Use interrupt to determine a change in IR detector

Question

I have build a schematic to read a counter. The counter has a white point at one of the digits, that I detect with a IR transmitter-receiver. The voltage at the receiver changes from 2.2 to 2.5 volts (with some noise between) when te white point passes. So I use a opamp diff amplifier to change the voltage range to 1 - 3 volts. That signal goes to opamp comparator that changes at 2.2 volt input. I measured the voltage and indeed, only when the white point passes, the voltage is more than 2.2 volts The output from the last one goes to a arduino that I programmed to count the signal via a interrupt. The problem is now that I have false counts, but I cannot find why. I use a software debounce that is sure big enough to avoid false detections caused by bounces. I think it has something to do with the hardware. Hereby a schematic. Unfortunatly I have no scoop...

^{simulate this circuit – Schematic created using CircuitLab}

Answer 1

You might need to apply hysteresis to OA2 - this has the effect of minimizing the effect of noise as the signal from OA1 gets close to the threshold point defined by R7. Try a 1k ohm in series with the signal from OA1 to non-inverting input of OA2 and something like a 100k to 1M ohm from output of OA2 back to its non-inverting input. The size of this feedback resistor will set the amount of hysteresis.

Answer 2

I agree with Andy's recommendation for hysteresis, and I want to also suggest that you use a better op amp than the LM358. With a supply voltage of 5V you can't guarantee that the output voltage of an LM358 will get any higher than about 3.5V, which is a pretty sloppy logic 1 level. You should look for an op amp that has a "rail-to-rail" output instead.