To: <Pine.GSO.3.93.961009132205.27901B-100000@autarch.acsu.buffalo.ed-
u>
>  I didn't realize this, but still, I even tried using VDD as my
> reference, I still get inaccurate readings on successive samples.

In my experience innaccurate readings are normally caused by not allowing
2Tad settling times before commencing the conversion. Either that or a
poorly regulated Vref.

Bruce