Did you get 3rd harmonic xtals?

Erik

Re guesses:

In cases like this you usually scope a few points to get your theories (un)confirmed. Ever so often looking for a you see b and THAT lead you on the path to recovery.

With "mysteries" the important thing is to observe what you see, not just check it aginst your current theory.

I was introduced to encoders with an existing design (a=clk, b=dir) and spent too much time checking all kinds of signsls, replacing the encoder and what not. Finally I drew out all possible direction changes (such changing to reverse with a low and b is high and moving forward) and saw that in one case a count would be missed.

In another case I could not see anything wrong on the scope and finally said to myself "that HAS to be it" correcting the possibility of a nanosecond pulse at point a fixed the problem.

The morale:
Do not believe what you see and analyze what you see not only in relation to your theory.