ever changing backlash

i ran the step timing a number of times on both axis(X-Y). i tried to lower the max_acc lower and lower each time. when i got to about 8 or 9,
it froze comp. i started going up in max and didn't see any different. so i resit to default( max_acc 10 - max_stepgen 12.5). i also noticed (running the test) that i have about .008.5 backlash in stead of the .003 to .005 i thought i had. i adjusted the backlash in the ini and everything seem to work now. one more question, while running the test i watched the dial indicators' needle. on the second part of the test(no pause) the needle didn't go to 0 traveling .5"
then to 1". about a .015" different. i know it's traveling a 1/2", but the needle should be at the zero each time. is it just that the comp/machine is moving faster then i can see ? MAX_LINEAR_VELOCITY = 0.34 or 20.1ipm. when running with pause the needle is about .001" different.



BigJohnT wrote:

The first test I would do is to comment out all backlash then run the stepper test for each axis to try and figure out if your loosing steps due to friction or something.

www.linuxcnc.org/docview/html/common_Ste...iagnostics.html#r1_3

John

not sure how to check that. is it possible ?

andypugh wrote:

piasdom wrote:

manually with an edge finder.

I wonder if backlash compensation is active during manual jogging? It is possible that it might not be,