strange head motion...

I'm setting up a delta printer using Repetier .92.9

I am having a heck of a time getting the basic bed level and motion working.

Actually my static measurements say everything is near perfect mechanically, towers, arms bed etc.
I am not using any firmware leveling, or rotation or flatness correction yet as I am trying to get the
hardware as close to perfect as I can.

My problem is if I print a few skirt loops about an inch from center, they do not touch each other where they get closer to the Z tower.  Opposite the Z tower they look really good and get progressively worse as they get closer and cross the center line of the Y axis.  They seem to actually spread out.  It isn't just that they don't flatten and spread against the bed, they are just farther apart that whole side the print just wider.  I have posted a picture here if a visual helps.  (the seam point is aligned with the Z tower when looking through the space between the X and Y towers.

I am hoping someone can tell me what parameter in the firmware I might adjust to fix this or alternately what mechanical parameter might be off....

I have levels and relabeled many times using many methods, I have measured everything I can many times.  
It seems to me that either both X and Y towers are off in a similar way or the Z tower is off.  OR the arm length
is wrong on the Z or something like this but I sure can't find it.  It doesn't seem to really get worse as it moves
out toward the tower , you can also see that there is some of this all along the the shorter sides in the photo though it is a X offset instead of Y a Y is not as bad but still present....It is possible the part that looks tight in the photo is actually where the problem is and the loops are closer than they should be.

Any ideas would be appreciated as I am all out.



  • Think about where which axis has most influence. Near Z axis moves of z axis have a big influence and for away it looses influence. So if a error increases near z axis it is most likely that z diagonal is wrong. If you rotate the object 45° you might get the same image also for x and y axis meaning all diagonal rods have wrong length set. Of course if you change them the center will not fit any more and you need to adjust horizontal radius when centered as well.

    Maybe have a look here
    and read comments for repetier adaption. 
  • Thanks. I had seen that thing before and I will revisit as the info seems to have increased in detail.

    Question: Do the Corr. diagonal A, B, C parameters I see in the EEPROM help with differing diagonal rod lengths. Is this a mm measure that is added to the base diagonal rod length? If C rod set is physically longer that the Diag Rod Length by .1 mm do I put -0.1 in the C correction to fix it?
  • Exactly each rod is assumed diagonal length + correction (in mm).
  • Okay thanks for confirming my understanding.

    Recent print measurements seem to indicate some sort of mechanical geometry problem. All the arms were jig built and within my ability to measure are identical. However the problems seem to be related to their length, that or a stepper motor or drive gear being out of spec. Or maybe a belt?

    I will play with some minor ABC corrections to see if they have a postive affect as this seems the more likely cause plus it is way easier to tweak numbers than change mechanicals.

    Thanks again for the great software and firmware and the assist.
  • Pure z moves move all steppes the same. You could use that to check steps per mm setting.

    Diagonal rod length is from rotation center to rotation center in case you measured a different distance.
  • Hmmmm good to know about the pure Z move....

    I am thinking if I put a dial gauge on the bed and zero it against the nozzle then descend 2cm and if I get 2cm on the gauge then I am probably good...if the gauge number is high or low I will have to adjust...I think, even on the delta, it should be proportional.

    If 2cm is not enough to see the variation I could probably do something similar with a caliper....maybe that is even better if I can find a way to mount securely and squarely.


Sign In or Register to comment.