using the millis() timer0 alternative

[alto777]

Any sense I made of using the millis() compare A and B mechanism seemed to need an additional ISR, in fact it did not work until I added the following nearly identical function in the section marked "/* Alternative using timer 0 that is used for millis()":

SIGNAL(TIMER0_COMPB_vect) 
{  
  computeEncoder();
}

I write in case that was your intent and it got dropped somewhere in an editing mistake. With both ISRs it seems like I am getting 2000 Hz attention to the pins, but I have not verified this yet until I figure out how. To.

a7