Rotary Encoders 1

[Ridly]

Is there any way to bench test a (0-499 pulse) enconder by
bench testing. I`m trying to trouble shoot just this area of the machine that giving me problems. Thanks!

 

[djs]

One way is to rotate the encode with a drill motor and look at the pulse outputs with an oscilliscope. Missing pulses can sometimes be detected this way.

 

[itsmoked]

Just count the pulses..

Just kidding. djs has a good idea there. Run it as slow as you can while looking at the scope. Get a good see-able scope picture with about 30-40 pulses on the display and you should be able to see any dropped pulses pretty easily.

Often "dirt" could bridge some engravings and cause lost pulses.

 

[Skogsgurra]

Better still (I use it routinely) is to run the encoder in the application. Trig on [pulse width < measured pulse width]. That will reveal any missing pulses. Remember that a bad bearing can produce a good pattern until there is some axial play - then the bad pulses show up.

Also, it is not sufficient to check just one channel. Most encoders have A and B channels, sometimes also /A and /B (complementary channels). Check that phase shift between A and B is 90 degrees +/- a tolerance, usually 20 - 30 degrees.

Also check zero level and high level as well as noise.

Same checks can be made out of the application. Run in a test bench or couple to shaft of asynch motor with a piece of rubber hose. Be careful - make sure that the encoder and motor do not get loose.

There is a glossary

http://www.gke.org/rapporter/Glossary for the encoder world.pdf

that can be of some help.

Gunnar Englund

http://www.gke.org

 

[itsmoked]

Good ideas there skogs. I was thinking... I would leave it in the application too. You could easily end up with a fine "bench encoder".