I also had some layer shifts on the x axis at the beginning.
I mitigated these by choosing two linear modules with the same backlash as y axis and limiting the maximum acceleration for all axis via the following startup Gcode:
M201 X1000 Y1000 Z100 E10000 ; sets maximum accelerations, mm/sec^2
But that may be totally unrelated to your issue…
As it’s not my printer I don’t have immediate access to do in depth troubleshooting in the code and console. I just like helping and figuring stuff out if I can. But as the rails have been swapped out so many times with absolutely no change it’s become quite perplexing. He is also in contact with Tracy in support. Bigger items tend to fail more. Cube printed perfectly multiple times in a row, until I scaled it up to 150% and failed first try.
With whatever the simplest reproducible object is (probably the cube) does reducing the flow significantly prior to starting change anything? M221 S80 after it starting printing will reduce the flow to 80% of nominal.
Wondering if that can conclusively rule out overextrusion issues.
Also, I can’t find above if you’ve check at end of a print if the step counts are the same with M114, for both successful and failed prints. Does the printer think it’s lost steps or is the printer unaware? Would also help identify if it’s firmware or hardware.
@brent113 I’ll have him run those commands. As far as hardware goes… the x rail has been swapped out with every other one, nothing was over heating, platform was confirmed to be installed correctly.
This machine is only a week old and has been calibrated so many times that a brand new user knows it as a second language. Before I joined in on the troubleshooting it had been completely disassembled and reassembled 3 times with no change.
Firmware… it has happened on 1.8, 1.9, 1.10.1, the first time updating the firmware it froze during update, second attempt was flawless. He then tried updating to the next and it froze again, reverted back to 1.10.1 and it updated flawlessly.
Luban was version 3.14 at the beginning of this, now it is the latest with no change. Zhop, retraction on both layer change and travel were enabled. Still failed. We’ve also noticed it fails far more frequently the taller a print is and only starts to fail when it gets to about an half-inch high with very rare failures starting lower than an that, first layers have always been really good. Calibration cube printed perfectly 4 times in a row. Scaled it up 150% and it failed first time with the same settings.
When I inspected the lamp model in multiple design software, it had a lot of problems so I thought that was a possibility, but after repairing it and confirming no errors in all design software, it still failed the same way, not to mention all the other models that also failed and had no errors in their design. As far as extrusion, I haven’t seen any major signs of it but would still be a good idea to completely rule it out.
The project has been finished. But no layers are shifted.
We are going to send you a beta version of firmware with which you can use the add-on-1 port on the controller as the X port.
Edwin
K cool. I didn’t want to bring up the converter possibility until someone from Snapmaker did since I don’t have immediate access to the machine and didn’t want a miscommunication causing a bigger problem. Thank you I will forward this on to him. He’s not a member of the forum but I have had him looking at this post as well. Really should tell him to sign up so he can join in ha.
Hi I am here. I want to thank Artezio on so many levels. To be clear everything he has posted on my behalf is 100% accurate.
please let me know where I can get the firmware file. I will update and start as soon as I know
I think the calibration cube will helps to see if any layers shift during printing.
Attach the calibration cube gcode file here.
cube-test.gcode (1.1 MB)
is that the normal sized one. I printed four of those without issue. So not sure what that will help. The one with the faults was sized up to 150%.
printing the gcode as you sent it
Well Edwin that was an epic fail. The printer head never left the home position. the heated bed was moving but the head was just shooting out in the start position. So to be clear the x module has zero movement left to right. Including calibration and printing. Please advise further.
@Deth I’m just confirming/verifying, did you unplug the cable for the x-axis on the controller and move it to “add-on 1”? The special firmware was to move the control of the x-axis from the designated “x” port to the “add-on 1” port instead.
Yes that is exactly what I did. I can’t upload video because I am a new user 
. But I flashed the system with the beta firmware you posted then I shut machine off and switched the cable like you have in the diagram. I turned the machine back on and tried to calibrate the machine. The print head went to home position the bed moved to home position and then during calibration the x axis and print head only moved up and down. The bad was moving as normal but the print head did not move once in home position on the x axis. So I tired to print the cube gcode you sent me and the print head did not move from home position. It did start to push the pla through but not on the bed. As the print head was again in the home position to the left of the bed.
@Deth Is referring to this video.
Sorry for the problem. I have tested the beta firmware with the USB stick. Users cannot update the firmware successfully and the add-on port does not work.
As for the controller pack of the firmware, users need to upgrade it with the computer and the USB cable.
Edwin
K I’ve forwarded this on to him. We will update with the results probably tomorrow. Sorry you are spread out between the posts.
One question was not cleared, what slicer did he use and what speeds did he print?
Interesting issue, hope you will catch the defective device soon.
Good question. I did not think of it. I printed this model, but it went well.
@xchrisd @Edwin Everything has been done in Luban 3.14 and up. Originally speeds were set at default fast print profile, slowed it down to normal print profile speeds. Default speed values were used, nothing was modified in that respect. Really the only thing modified from default was enabling z-hop, retraction at travel, retraction at layer change.
