ever changing backlash

i used backlash before and never had this problem. i set the backlash to .003" on all three axis,
ran the code and measured. ran the same code without changing anything. it was off by about .003".
ran the code again(no changes) and it was off another .003"(total .006" now). ran the code again(still no
changes) and saw it off by .009" now.
should the backlash code change every time it's ran ? i thought it would only set the backlash from zero
and not every time the codes ran. Thanks for any help !!

P.s. attached is my ini.

thanks, but i don't have any limit switches. it's one of the sherline mills. i have to set the limits
by knowing the limits. limits now are 10,10 and 8, but my machine doesn't go that far. if it
gets to the limit, i lose zero. again, Thanks for your reply !


andypugh wrote:

I wonder if it is homing to the wrong side of the backlash?
What happens if you reverse the latch velocity? ie, if you now have rapid to switch, move off, slow move on, switch to rapid to switch, slow move off.

never had to do/use stepper test. in fact, never heard of it and if it's part of HAL,never used that either.
never had to BUILD a machine. thanks JohnT

p.s. it is a stepper motor. www.sherline.com


BigJohnT wrote:

Backlash will move the amount specified each time the direction changes. So when X for example goes from + direction to - direction a rapid move of the backlash amount occurs in the minus direction at the start of the move.

Is this a stepper machine? If so have you tried the stepper test to see if your losing steps?

John

piasdom wrote:

i used backlash before and never had this problem. i set the backlash to .003" on all three axis,
ran the code and measured. ran the same code without changing anything. it was off by about .003".
ran the code again(no changes) and it was off another .003"(total .006" now). ran the code again(still no
changes) and saw it off by .009" now.
should the backlash code change every time it's ran ? i thought it would only set the backlash from zero
and not every time the codes ran. Thanks for any help !!

P.s. attached is my ini.

File Attachment:

File Name: Axis.ini
File Size:2 KB

Most likely, the 10 inch/sec/sec acceleration you have set in your .ini file is too high, and the rapid jumps when the backlash correction attempts to jump to the opposite side of the backlash is exceeding the ability of your driver/motor to accelerate, causing lost steps. Try setting the MAX_ACCELERATION parameter in each axis section of the .ini file to a lower value and see if the problem improves.

Jon

manually with an edge finder.