I am using Timer A to measure an RC decay. I am comparing to Vcc/2. At Vcc=2.2V or greater I get very consistent results from the comparator every time I make the measurement (+/- 2 counts). If I drop Vcc to 2.1V or less the spread suddenly jumps to +/-7 counts. I see that the comparator and the reference are only specified down to 2.2V. Does the expected performance change that much at 2.1V?. Any suggestions how I might be able to improve this?
I saw some mention of adding an external resistor for more hysteresis. Would this help?
