search

prusa3d / Prusa-Firmware

Firmware for Original Prusa i3 3D printer by PrusaResearch
GNU General Public License v3.0
1.99k stars 1.05k forks source link

[BUG] MK3S (3.13.3) stops mid print and reboots/resets itself #4630

Open N0ciple opened 3 months ago

N0ciple commented 3 months ago

Printer type - MK3S Printer firmware version - 3.13.2 and 3.13.3 (I had the problem with both)

MMU upgrade - N/A MMU upgrade firmware version - N/A

SD card or USB/Octoprint Both SD card and Prusa Link (raspberry py zero W) version 0.7.2

Describe the bug The printer resets/reboots itself during the print and display "Print aborted"

To Reproduce

  1. Download the files thumb_bolt.stl and thumb_nut.stl from here : https://www.printables.com/fr/model/3407-articulating-raspberry-pi-camera-mount-for-prusa-m/files
  2. Import the files in PrusaSlicer 2.7.2 (Flatpak on Ubuntu 23.10)
  3. copy/paste the 2 items so that you have 3 bolts and 3 nuts (works with other number of part but I am sure of this one)
  4. press Ctrl+A and then shift+A to place them automatically
  5. Change the number of perimeters to 4 or 5 (I had the issue with both)
  6. export the Gcode to the SD-Card or with Prusa link and print
  7. After a while the printer stops (I think always at the same spot but I am not 100% sure) and reboot or reset itself (I think). I noticed that the screen displays a loading bar and then the splash screen "Original prusa etc..." with the firmware version, and then the info screen with "print aborted" (or something similar, I dont recall the exact phrasing)

Expected behavior The printer prints all the objects without reseting / rebooting itself

G-code Here are 2 gcodes that causes the problem with Prusa Link, but I tried similar configurations straight from the SD and I had the issue too. bad_gcodes.zip

Video /

gudnimg commented 3 months ago

I think always at the same spot but I am not 100% sure

This sounds very important. How far into the print does this happen? What Z height?

I noticed that the screen displays a loading bar and then the splash screen "Original prusa etc..." with the firmware version, and then the info screen with "print aborted" (or something similar, I dont recall the exact phrasing)

At that point the printer has already rebooted.

N0ciple commented 3 months ago

I think always at the same spot but I am not 100% sure

This sounds very important. How far into the print does this happen? What Z height?

Honestly I don't remember.

But I narrowed down the problem. The following GCode test_gcode.zip works or fails (the printer reboot itself) in the following conditions :

So it seems that having the raspberry pi zero W running prusa link connected causes the issue.

For the record, my RPI Zero W is connected directly to the Einsy Rambo board (not via USB) as per this tutorial : https://help.prusa3d.com/guide/prusalink-and-prusa-connect-setup-mk3-s-_221744

avehome commented 3 months ago

Did you check the flex cable on cracks or damages? I had a small crack in the cable (pi zero to pi camera) that made my Prusa reboot. I replaced the cable and now it's running fine again. If the cable is twisted it can crack at some time... (see example below)

Scherm­afbeelding 2024-03-11 om 22 14 44

TojikCZ commented 3 months ago

I had my raspi soldered poorly. Printer did the same thing. If you want to be sure this is a hardware issue, send me please the syslog file from the settings page in PrusaLink local web interface

Arcadian1977 commented 3 months ago

I've had random rebooting during prints and also when doing an auto home. I was mostly trying to print lithos in pla. I've had lots of issues with this release. 3.13.3 revo version.

TojikCZ commented 3 months ago

Are you willing to check whether this is a hardware issue? (send the log, confirm you are using the pi on the einsy pins, jiggle it a bit to see if printer reboots) if you just pile on while not providing little to no info, we unfortunately cannot just snap our fingers and fix your issue magically. We'd like to, but real world is dissapointing like that ☹️

Arcadian1977 commented 3 months ago

Pi 3b connected to usb port on the board. Primarily printing through prusa connect or prusa slicer.

Arcadian1977 commented 3 months ago

I'm not at the machine at the moment. I'll try to get the log later on.

N0ciple commented 3 months ago

I made a fresh install of PrusaLink on my raspberry pi zero W with another SD Card (I had no reason to suspect the sd card but I had to swap it so I took the opportunity to make a fresh install). I printed the last GCode I sent without the camera connected and it printed without error. I inspected the camera ribbon cable but I did not see any cracks. I'll try to print the same GCode with the camera connected to see if it was the issue. @TojikCZ Do you want to see the syslog files even if there is no error?

GaryAitken commented 2 months ago

I've had similar issues with 3.13.3 and 3.13.1 (backed out 3.13.3 to check). It doesn't always happen at the same place, although several times it seemed like it was close to the same place. Big print covering most of the plate, happened multiple times about 12mm height, then on 3.13.1 at maybe 2.5mm. Stock MK3S from kit, working several years, printing from SD card. Two slightly different .gcode files. Small prints (calibration cube, etc.) seem to work. Using 0.8mm nozzle. However... This only started happening after I had to replace the hotend (stock V6 hotend) a week ago. I've checked all connections and nothing seems amiss. I have another (different) hotend so I may try installing that to see if it makes a difference.

GaryAitken commented 2 months ago

I checked all cables and wiggled them with the power on, with no effect. After doing more research I backed out to 3.12.2 and that solved the problem. I can provide the gcode that fails, but not the model or a .3mf file.

shawneily commented 2 months ago

I am also having the same issue on 3.13.3

Printer: MK3S+ FW Version: 3.13.3 Method of Printing: SD Card, Pronterface

I updated the firmware to 3.13.3 about a week ago which is when the failures started:

Video of failure

Note: I sped the video up after the failure 16x until I came to clear the failure which is why the error messages are erratic.

Troubleshooting steps attempted:

Final Resolution: Flashed back to 3.12.2, was able to print without issues.

No matter the steps that occur that issue still persists in 3.13.3. I did attempt to roll back to 3.13.2 but, experienced the 'ghost layer shift' so I attempted to roll it forward again to 3.13.3.

I experienced I believe about 10 failed prints, might be more I lost count. 80% of the prints were failing around the same time, then the others were also failing at the same time, just later in the gcode.

print-examples

When connected to PrintRun (Pronterface) I was able to see where it failed, and both times I had it connected it failed right after a WIPE was completed.

Here are the lines of failure from PrintRun, along with the associated GCode from each file.

PrintRun - 1:

RECV: ok
SENT: N62822 G1 X153.224 Y133.224 E-.22294*122
*** FIRMWARE CRASH HERE ***
RECV: start
SENT: M105
RECV: echo: 3.13.3-7094

GCODE:

G1 X153.224 Y133.224 E-.22294
;WIPE_END
G1 Z13.2 F720

PrintRun - 2:

RECV: ok
SENT: N14001 G1 X92.12 Y131.992 E-.22268*126
*** FIRMWARE CRASH HERE ***
RECV: start
SENT: N14002 G1 Z2.8 F720*34
RECV: echo: 3.13.3-7094

GCODE:

G1 X92.12 Y131.992 E-.22268
;WIPE_END
G1 Z2.8 F720

gcode file (renamed to .txt to attach, rename to .gcode): test-gcode.txt