I am decoding a quadrature hall sensor (allegro A3425).
the 2 pins go to 4.7k pullup with .01uF cap to gnd. then they go to pins C1 and C2 on a mega168P (I have not tried it on a 168PA). The processor is using pin change interrupt.
we have built a few thousand units, and suddenly a few % are failing. I have 3 units that all show the same issue:
rising edge on C1 (measured risetime is about 100usec) very occasionally does not fire the interrupt (I have added a toggle pin B3 every time it fires).
I have never seen C2 fail to fire.
I have stripped the code down to just initialization and this interrupt, and I am turning the magnet by hand very slowly. the scope catches the condition easily.
Anyone else seen missing pin change ints on this pin?
there is no errata in the data sheet.
Most of the units work fine, so I am expecting some weird hardware anomaly.
thanks in advance