Did Backlash Compensation Need Linear Encoder To be Feedback ?

[compoundMR]

im trying use a backlash compensation , and first , when i try use it by just moving G1F100 X-0.1 and then G1F100 X0 it seems the backlash compe work normally , but after i use a long gcode , my 0 point is change and not consistent do u have any idea what happen is it because of inertia movement ? or backlash compensation need linear encoder or ballscrew encoder to be feedback ?

[terjeio]

backlash compensation need linear encoder or ballscrew encoder to be feedback ?

No. The backlash compensation is only added on direction changes and based on the configured distance.

my 0 point is change and not consistent do u have any idea what happen is it because of inertia movement ?

Is the long movement a single movement or comprised of many small that also includes direction changes? And are you sure the configured backlash distance is correct?

[compoundMR]

thanks in advance for replying ,

im sure i get 0.02 backlash on my machine and i only activate X axis, and here is my gcode i use for test

backlash Test.txt

[terjeio]

When I test with 250 step/mm and an odd setting just above that I get 5 steps out of sync (by counting steps). What is your step/mm setting and how much is zero off?

[compoundMR]

sorry for late reply , im setting it on 500 step/mm ( 0.002 accuracy ) , and get 3 miss step per section ( make a counter pulse on another controller and count the steps) , and get alternatively solved by increasing the step/mm into 1000 step/mm (0.001 accuracy )