But what if the encoder is sitting on the the edge of the optical bar one of the phases is sitting there toggling back and forth at some cadence that is caused by any of:

- Machine vibration
- Electrical noise
- Wobble of nearby black hole

Now consider that this toggling can happen at rates of several megahertz ( I have seen it so do not say it cannot happen). This means that your slower timer interrupt that comes in and samples asynchronously may sample when the phase signal is low or high. Even worse it could be sampling when the phase signal is just now transitioning. At any rate you could be seeing it showing sometimes low and sometimes high. You can never be sure and certainly the phase signal may have transitioned an uneven number of times when your sample routine may be thinking that this signal is up-down-up from timer to timer interrupt. You will realize a net loss or net gain in your accumulator that does not match what the encoder has actually done!!

I'm not convinced that your "tried and true" solution is really any better than one that tries to operate off the edges of the signal.

Michael Karas