J
Oops - forgot about the OP's other question about the CNC welder..
I'm a big fan of periodic volumetric testing ("ballbar testing") of CNC equipment. Most CNC's nowadays have software that allow a volumetric test to feed backlash and lead error compensation into the control to effectively 'rejuvenate' the CNC from error due to wear in the lead screws and slideways. I certainly wouldn't think of this as a calibration system requirement, but it's a nice tool for preventing errors before they happen as opposed to detecting them and reacting to them after the fact.
One quick and dirty way to see if you need backlash compensation: in each axis, drive the CNC in one direction to various points in its travel. Place the measuring point of a test indicator on some part of the driven component (spindle, welding head, etc.) so that the indicator is secured on a magnetic base or by other means to the table of the machine. Move the machine in the opposite direction and compare the results from the indicator to the distance the machine 'thinks' it has moved. If there is a discrepancy between the two, you need backlash compensation.
Kind Regards -John
I'm a big fan of periodic volumetric testing ("ballbar testing") of CNC equipment. Most CNC's nowadays have software that allow a volumetric test to feed backlash and lead error compensation into the control to effectively 'rejuvenate' the CNC from error due to wear in the lead screws and slideways. I certainly wouldn't think of this as a calibration system requirement, but it's a nice tool for preventing errors before they happen as opposed to detecting them and reacting to them after the fact.
One quick and dirty way to see if you need backlash compensation: in each axis, drive the CNC in one direction to various points in its travel. Place the measuring point of a test indicator on some part of the driven component (spindle, welding head, etc.) so that the indicator is secured on a magnetic base or by other means to the table of the machine. Move the machine in the opposite direction and compare the results from the indicator to the distance the machine 'thinks' it has moved. If there is a discrepancy between the two, you need backlash compensation.
Kind Regards -John
Last edited by a moderator: