Optical rotary encoders and Schmitt triggers

[atferrari]

Optical rotary encoders as in PC's mice to be read through two PIC's pins.

Schmitt triggers in between, are they convenient? Or a must? Or not necessary at all?

 

[JimB]

In my experience the schmitt is a must.
Otherwise the slow rise time of the signal while the photo-interupter passes through the light beam, combined with a bit of noise on the signal, can cause all sorts of problems.

JimB

 

[Mr RB]

I would just debounce in software. Quadrature encoders are reliable as the noise is always 1/4 of the quad (so they won't lose position).

 

[ericgibbs]

In my experience the schmitt is a must.
Otherwise the slow rise time of the signal while the photo-interupter passes through the light beam, combined with a bit of noise on the signal, can cause all sorts of problems.

JimB

hi,
I go with JimB on this one, I have always found most optical encoders have slow rise times, use a Schmitt.

A software solution sounds attractive, but long debounce delays are required to ensure they covered the worst case encoders.

E,

 

[atferrari]

Sensor output

This is what the ST will have to deal with: output of one of the sensors obtained while turning de encoder's disk slowly back and forth.

 

[JimB]

What I would have expected.

As I said, Schmidt it.

JimB