Whit's End Layer Shifting Max

[ATC3D]

About to lose my mind.

Have had my Max since March 2026. Early on produced some decent prints. Started developing layer shifting issues primarily on larger prints. Contacted Solvol support (I bought from Ebay so have to use the Ebay support function) and they had me do belt tensioning. Thought it may have made a slight improvement but problems persisted. Discovered that the X-axis rail was kinda rough to the left of center and Support kindly sent me a new rail and bearing. While i had it apart I also did the CN3D gantry mod on Printables, installed new LDO X/Y motors, and installed a brand new OEM toolhead. Also brand new EMMC card with stock Sovol firmware (upgraded to latest available).

FInally get everything back together and layer shifting (if can even call it that- see photos) worse than ever!

While I am not a Printer expert, I am a very technical person and have been painstakingly methodical in my approach to this. I’ve made my share of oopses and redos, and to the best of my knowledge have things as they need to be.

the two horrific pictures show the result with the model printed facing bow to the right. The second set of pictures with the bow pointing forward, and with the exception of the one layer by the bow holes the quality on the latter is acceptable.

Any recommendations greatly appreciated. Will also submit to Ebay Support.

I’m a patient guys and I love this printer, but even I have my limits.

Thanks

[Lion]

If it’s worse after tearing it down, then you need to tear it down again & rebuild…JMO

[cardoc]

Shifts of this type are almost always due to the nozzle hitting the print on a non-extrusion move.

Check for looseness between the nozzle tip and the X rail. With everything cool use your finger to push up on the nozzle.

In your slicer enable Z hop and check “avoid crossing walls”

What filament are you using. Warping of the print as it cools will often cause 1 corner to lift putting it into the path of a nozzle move.

Did you adjust the TMC driver parameters for the new motors? Improper settings can cause lost counts (layer shifts)

[ATC3D]

Check this out-

I noticed the X motor looked out of alignment. Was piping hot. Upon inspection it pulled the mount screws halfway through the ABS mount plate!

When spin the shaft by hand it’s binding up at 180 degrees. Can’t say I checked it before installation but the other identical one spins freely.

Cardoc I do have someone looking at the OEM printer.cfg to see if there are any parameter mismatches TYVM.

[cardoc]

Ouch.

New motor with a bent shaft?

[ATC3D]

So looks like the default config has the motors as 3A which folks are telling me is really high (these new LDOs are rated for 2.5) and likely contributed to the melting. Waiting to hear on whether the seemingly mechanical issue is related/covered.

Regardless, took Lion’s advice and tore the whole thing down again and put it back 100% stock. Only real change was replaced the X-rail and bearing that Sovol covered under warranty.

After tuning it up a bit it seems to be pretty good. I’ll run a couple larger prints and let you know if solved the original layer shifting issue.

[ATC3D]

First off Happy Father’s Day to all the dads out there.

Here is the result of my first meaningful print after calibrating everything the best that I know how to. R2 full dome at 50%.

Reminder the printer is now back to 100% OEM hardware with a fresh firmware load.

I am now truly at my whit’s end. Don’t have the massive layer shifts I had before but this is certainly not acceptable quality. That banding has to be symptomatic of something. As you can see even that seam (it that’s what it is) isn’t consistent.

I will forward these pictures to Support as well for advice. Keep telling me to adjust the belts.

[mich0111]

Hello,
Have you tried changing the seam in OrcaSlicer?
For example, you can choose a random seam.

[ATC3D]

also see below

[Lion]

Some users have found issues with the filament sensor motor, try bypassing the sensor & bowden tube…go straight into the toolhead.
There are some threads about this in the MAX forum.

[ATC3D]

Thanks- will give that a try.

Any thoughts on that bed mesh?

[cardoc]

@ATC3D

Any thoughts on that bed mesh?

Excellent for a bed this size. Does the single point hump show up every scan? Check for a bit of debris between the magnet sheet and the build plate.

[cardoc]

@ATC3D

but this is certainly not acceptable quality

Sovol’s stock configuration is focused on speed benchy results not surface quality.

[printer]
kinematics: corexy
max_velocity: 700            
max_accel: 40000             
max_accel_to_decel: 10000     
max_z_velocity: 10           
max_z_accel: 500             
square_corner_velocity: 5.0  

[lis2dw]
cs_pin: extra_mcu:PB12
spi_software_sclk_pin: extra_mcu:PB13
spi_software_mosi_pin: extra_mcu:PB15
spi_software_miso_pin: extra_mcu:PB14
axes_map: x,z,y

[resonance_tester]
accel_chip: lis2dw
probe_points:
    250, 250, 30  
accel_per_hz:100
min_freq: 35
max_freq: 45
max_smoothing: 0.25
hz_per_sec: 1.0

[input_shaper]
damping_ratio_x: 0.001
damping_ratio_y: 0.001

The damping ratio at 0.001 90% disables input shaping.

[input_shaper] - Configuration reference - Klipper documentation

Set it to 0.1 for both axis. Re-run resonance test and set acceleration to 75% of the recommended value.

The seam is VERY strongly effected by pressure advance. Run Orca’s calibration (I use “tower”) and set the correct value in your filament profile. Then for a object with no corners set seam to Random in Orca.

[cardoc]

@cardoc

probe_points:
    250, 250, 30  

I’d experiment with the gantry higher, say 250,250,200. Logic would indicate that higher gantry = more shake.

[ATC3D]

Lion

8h

Some users have found issues with the filament sensor motor, try bypassing the sensor & bowden tube…go straight into the toolhead.
There are some threads about this in the MAX forum.

I gave this a shot. Did not seem to make any difference.

Thanks for the info!