Omron Rotary encoders - stepper motors.

andypugh wrote:

It _might_ be possible to connect the pid error value to the adaptive feed pin so that step-rate reduces while still making up the lost steps, but I don't know if anyone has ever tried it.

My Allen-Bradley 7320 had this feature, but it did not have a smooth on/off logic. When the feedrate caused increasing error, the machine jerked to 1/2 the feedrate, the motors caught up, and the cycle repeated. I had to disable it to prevent tipping the mill over.

Jon

@andypugh i wonder how those "closed loop stepper motors" drivers deal with this then, they are basicly a rotary encoder with a closed loop in its own.. if it would counteract the purpose of it, it would just end up being an interesting experiment i guess..

of course that makes sense.. hmm.. Wonder if one could "copy" that behavior within linuxcnc tho.. but some form of "measuring" needs to be performed then.

Theoretically yes, practically I'm not so sure. Biggest limitation might be timing resolution and granularity (a 1kHz servo thread likely isn't going to be fast enough.)

I've wondered if a high resolution encoder couldn't help with improving the accuracy of a high microsteoped stepper (32x microstepping or so) and limit max following error to something less than 1 full step.