Yes, Bruce it was me. I do not have the pulse chart and do not have the time to recreate it, but I think it is 2 in 16, not 1 in 15 that misses. In my previous posting in this tread I fell in the trap of suggesting uC inputs, seing his very low pulse rate. Even a 1 pulse per hour encoder interface must be capable of high speed. The situation is that the encoder may stop with the lamp and cell just across a transition line and even the most minute oscillation (we are talking gears) will create very rapid direction changes.
You must use the HP chip or a CPLD or accept occasional errors.

Have fun,

Erik