PIC16F1827 at 32MHz and PBP.

My AN channels, specifically AN5 and AN6, are each connected to the wiper of a different pot giving a voltage range of 0-5VDC.

Based on what I see on the oscilloscope, with 0VDC present on the ANx pin, the ADC for that pin is reading about 60 counts high out of 1023. When I advance the pot, no effect is evident until about 150 counts is reached (i.e., an additional 90).

It does not appear to be a circuit problem; I have checked the voltage on each pin at the millivolt level and even gone so far as to strap each pin directly to the PIC's ground to verify that the pin is at 0VDC. The "bias" is still there. According to my math, I'd need at least 0.3 VDC to account for it.

I suspect, from what I'm seeing, that this also may be occurring on signal channels AN0 through AN4.

Ideas, anyone?