Strange step loss problem

[CyberBorder]

Printer Model

Voron 2.4

Extruder Type

Clockwork 2

Cooling Type

Stealthburner

For a month now I have been having problems with my Voron 2.4, it loses steps when it prints. Now if I perform a TEST_SPEED at a higher speed and acceleration the printer does not lose steps.
I thought it would be a problem with the printer colliding with the print due to a height problem and I printed without filament and it lost the steps anyway. Adjust all gantry check all straps but still losing steps.
I leave a list below of all the things I have already tried, I also leave my github where I am going to leave the CFG printer and the Orcaslicer profiles that I am using.
I leave a link to a YouTube video where you can see the exact moment where he loses his steps and the characteristic noise (TAC) that he makes. I also put photos of canceled impressions where he loses his steps.

Data from my voron 2.4
350x350mm
Electronics: MKS SKIPR
Canbus: MKS THR36
Hotend: Trianglelab Rapido
Extruder: clockwork 2
Gantry: All metal
Quad Gantry level and Home z: SN-04 inductive sensor

GitHub - CarlosTortonese/MyVoron

Contribute to CarlosTortonese/MyVoron development by creating an account on GitHub.

github.com

At second 10 of the video you can hear the noise he makes when he loses his steps.

 

[CyberBorder]

These are all the things I already tried:

Adjustment of all straps

Gantry belts

Calibrate with Probe_calibrate

Motor current

The engines do not overheat

Gantry Alignment

Update Klipper

Try another slicer

Align belts well and have the same tension

The test_print works perfectly

change layer height

Go to 16 microsteps

Print without filament to see if you lose steps anyway

 

[NoGuru]

I don't know much about the MKS boards but they could be the issue if they cannot handle the speeds. Again, I don't know about them.

Check your belt path, I have seen so many people run the belts incorrectly. Many show some pics of the back of the A/B motors.

When the motors are not energized, does the toolhead move freely all over?

 

[Isue]

Hi! In my case loosing steps is most certainly due to low quality Chinese linear guides. On some range of speeds, the carriage generates noise, and what I imagine balls in it starts blocking the movement. Maybe it’s not your case, but it is worth to try X movement at different speeds and looking for grinding/rattling noise coming from carriage? The same for Y.

 

[Grimshadows]

Sorry if I am necroing this thread, as for rails doing weird binding things at speed it can happen, but the sound that comes from that is horrifyingly loud.

I struggled with this issue for a while trying to understand. Then it hit me in an obscure section of the driver config section in the Voron wiki.

Spread cycle Vs. Stealth Chop

Evaluate your speeds and your accelerations, pick a camp for your X and Y (A and B) motors. While either mode generally works, its better to pick one side of the fence or the other. If silence is your more important than holding torque you go with Stealth Chop, if fast is more important Spread cycle.

The problem occurs when you have the "stealthchop_threshold: 0 - 999999" set to somewhere in between. If the motor is not at zero speed and the driver changes mode it produces errors. SO I had mine set to 200 Which would operate silent below 200 when above that Spread Cycle. Step errors acumilate everytime some errors can even produce 75% of a a full step. Get a few of those and POOF, crazy shifting.

Here is the link, https://www.klipper3d.org/TMC_Drivers.html#setting-spreadcycle-vs-stealthchop-mode

I personally set it to Spread Cycle 100%, my build even at warp speed is almost dead silent due to my choice of isolation and insulation.

Good luck.

Also OP, you are having nozzle crashes with the print too. Mix these two together and poof!