Help: Klipper crashes mid-print

General Discussion
21

In this case, your SBC crashed hard:

  • Check for potential overheating
  • Check the Power Supply / use a properly rated one with good cables
  • Remove all additional hardware
  • Replace the SD card

Very unlikely that it is caused by Klipper.

22

  • Check for potential overheating

    • I have had it crash while I was watching it several times and temperatures were steady. The MCU at 40C, the Host/SBC around mid 60C

  • Check the Power Supply / use a properly rated one with good cable

    • I have tried two different power supplies and I have a third arriving today

  • Remove all additional hardware

    • Right now it is down to no additional hardware. I had previously tried it with different combinations of attachments.

  • Replace the SD card

    • I have a new SD card arriving today as well. As well as a new rPi 3b+. If after replacing all of that, I guess trying a new SKR Mini e3 would be just about the last thing to replace. I have replaced the USB cable between the rPi & SKR a few times already as well.
23

Is there any passive or active cooling for the SBC? Keep in mind that the USB chip on a Pi3 can get way hotter.

BTW: On my constantly running RasPis (4) the temperature is between 35°C and 40°C.

24

There is not currently. Ironically, the print that I have been testing with is a case for the pi that has a fan! I’ll take it out of the current case and put a fan on it.

25

Don’t forget to attach heat sinks to the chips.

26

I had high hopes for temperature being the issue. I took it out of the case and put a fan on it. Host temperatures reported in Mainsail dropped from mid 60sC to mid 40s! But it just crashed again. Next step is to replace the rPi.

27

As users sometimes have issues like you are facing, we put together some background information.

See Troubleshooting Spontaneous SBC Reboots and Crashes in Klipper

28

I replaced the SD card, no luck. I replaced the rPi, no luck. This is turning into a ship of theseus situation.

The only things left to replace are the SKR Mini E3 and the power supply that powers it.

29

So you power the RasPi by the printer?

30

Separately. I have a 5v 3A supply for the rPi (actually I have tried two different supplies already; and I have a 5.2v 3A supply arriving today). I have a separate PC PSU powering just the SKR Mini E3

31

It really seems to me that you have something fundamentally wrong here.

Rather than continually trying different things in hopes they’ll fix it could you:

  1. Give a complete history of the problem in relation to the printer. When I look at the OP, it sounds like you took an existing commercial printer and modified it - did the problems start from here or did they come later? If they came later, what kind of changes happened with the printer (was it moved, sitting for a long period of time unused, did you do an update, etc.)?
  2. Provide an accurate and complete wiring diagram. A scribbled drawing on a notepad isn’t good enough. Make sure you mark wiring polarities and, ideally, the gauge/thickness of the wires that you are using. Include any peripheral devices.
  3. Post the latest klippy.log complete and unedited.

With full and accurate information, I’m sure that the source of the problem will become apparent.

32

  1. The printer is a diy Hypercube Evolution. It had been running well for a number of years. I recently reworked the xy gantry to use linear rails and switched to an Eva 3 tool head. At that time I also switched from Marlin running on rails/Mega 256 to running Klipper on a rPi 3b+ host & SKR mini e3 mcu. It has been about one month since I made these changes. I think that the issue has likely been happening since I first switched to Klipper. It sometimes take several hours to exhibit the problem and until recently I had only attempted shorter prints and calibration.

  2. There are now no periphery devices. The circuit is only the power supply → SKR Mini and a separate power supply → rPi with a USB connection between the two. And of course the necessary thermistors, steppers, hot end, fans, etc. There is nothing unusual about the printer hardware itself. If I find the time I will draw up a more official wiring diagram.

  3. Attached

klippy_2025_08_04.log (2.8 MB)

33

I also wanted to add my thanks for sticking with me through this! It’s been a frustrating distraction for too long now.

34

(forgot to add: There are two crashes in the klippy.log. The first crash in the klippy.log was after changing the SD card. The second was after changing the rPi)

35

You’re still running the rPi 3B+?

Would you consider a Raspberry Pi 4B or 5B instead?

This is just throwing an idea out there as the 3B is generally considered the minimum to run Klipper on.

When I look at the performance differences between the 3B+ and the 4, there is a pretty big difference:

36

I don’t think I’m even close to taxing the 3b+ processor capabilities. Except for during the stress test it doesn’t seem to go above few percentage of usage.

I’m running out of things to change. Since my last post I have changed the mcu power supply, and the SKR Mini e3 v3 (and updated Klipper to the latest version). At this point none of the klipper → print controller hardware or software is the same as I started with and the problem remains.

Someone on reddit is having the same exact issue with the same exact hardware and speculates that the issue could be a klipper + SKR mini e3 v3 issue (they have 3 printers and only one has the issue and the other two are on lower versions of the SKR Mini e3)

Is it possible for something in the printer electronics/hardware (i.e. thermistors, steppers, hotend heater, etc.) to cause a hard crash of klipper and the printer?

37

Personally, I’m loath to blame electronics when there are so many other variables. I used a Mini E3 V3.1 for a couple of years on Klipper without any problems at all. I ran various hosts on them including a rPi 4B, rPi 5B and CB Pi 1.2.

I’m suggesting the rPi 4B over the 3 series because it does have significantly more CPU horsepower and, I beleive, the OS base is more actively maintained and tested.

I just took a look at your latest klippy.log and I noticed that the MCU temperature never goes beyond 40.7 and each time there’s a reset the temperature is higher than 40C.

image
image1977×396 29.4 KB

How are you cooling your Mini E3 V3? I’m wondering if you have a temperature related issue here.

Is the board located inside your printer? Does it have any kind of active cooling (ie a fan)?

If it is located inside your printer, could you mount it outside the printer and try aiming a fan at it?

38

As mentioned before and also put forth in the KB article: If your whole system becomes unresponsive, then it is not Klipper that is crashing but your operating system / SBC. A Klipper crash is somewhere between extremely unlikely and impossible to kill the entire OS, as it is a pure user-land application running under restricted permissions.

For troubleshooting purposes, it is extremely important to be precise about what is happening. A crash is not a crash.

39

The Mini e3 v3 is cooled by a fan which is powered by the mini e3 v3 itself and configured as a heater_fan so it turns only during a print which is probably why it is cooler during the print.

This is a diy corexy printer (Hypercube Evolution). There is no enclosure and the Mini e3 is attached to the side of the printer but is not in a case.

I also have a fan on the pi (currently just a portable house fan blowing directly on it) but I will try something more “local”. That same house fan is blowing on both the pi and skr (they are currently attached by the short USB cable provided with the SKR).

40

I don’t doubt that this isn’t a klipper specific issue.

”For troubleshooting purposes, it is extremely important to be precise about what is happening. A crash is not a crash.”

How do I describe the crash more specifically? What additional information can I provide that will let you distinguish what is happening from other crashes?