Hello. With dial mastering of an older krc1 equivalent robot, at what position of the dial should the axis be mastered? exactly at the turning point, or just as it begins to reverse direction? If the mastering should be done after, then how many fractions of a mm should it be? This is an old robot and with a short probe I get min 1.7 mm of a tool calibration error.
dial mastering
-
mbalazs -
January 19, 2021 at 2:38 PM -
Thread is Unresolved
-
-
with Dial mastering you need to find as exactly as possible center of V-slot. so normal approach is to find the dip, then reduce speed and pass it again (and repeat as needed) until you are right at the center, not before, not after. 1.7mm error for tool calibration is quite large but that also depends on how well defined your pointer tips is, how far the TCP is from flange, size of the robot, how well it is maintained (belts tensioned correctly?) etc.
-
Thank you. I've got 1.7mm at best, with a precision machined tipped rod mounted in a small spindle. Now I've switched to a larger spindle with an aprox. 500mm tip distance from the flange. This way I can't get below 6mm error. I once got 2.mm, but with little orentation change between xyz4 point measurements. The robot is a kr125/2. The weight of the spindle is around 40 kg. First I will try to get better results by making a more careful remastering, as you've mentioned. How much does belt tensioning influence accuracy? Does incorrect tensioning introduce more backlash?
-
too much tension will damage the wrist, usually it is the belts and eccentric shaft that pay the price. too little tension and wrist may be skipping the belt teeth and of course damaging belt. tension need to be really accurate (1% or so) and therefore cannot be done without proper equipment and training. if used, belts are usually in the wrist but some robot models also have belts in the robot arm.