Results 1 to 10 of 10
  1. #1
    Jailcee's Avatar
    Title
    Administrator
    Join Date
    Apr 2017
    Location
    Canada
    Posts
    80

    Reducing ringing / ghosting in the prints

    The CR-10 is prone to ringing / ghosting when printing, especially at higher speeds (faster than 40mm/s). The heavy bed creates the ghosting along on the X axis faces/features and the print head creates creates the ghosting along the y axis face/features. I recently had the chance to pick Thomas Sanladerer's brain on this and he explained it well.

    What happens in ringing is the part of the printer is moving in a direction, say along the X axis. As you can see in case A on the left side of the image the mass of the printer (black square) is moving along X and suddenly stops. When it stops it bounces back and forth along the X axis as the drive belt acts like a spring. Imagine the drive belt was a rubber band and you can see there would be lots of bouncing. The same thing happens, only on a smaller scale.



    In case B on the right side of the image as soon as the mass (black square) gets to the end of its X move it immediately starts moving along the Y axis. The bouncing is still there except now because the mass is moving along the y axis while it is bouncing on the X axis the path it takes isn't straight but instead draws out a wave. This is the source of the echo or ghosting we see on prints.

    Here are two prints I did with the left one at 40mm/s and the right one at 60mm/s. Both have the same jerk and acceleration settings. You can see the ghosting in both however in the right one it is more pronounced because the movement was faster and spread that bouncing motion over a larger area.



    The stock CR-10 see's this mostly on faces facing the X axis since the mass of the print bed induces more bouncing. In my case I see this effect strongly on both axis since I modified my printer with a direct drive volcano print head.

    One strategy to reduce this effect is to print slower and with lower acceleration and jerk settings (800 accel, 8 jerk). This works but I want to print FAST so I need to do more work. I have ordered new high quality Gates 2GT belts that I will use to replace the stock ones. Once those come in I'll be putting them on the printer and running some tests.

    Stay tuned!

  2. #2

    Title
    Junior Member
    Join Date
    May 2017
    Posts
    1
    Very interesting. Have you noticed a difference in ringing depending in the tension of the belts?

  3. #3
    Jailcee's Avatar
    Title
    Administrator
    Join Date
    Apr 2017
    Location
    Canada
    Posts
    80
    Quote Originally Posted by Elac View Post
    Very interesting. Have you noticed a difference in ringing depending in the tension of the belts?
    I haven't experimented with that but it would make sense that as tension in the belt is increased the frequency of the vibration increases as well. The relationship should be proportional to the square root of the tension as well as how much belt is behind the mass. For example as the hot end moves to the right the length of belt between the hot end and the left side of the gantry increases. When the hot end stops moving to the right that's the length of belt that will be ringing and the longer it is, the slower the vibration. This would be independent of the tension. I suspect the vibration would follow this formula where n is the harmonic, L is the length of belt, T is tension, and u is the density factor of the belt.

  4. #4

    Title
    Junior Member
    Join Date
    Jun 2017
    Posts
    2
    Have you tried it out yet? I've been getting really bad ghosting.

  5. #5

    Title
    Junior Member
    Join Date
    May 2017
    Posts
    11
    i found upgrading the printed Y tensioner and swapping to steel belted gt2 helped (not totally fixed but i'm down to 1-2 layers of ringing as opposed to several) , but later found my tie offs have been slipping so i've been out of action completely the last few

  6. #6

    Title
    Junior Member
    Join Date
    Jun 2017
    Posts
    2
    Quote Originally Posted by Ralgon View Post
    i found upgrading the printed Y tensioner and swapping to steel belted gt2 helped (not totally fixed but i'm down to 1-2 layers of ringing as opposed to several) , but later found my tie offs have been slipping so i've been out of action completely the last few
    What speed were you printing with your gt2 belt?

  7. #7

    Title
    Junior Member
    Join Date
    May 2017
    Posts
    11
    Quote Originally Posted by cotang View Post
    What speed were you printing with your gt2 belt?
    mostly in the 60-80mm/s .16+ layer with petg on the xyz 20mm cube for the comparisons.

    60mm/s with esun PA-CF had no ghosting at all but as i don't have a comparison that could be more the properties of the filament rather than the upgrades.

  8. #8

    Title
    Junior Member
    Join Date
    Jun 2017
    Posts
    2
    The y axis motor brace helped with ringing on my S4, but only at 45mm/s or slower. It made no difference at 60mm/s. I do not have upgraded belts, but do have the y axis tensioner. I had photos and input in the FB group.

    eta: my accel is 1500, jerk is 15. I run marlin 1.1
    Last edited by RobbyMaQ; 07-04-2017 at 04:26 AM.

  9. #9
    Jailcee's Avatar
    Title
    Administrator
    Join Date
    Apr 2017
    Location
    Canada
    Posts
    80
    Quote Originally Posted by RobbyMaQ View Post
    The y axis motor brace helped with ringing on my S4, but only at 45mm/s or slower. It made no difference at 60mm/s. I do not have upgraded belts, but do have the y axis tensioner. I had photos and input in the FB group.

    eta: my accel is 1500, jerk is 15. I run marlin 1.1
    So you know Marlin 1.1 expects the jerk set command to specify both X and Y values. It used to take just X and set X and Y to that so that:
    M205 X15 would set your X and Y jerk to 15 but now with Marlin 1.1 it will only set your X to 15. To set both X and Y it needs:
    M205 X15 Y15

    I don't think slicers have adjusted for this yet. I found this error and it is fixed and will be in Cura 2.7. Don't know if the other slicers are correcting it.

    So what does your Y jerk end up being? Whatever the default is in the firmware which is, unless you changed it before compiling, 20.

    In other words your printer is running with a Y jerk of 20 regardless what you put in your slicer. The only way I know right now to change that is to change the max jerk setting to something lower than 20 (like 15) in the EEPROM.

    P.S.
    I discovered all this when I noticed that some settings were improving the ringing on one face of a cube but not the other. I went deep into the firmware to find the cause and this was it.

  10. #10

    Title
    Junior Member
    Join Date
    Apr 2018
    Posts
    1
    Jailcee,

    Digging up an old thread I know, just curious, did your GT2 belt upgrades do the trick or was further adjustment needed?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •