Skewed print any suggestions. SOLVED



  • So printing a cylindrical object and it leans towards the -x direction. I am printing it on an all metal xl kossel, which so far has been faultless. I was using 1.17rc1 and assumed it was a firmware bug, but I've tried grid compensation on and off and gone back to 1.17 dev8 and 1.16, same issue so it's not the firmware.

    Towers are 90degrees to the bed and obviously the bed is level as a result. Arms are Hayden Huntley magnetic type, bed is tooling plate.

    It autocalibrates to 0.03mm deviation and until 2 days ago nothing has been skewed like this, the only thing I changed was the firmware.

    Stumped….



  • How did you determine that the towers were orthogonal to the bed? If it was using that aluminum square, did you verify that the square is itself accurate? You can do this using a sheet of paper and a pencil, tracing the square, flipping it and seeing if the lines are still parallel.

    Edit: I see now that you say this started happening out of nowhere, so that leaves me a bit stumped too. Did the frame get shifted or bumped somehow? If you are in the Northern Hemisphere, maybe the cold weather has caused the frame to shrink and deform?



  • I've got that square and a digital angle finder which both agree with 90 degrees for the towers. The printer is indoors so hasn't varied by more than 3-4 deg C ambient over the last week and hasn't moved a millimetre. I did post this on reprap and someone suggested a slicing error related to the fact that it is a cylinder and possibly related to its wall thickness, which seems very odd, but I'm just going to print a selection of cylinders some solid some thin walled and some thick to see if it makes any difference. I'm going back to 1.17rc1 as I see no reason not to and reenabling levelling grid, since this made absolutely no difference either way.



  • You could look at the gcode and verify that each circle is printed with coordinates that only change in Z, in order to rule out a slicer error. Very odd that it would appear out of nowhere.



  • That's a good idea, I'll do that.

    Other things I've tried:
    -a selection of cylinders one single walled, one 2mm one 4mm one solid. All the same lean, irrespective of bed position. I tried one with random seams and printed slowly.
    -I've checked the mechanicals the pulleys are not loose, the belts are tight, between the towers measured at the top and bottom they are all within 0.1mm - for 1000mm towers.
    -The PSU is giving a solid 23.5-24.5v throughout.

    I am quite puzzled but enjoying trying to work it out.



  • you can also load up Gcode to Gcode.ws to inspect it layer by layer etc


  • administrators

    Is the height of the cylinder correct, all the way round?





  • Solved it was the steel reinforced belts losing their reinforcing and stretching.



  • Well, that is a relief!



  • I've not heard that the steel reinforced belts are useful for our machines. Particularly, small bending radii damage the steel belts.



  • Yeah I'd agree and to give DC his due he did question my selection of them when I was putting the machine together. However I'd used them on my corexy for a year or so before they failed, I suppose the section of belt typically spending most time around the pulleys is smaller on the delta, if often printing small object in the bed centre. I suspect with 20t pulleys they would last longer, indeed they do, I had them on my kossel mini which I recently sold for quite a while with no degradation.


  • administrators

    We have found the GT2 belts that are commonly used (including by us) on deltas from small to ~1.2m large work well. They are fibreglass core AFAIK and have no appreciable stretch. The bend radius has no issues in 20T pulleys that we use.

    What I don't know is how well they scale to larger printers where stretch may be more pronounced.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.