A very, very strange problem that is driving me crazy!

[valsimdim]

Basic Information:

Printer Model: P1: SKCUBE heavily self-modified; P2: Self made Voron Trident like
MCU / P1: Printerboard: Octopus Pro; P2: BTT Manta M5P
Host / SBC: P1: Rpi 3B+; P2:CB1 EMMC+EBB42 Canbus
klippy.log (8.6 MB)

klippy.log

Describe your issue:

I build these 2 printers about 3-4 years ago and was happily printing since. Not too often, just when I needed. All this time I was using Orca slicer exclusively. After 2-3 months pause, a week ago I started a print (on Printer 1 - SKCUBE) of a box enclosure about 140x140x200mm, 350 grams, 10 hours print. After about an hour It failed with error “Move out of range” with some weird value Y1593.30256. I checked all printer parameters both in Orca and printer config file, re-sliced and started again, it failed again with a different error “Cannot parse move”. Resliced, printed again… “move out of range” again…, resliced then extrusion failed, and so on. Sometimes after 30 minutes, an hour, 2 hours, never made it pass 2.5 hours.

Then I started troubleshooting, cloned and changed the SD card, updated everything to the latest, OS, Klipper, Mainsail, re-flashed firmware on the MCU… nothing helped, random movement errors and printer just stops printing. Just to clarify, before this I had a print success rate of about 95%. Then I gave up on this printer and switched on Printer 2. Same issues, same random errors. Next step I decided to analyze the .gcode file, and I found weird commands. I blamed the slicer. Installed fresh Orca last version on a different PC and created and configured new profiles, same issue. Then I got pissed off and yesterday spent 200 USD buying Simplify3D (I know, I know, don’t say anything). Same! Then I switched back to printer 1 and decided to dig even deeper into the gcode file. I generated the file, save it on my PC, then uploaded it to the printer, then without starting a print downloaded it back from the printer with different name and loaded both the original and downloaded files in Notepad++ and used the Compare plugin.

And the files were different! The movement commands on the local file were looking fine, and the file that was uploaded and then downloaded back from the printer had strange differences (more than 1) with invalid movement values. I did this many times comparing the same original file against the re-downloaded files and they were always different at random places. I even tried copying the original file to different folders and disks on my PC (without going to the printer and back) and the Compare in Notepad in this case is always showing “Files are identical”. So definitely somehow the file was changing between the PC and the printer and back to the PC

I was always using WI-FI to connect my 3d printers to the network and LAN for the PC, so next step was to switch off the Wi-Fi AP and connect the printer’s RPi by cable to the same switch where my PC is connected. And I saw difference and improvement. The re-downloaded file was still slightly different, but only a few rows and nothing weird and “out of range”. I still wasn’t satisfied, changed the network switch, sliced, uploaded, re-downloaded, compared and still a few rows were always different but with no noticeable invalid values.

At this point I just started another print with the “slightly different” but looking better gcode file about 110 minutes ago and it’s still printing. I couldn’t just look how the printer is working, waiting for it to fail, so I decided in the meantime to share my troubles with the Klipper community and eventually get some advice and clarity. So it’s time to ask some questions:

1. Did anyone else experienced similar issues?
2. Does the gcode file gets always modified ones uploaded to the printer, I thought it’s supposed to be 'read-only"?
3. Is someone willing to test this and confirm if the same thing happens? Slice, safe the gcode file on a PC, upload to printer, download back to PC and compare the original with the ‘traveled’ file in Notepad++
4. I will highly appreciate any help, hint, advice or really anything that can lead to resolving this issue.

For anyone who found this topic interesting, I’ll be back with status update, the print is supposed to finish in about 8 hours from now. Thank you for reading my novel to the end and happy printing!

[EddyMI3D]

Hello @valsimdim !

The two errors you mentioned in your very very long text do not appear in the klippy.log.

But I found:

Unknown command:"T0"

[valsimdim]

Hello, @ EddyMI3D!
Yes, the Unknown command:“T0” is from the Simplify3D slicer, it doesn’t harm the print. And the klippy.log doesn’t show errors probably because after the last fail I restarted the printer, and this log file is from after it was already printing for 90 minutes.

Update, so far it didn’t fail, 5.5 hours already, more than halfway. I think connecting by cable and stopping the WIFI solved the issue, still curious what is wrong with my WIFI, all other devices (and I have a lot) looks to be working fine. And still curious how is this even possible, a few rows in the gcode file to change during transfer from the PC to the RPi. The file is not corrupted, it opens normally, its just different, I’ll attach a screenshot below:

The difference starts at row 153,526 and you can see the 'Compare nav bar" circled in red the difference section. This is the first almost good file, it already passed the section in red, so I’m hopeful it will print to the end, but why and how is this even possible???

[valsimdim]

Ok, print finished, and another one also, no errors, gcode files now uploading and not changing. All I had to do is connect the Pi with LAN cable and disconnect the WIFI. When I have time, I’ll test the same with my second printer and update here, but I’m pretty sure it will be fine.

Hopefully someone with the same problem will find and read my novel and save some time and plastic.

Happy printing everyone!