Pleased with myself

[throbscottle]

I found how to add hysteresis to a photo-interrupter, now mechanical noise is banished!
So anyway, the interrupter's transistor outputs into the gate of a p-mos, originally to make the gear-tooth shaped waveform rectangular-ish I finally realised I can feed some signal back from the p-mos drain to the LED of the interrupter. Viola! An opto-schmitt trigger.

 

[AnalogKid]

s.c.h.e.m.a.t.i.c - ?

 

[throbscottle]

h.a.p.p.y. t.o. o.b.l.i.g.e.-

I used the BSS84 in the original design because it's threshold voltage is similar to the nominal output from the photo-transistor, and the old circuit needed a p type transistor to drive it (also, it's simple) - so I just left it in. PL6 carries the 0v, LED and photo-transistor's collector for the interrupter. I need to change the LED feed resistor for a higher value now because on the high outputs it's in parallel with R48. C33 probably needs to be removed - it was part of my original noise suppression scheme (the other part of which was a 1uF to 0v on the collector pin which I also haven't actually taken out yet).

The input to the circuit is gear teeth passing through the photo-interrupter, so any vibration around the transition point created a lot of extra pulses. The next stage of the circuit is actually a noise discriminator (discussed elsewhere) followed by a monostable, but even these are fooled by the very worst of the vibration - so I needed to catch the signal at source. Without properly checking I seem to be getting a volt or so of hysteresis. Seeing noisy 1 and 0 states, but only in the area of about 500mV so that's ok.

 

[AnalogKid]

Looks hysterical to me.

ak

 

[throbscottle]

I just realised the load is missing from the BSS84 - you'll have to visualise a 47k from drain to 0v. Otherwise it's merely funny...