Hi All,

Uludere, I tried to send you circuits etc but I'm a bit of a novice to this forum and the files are too big. I'll cut them down to size this evening and send tomorrow, sorry for the delay.

Sougata, of course for voltmeters, the PIC ADC is really not suitable for any real precision. Nor is any other normal µcontroller for that matter. RMS-type measurement is easier but not better, especially for low levels and changing frequencies, which is why I even contemplate AC chunks. It isn't necessary here to check the zero cross though, assuming we're sampling offset peak to peak and calculating RMS.

It isn't necessary to measure the exact voltage value, either, and that in all cases. Of importance is that the measurement system used is consistant, that means it detects the same value of any given voltage at all times. Software correction to convert to the correct RMS values is then easy. The above statement includes your remarks about temperature correction.

A trick I have used, not for AC voltage but for reading optical sensors, was to couple a number of ADC inputs together and to let one intergrate while another converted. This allows a higher reading rate when reading chunks, and of course for all other cases too. Watch the delay requirement for switching ADCs though.

Greetings,

Rob