Chasing an X vs Y movement issue on a DELTA....checking Steps/mm

edited July 2016 in Motor Control
I have been chasing a problem in the X-Y axis relationship on a delta printer...
By this I mean the actual Cartesian X and Y don't seem to be the same when commanded to move the same..

I am not 100% certain but I suspect after numerous test that maybe the towers' steps/mm are not identical.  I see that there are settings for each  in the firmware config tool, though they are labeled XYZ I think., while I refer to them as ABC.  Can I alter one to correct if one moves differently from the other?  I know in reading up on the subject that there was some discussion that they had to all have the same steps/mm setting even though the configurator uses multiple numbers.  Does the system always use just one of the numbers when doing the math?

I am hoping to make a mechanical fix like a belt adjustment or something to correct my issue as it seems likely caused by mechanicals anyway but is there a software way of doing it or not?

Next question assuming all the columns are using one step/mm or they are all set the same can I assume that if I move the extruder down vertically from the home position a fixed amount I should see a identical motion at each tower...I don't expect it to be the same as the extruder move but I am assuming that the motion should be the same at each tower so I can measure and compare motion of the 3 cartridges on the rails.  Is this a good way to measure the steps/mm or is there some other method that can be used?  My guess is this might be a belt tension or pulley alignment thing...I have never found a definitive way to set belt tension the same or even how tight that tension should be. 

Any help would be appreciated.



  • All towers use z steps per mm for computations as they need to be identical. Pure z moves move all axis the same distance, so that can be used to test difference. Adjust belt tension to make them equal in steps.

    Apart from this geometrical imperfections also can make xy different.
  • edited February 2017
    Thank-you for the tips...

    I have recalibrated things again following your advice...and things are a bit better.

    I am wondering however, it is possible that shifting extruders while a Delta is homed causes some problem, like end stop
    collision?  I added 10mm to the back away last time I built the firmware but maybe this was not enough (though better that the 5mm I used previously). I notice now that my start code which issues T0, T1 commands (is that the right way to do this is or are there M or G commands) to change and heat/prime the extruders seems to cause a 'clunking sound' not the smooth transition from ext0 to ext1 and back.  If I remove these the result seems to be more level travel over the bed during print.

    Also I get odd sounds when I adjust the travel accelerations in the start code, even without the extruder changes.  I use something like this M201 X500 Y500 which I assume is correct for Repetier firmware as opposed to the M204 some other systems use.

    Thanks again for all the work and assistance.
  • - M201 - Set max acceleration in units/s^2 for print moves (M201 X1000 Y1000)
    - M202 - Set max acceleration in units/s^2 for travel moves (M202 X1000 Y1000)

    so your M201 are not used for extruder changes.

    And yes, if extruders offset you need to go down after homing a bit more so you do not trigger endstops are position is wrong after homing. That is no problem since you can not print at that height anyway. It is just an illusion you can, except pyramids of course.
  • edited February 2017
    yes I didn't think the extruder change would use those accel values I only noted that they are in my startup code as it prints a small start up pattern and I wanted to set the accel values for that...oddly however I get clunking sounds and bad first layer level if I do use the M201 X500 Y500 in my code but not if I comment that one line out...the clunk happens on the first move down to what I call extruder heat position.  It isn't critical as I don't need to set accel that badly, but just odd...maybe it is still due to needing more initial offset but things work if I just take out the M201.
Sign In or Register to comment.