What is the problem?
Recently set up a CR-10 with octoprint.
I've done a few shorter (3-4h) prints which has worked great.
Then I've tried a 24h print twice, reslicing the file in between with other settings.
Both times the print has randomly stopped after >7 hours.
The log has said "WARNING - Printer sent 'start' while printing. External reset?"
What did you already try to solve it?
After the first failed print I disconnected the LCD/button combo since it would sometimes randomly beep like someone clicked it, and enter menus. So I figured this could be the culprit.
I removed the cable so it can't act like an antenna or something.
Despite this, the second try also failed in the same way.
Have you tried running in safe mode?
No
Did running in safe mode solve the problem?
n/a
Systeminfo Bundle
Attached.
octoprint-systeminfo...zip (20.9 KB)
octoprint.log (180.9 KB)
Additional information about your setup
OctoPrint version: 1.10.2
OctoPi version: Build 2024.06.18.085851 with "webcamd", based on OctoPi 1.0.0, running on Raspberry Pi 3 Model B Plus Rev 1.3
Printer: Cr-10 modified with inductive probe and direct drive.
Firmware: Marlin 1.1.9 but modified to work with above.
Printer has previously worked flawlessly printing with SD card.
PSU measures 12.27 to 12.35V, can't see any voltage drop/instability.
Raspberry runs off a 12V to 5V converter which is sitting steadily at 5.07V, fed from the printers 12V PSU. Capable of 2.5A and has extra forced cooling so should work.
Any help is much appreciated, feeling kinda lost as what to troubleshoot/do!
Hello @Mr.Crankyface !
Please upload your files here and not to (another) file server.
And BTW: The systeminfo bundle contains all logs - that's why it is called bundle.
1 Like
Hi! Sorry didn't see the upload icon, not the first time I'm bad at looking. 
Fixed my initial post and think I managed to upload the bundle to the forum.
1 Like
Have you extracted some files from the bundle?
The octoprint.log is missing. It belongs in that bundle.
1 Like
I haven't knowingly changed anything with the bundle, it is as I downloaded it from octoprint.
I also attached the log that I know was from the latest "incident".
I cleared the logs after I had saved down the octoprint.log, I must've saved the bundle after that.
I wasn't sure what the bundle was, just following instructions of what the "new thread" said.
I realize now that it was stupid of me.
The printer is running at the moment so I'm hoping it will reproduce the error again for more data.
I enabled some extra logging stuff I found in the settings that could possibly give more info.
2024-10-08 20:56:37,840 - octoprint.util.comm - WARNING - Printer sent 'start' while printing. External reset? Aborting job since printer lost state.
For this issue is generated by the printer - and this is a Crealty, you may try
1 Like
I taped the 5V pin before as I saw that was recommended for creality in some thread, should've mentioned that earlier. So many details that I forget. 
Will test with a sd card and report back!
Seems to take on average ~7h to make the error show up so it's not a fast process.
It printed without filament no issue for 20 hours at which point I cancelled the print, thinking maybe removing all plugins had an affect.
Started the print with filament which worked for 5 hours at which point my filament holder decided to disintegrate forcing me to cancel print.
Started again with repaired filament holder and got around 1h into it when it had the initial error, external reset etc.
Attached the new bundle that should have catched this event.
octoprint-systeminfo-20241009201953.zip (131.5 KB)
I have now put in a SD card and restarted the print to see if it makes a difference.
Unfortunately SD-card did not help.
Same issue after some hour of printing.
octoprint-systeminfo-20241009231724.zip (267.8 KB)
Time to try some more troubleshooting I guess...
I'm still worried about the PSU not being good for some reason, I don't have expensive logging equipment but I think an arduino can do the job.
Initially using a 16x2 LCD.
This quickly showed issues of data corruption both when hooked up to to the 12V line of the printer(for measuring/logging, not powering arduino) and when disconnected.
It ran fine when I moved it further away from the printer setup which makes me highly suspect EMI issues.
Added a capacitor across VCC and GND which helped slightly when not connected to the 12V line, once connected it didn't help anymore.
Those LCD's are semi-known for being sensitive so rewrote the program to work over serial instead and directly monitor from my laptop.
All it does is measure as often and fast as possible and then save/show the min, max and current values once a sec.
ie the values refresh at the max speed arduino can handle(10k per sec or something) but I refresh the serial less often to not spam the terminal..
Will do another testprint and see if it can catch any voltage drops.
I still need to investigate the EMI issues as that could explain both the printing issues and why the cr10 lcd kept beeping as if someone was clicking the button every now and then, if it was also affected by the EMI.
edit: I've also stopped using the 12V to 5V converter for the pi, running off the creality power supply. Considering it's price and lack of case I suspect it's creating more EMI.
Using a 5V 3A phone charger atm, outputting 4.9V fairly consistently.
Phone charger wont cut it. My Pi 3 prefers 5.2V at 2.7A and over depending on what other USB devices I have plugged in. Get a Pi Power Supply...
2 Likes
Yes, long term I agree that's the better option. I am however currently broke and since this seems to let the pi run with only a 0.1V drop it will have to do for testing.
It is currently only hooked up to the printer to make sure there's no accessories pulling extra power.
It could cause new issues but previous issue has been that the printer itself seems to shut down, which this is good enough to test with.
Maybe celebrating early but I think I've found the fault!
To clarify for future reference:
I was using a cheap 12V to 5V 3A buck module hooked to CR-10's 12V power supply, this powered the raspberry.
Since I suspected EMI issues I removed the buck module and the printer has now successfully run nonstop for 30 hours.
For me that's enough proof that the buck module was causing interference, either causing the printer to reboot or octoprint to think that it had rebooted.
Last thing todo is to replace the 5V supply with a proper one.
Will update if the issue comes back but hopefully it's solved.
1 Like
