I have built an IR sensor to measure RPM:

Some tests:

And the sensor build up:

With the proper encapsulation and the holder already attached to the engine:

Some preliminary test did work well unless the sun was lighting it directly. I adjusted one of the resistances of the circuit to reduce sensitivity and it now works perfectly in all condition:

My sensor says 5085rpm while the commercial one measures 5160rpm... not too bad! BTW, it has a precision of +/-15rpm because I make 4 measurements per revolution in a period of 1s (actually this may change, but at the time of taking that picture, that's how it was).

I have a video, but there is no way to read the tiny screen anyway... I'll put the schematics of the circuit another day as I don't have that stuff with me now.