encoder not working

This c at the opto's input does absolutely cut higher frequencies.
Maybe the circuit starts sometimes oscillating without.
Beside of that I would even see no problem to take it out totaly. Or test with some quite small c's.
Can you send a photo, so maybe it is possible to recognize the value...
Mike

Thanks Mike, unfortunately the caps are both surface mount with no markings at all... I'll take one of them off and see if I can measure the value although I've never tried that before.

Nop
The one on the left is OK.
The one in the middle of your drawing is the guilty.
If you take that out, it might work up to the opto's highest frequency. But then you risk that the circuit starts oscillating. So if I would need fx double speed that worked unmodified I would measure the c and replace by half the value.
Mike

OK, I took the cap out and measured it at 100nf. I'll try the opto without the cap in this morning and, if it misbehaves I'll do as you suggest an try a lower value. Thanks.

If this is the middle one, I think this value is crasy high. It try 1-10 nf.
Mike

Well, taking the cap out altogether seems to have wrecked the sensor - it now apparently gives no output. I will try to resurrect it with a smaller cap but I don't hold out much hope. Fortunately, the postman brought me another couple of sensors I had ordered and I've fitted them for now. They still use an LM 393 comparator chip but now I have a different situation.... The PyVcp speed display now ramps up to almost 700 rpm but it then stops there even though the spindle is doing 2500rpm. However, a big difference from the original sensor is that the speed keeps being displayed as 6 - 700 (it does bounce around a bit) and doesn't just return to zero after a certain speed. It all seems to get more baffling and frustrating! I'm more used to seeing gears go round and easily spot problems!!

What you mean by wrecked? Distroyed the pcb by to heavy soldering.
I don't think that taking out the capacitor can effect either the optocoupler nor the lm.
Resoldering should be enough. But I agree it seems to be a pai.. In the a... story.
Mike

Nope, the pcb is fine, the sensor worked for a few seconds with the sensor LED flashing as the disk rotated but then just ceased and, while the power LED was still lit, the output LED didn't flash any more.

And you tried to connect the c again?

Thanks, I've been away for a few days. I think I have solved the problem by changing the cap as you suggested. I have now changed the 100nf caps for 3nf ones and the phase a signal is now reporting the correct speed. I first tried changing to a 33nf but it still cut off at about half speed but the 3nf (just a size of surface mount cap I already had) now seems to have done the trick. I will have to keep an eye on possible oscillations etc though.. Thanks for your help - I now have a working cnc lathe conversion!