Hello Guest it is November 22, 2019, 02:33:28 AM

Author Topic: Esab flame cutter retrofitted with Mach3 and CSLABS CSMIO-A part accuracy issues  (Read 3729 times)

0 Members and 1 Guest are viewing this topic.

Several years ago we retrofitted a large Esab flame cutting table with Mach3 and a CSLABS CSMIO-A analog servo interface.  We were never that impressed with dimensional accuracy of the parts.  They seem to vary around from 0.060 to .125 inches in length and width.  Whatever is going on it seems to occur at exactly the same rate in both axis.  One can literally cut a line of parts orientated the same and get one that is within .010 length and width, the next one 0.030, 0.090, 0.120, etc, not necessarily progressively larger in size.  We use no kerf compensation in the program, I do all of that with the post processor.  The DRO follows the path perfectly on the mach interface.  I do call a G92 X0.Y0. at the beginning of the program to set the program zero point. Has anyone ever had any trouble with the pulse train loosing or gaining steps when talking to the csmio-a interface.  I really need to get this issue fixed as it is makes fixturing up parts to machine interesting.

Offline Davek0974

*
  •  2,580 2,580
    • View Profile
    • DD Metal Products Ltd
I have a CSMIO-A on my mill and have not lost a step yet, in fact when it does lose position while tuning it will fault out and stop. Do your servos have encoders connected back to the CSMIO?

Have you eliminated all mechanical possibilities?
Bridgeport Mill, Mach3 V062, CSMIO-IP/A controller, AC Servo Drives