Vasile, thank you for the compliment with the perpetuum mobile, but I
think that it works. I have good resons for that since I once wrote an A/D
that works like that and attained 10 bits resolution in the range that
interested me (near 2V input).

In short, you turn the PIC pin to output for a certain (short) time and
output a 1 or a 0. The T filter circuit acts as the integrator. Then the
PIC pin is turned input and A/D conversion is done. The small charge
change in the 0.01 uF capacitor induced by the output pulse biases the LSB
in the desired direction. By using 16 different combinations (length and
polarity) of output pulse and 16 measurements a resolution improvement of
4 bits should be possible. It is possible that the value of the capacitor
will need to be changed somewhat. I'll need to simulate the circuit to
tell ;-)

Peter

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu