Hi Erik,

the trick on my debouncing was also to sample very fast. So a false code on 1 sample can be corrected on the next.
E.g. a sequence of
0-0-0-1-0-1-0-1-1-1 give in result 3 increasing and 2 decreasing = 1 pulse.

Since the encoder give out gray code, my detection of wrong codes was only needed to see if the scanning speed was to slow.


Peter