Cumulative layer shifting on a Voron 2.4 r2

[Mykola]

Hi everyone!
I’m having an issue with my Voron 2.4 R2 (350x350x350) – I’m seeing cumulative layer shifts during long prints (the longer the print, the more the shift is visible on the X and Y axes).

Hardware setup:

• Mainboard: BIGTREETECH Manta M8P V2.0 with CM4
• Toolhead board: EBB SB2209 (CAN)
• Drivers: TMC2209 in UART
• Input shaper configured using ADXL345
• Z-probe: Voron TAP

What I’ve tried so far:

• Belt tension verified and even
• Motors and pulleys are secure
• I swapped X/Y motors and drivers – no improvement
• No overheating detected

Could the problem be caused by the mainboard, CAN communication, or possibly power issues?
Any help or suggestions would be highly appreciated!

 

[NoGuru]

That is not layer shifting, that is skew. Look into Klipper skew compisation or try and square your frame.

Skew correction - Klipper documentation

www.klipper3d.org

 

[Mykola]

That is not layer shifting, that is skew. Look into Klipper skew compisation or try and square your frame.

Skew correction - Klipper documentation

www.klipper3d.org

The diagonals of all sides of the cube are as equal as possible +/- 0.2 mm

 

[NoGuru]

Humm, if your sure its not skew and is in fact layer shifts.
Is the config for the motors correct?
What speed and accecls are you running?
Current of drivers?

 

[Mykola]

---

The current is set to 0.9 A, speed is 150 mm/s, and acceleration is 5000 mm/s².
I’m using TMC2209 drivers.

 

[NoGuru]

Anyway to get a fan on the drivers?
What about the tension in the Z belts? Could there be some wobble in the frame or gantry?

 

[Bambel]

Hi, did you find a solution? I recently noticed that a lot of screws inside my 2.4 got somewhat lose after 400h of printing and since i retightened the screws, i got a similar problem. I didn't change anything else but from then on the printer didn't work. Already did a lot of things: re-belting 3 times, swapping 2209s, checked every idler/bearing stack, squared the gantry, ... Everything feels smooth and still no improvement.

b.

 

[OkieVet]

Hi, did you find a solution? I recently noticed that a lot of screws inside my 2.4 got somewhat lose after 400h of printing and since i retightened the screws, i got a similar problem. I didn't change anything else but from then on the printer didn't work. Already did a lot of things: re-belting 3 times, swapping 2209s, checked every idler/bearing stack, squared the gantry, ... Everything feels smooth and still no improvement.

b.

Not intending to sound like a smartalec but I gotta ask, did you use a thread locker like Lock-Tite or Permatex on all your T-nuts for at least the frame and gantry? I used the blue stuff for things that might need to be re-adjusted later - can be removed without heat, and red for more permanent things. I actually used the red for the hammer-head nuts and panel clips so they act like 1/4 turn fasteners now.

 

[Bambel]

Not intending to sound like a smartalec but I gotta ask, did you use a thread locker like Lock-Tite or Permatex on all your T-nuts for at least the frame and gantry? I used the blue stuff for things that might need to be re-adjusted later - can be removed without heat, and red for more permanent things. I actually used the red for the hammer-head nuts and panel clips so they act like 1/4 turn fasteners now.

It was in fact something completely different: Kalico!

So after i tried further things on the hardware i had the desperate idea that it could be software related. I then converted back to mainline Klipper (host and MCUs), and it worked fine! As an experiment, i ran Kalico as the host and left Klipper on the MCUs and aside from a warning that the MCUs should also use Kalico, it also worked fine. Finally also flashed the MCUs with Kalico and the skew was back. Already filed a bug report.

I was originally investigating an extruder problem and found the many lose screws accidentally. During all this i made an update of Kalico and didn't notice the skew because no print went far enough to show the skew. So i falsely attributed the skew to hardware.

BTW: i only used Loctite on all the grub screws but no other ones. Loctite seems to dissolve ASA so should not be used on screws that touch plastic. I could use it on the blind joints and for the linear rails but funny enough, these were the only screws that were still very tight.

I have setup a reminder in 100h printing and will check all screws then..

B.