Vertical ribbing/lines



  • Hello,

    I am using a CoreXY machine with a Duet 0.8.5 board and am suffering from what appear to be vertical lines, ribbing, or pulsing artifacts. This only occurs at lower speeds, and almost completely disappears when printing at high speeds (100mm/s or so).

    The datasheet from the manufacturer states the stepper motors are rated as follows:

    Rated voltage: 4.2V
    Step angle: 1.8
    Resistance per phase: 2.8 ohm +/ 10%
    Inductance per phase: 4.8mH +/- 20%
    Current per phase: 1.5A
    Holding Torque: 5.5Kg.cm

    The printer manufacturer has the following settings for motor currents in config.g:

    M906 X1500 Y1500 Z1600 E1500 I100

    Can anyone shed any light on what might be causing this?

    I have tried running the extruder motor at a variety of lower amperages with seemingly no affect on the lines, although these are not scientific tests.

    Thanks

    https://ibb.co/i331Da
    https://ibb.co/kQfgDa



  • Are those lines on all four sides?



  • Yes, but less apparent on X



  • Make it easier for everyone else to see. I have never seen anything like that, so hopefully someone will chime in.

    What are you M201, M203 and M566 settings?



  • M201
    Accelerations: X: 2000.0, Y: 2000.0, Z: 300.0, E: 4000.0:4000.0:4000.0:4000.0:4000.0:4000.0

    M203 X33000 Y33000 Z1000 E10000

    M566 X600 Y600 Z40 E950



  • Try cutting all of your M201 numbers in half.



  • Yeah, tried that, still no luck. The issue actually becomes worse the slower the machine prints. It's really odd.


  • administrators

    It could be friction in the printer mechanics. Command a long slow movement in the X or the Y direction, and watch the movement to see if it is smooth.

    What are the X and Y steps/mm of your printer?


  • administrators

    I have not seen this before either, except on corners when the acceleration is not tuned well to the printer mechanics. How does printing in different layer heights change it?



  • @dc42:

    It could be friction in the printer mechanics. Command a long slow movement in the X or the Y direction, and watch the movement to see if it is smooth.

    What are the X and Y steps/mm of your printer?

    M92 X43.96 Y43.96 Z426.67 E175.0



  • @T3P3Tony:

    I have not seen this before either, except on corners when the acceleration is not tuned well to the printer mechanics. How does printing in different layer heights change it?

    Printing in higher layer heights is a definite improvement



  • Those pictures were 0.1 layer height.


  • administrators

    @darkstarone:

    @T3P3Tony:

    I have not seen this before either, except on corners when the acceleration is not tuned well to the printer mechanics. How does printing in different layer heights change it?

    Printing in higher layer heights is a definite improvement

    In that case it could be extruder pulsing, which you get with the low steps/mm of ungeared extruders. Divide the filament cross sectional area by the layer height, extrusion width, and extruder steps/mm. That will give you the extrusion distance per extruder microstep. Compare that with the separation of the ridges.

    Another test worth doing on a CoreXY machine is to rotate the cube 45 degrees about the Z axis and see whether the ridges are still there.



  • 43.96 steps/mm is lower than any printer I have ever seen for X/Y axes. Has the printer been modified?



  • That is how it comes from the factory. It uses a spectra string drive setup.



  • @dc42:

    @darkstarone:

    @T3P3Tony:

    I have not seen this before either, except on corners when the acceleration is not tuned well to the printer mechanics. How does printing in different layer heights change it?

    Printing in higher layer heights is a definite improvement

    In that case it could be extruder pulsing, which you get with the low steps/mm of ungeared extruders. Divide the filament cross sectional area by the layer height, extrusion width, and extruder steps/mm. That will give you the extrusion distance per extruder microstep. Compare that with the separation of the ridges.

    Another test worth doing on a CoreXY machine is to rotate the cube 45 degrees about the Z axis and see whether the ridges are still there.

    I printed a 3D benchy rotated at 45 degrees and it does make a massive difference in quality. Given this discovery what does this point the problem to?



  • @Nylkos:

    I might be completely wrong but when I started building my CoreXY, I was told that if I tightened the belt(s) too much I would get lines on the parallel side of the axe that require the most resistance…

    I am only posting this since it's pretty close to what I am reading.

    Thanks for the tip. I could try loosening up the string tension. The principle should be the same as a belt.


Locked
 

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