Fill out above information andin all cases attach yourklippy.logfile (use zip to compress it, if too big). Pasting yourprinter.cfgis not needed Be sure to check our “Knowledge Base” Category first. Most relevant items, e.g. error messages, are covered there
Describe your issue:
Hi everyone, I’m running into an issue where my printer (display freezes) and Raspberry Pi freeze shortly after a print starts. Usually, the web UI loses connection and shows: “Cannot connect to Moonraker (192.168.x.xxx)”. And i cannot ssh into the pi. i have to power off and on by pulling the power, as there is no way to issue any commands as i know atleast. Im using the powersupply that came with the pi, 3A 5V via the micro USB connection on the pi
The Symptoms:
The printer completes the start g-code (probing/heating) and begins the first layer.
It freezes randomly—sometimes at layer 1, sometimes at layer 25.
When it freezes, the entire Pi seems unresponsive via the web interface and ssh. The printer also becomes unresponsive.
What I’ve tried so far:
Swapped between 3 different SD cards (Transcend 32GB Class 10/A1).
Performed fresh installs of Klipper/Mainsail.
Searched logs for MCU or Shutdown errors but couldn’t find a clear “smoking gun.”
Tested with and without a webcam active in the UI (also tried dissconnecting it physically from the board). This time it got to about layer 4. (Latest print when uploading the logs)
Checked for the “common klipper error messages” in klippy.log.
Checked other users similar problems, but i dont get their error messages in my logs as they seem to do, unless i miss it when searching for them.
I’ve attached my klippy.log , moonraker.log below. Since I’m having trouble interpreting them, I’d be very grateful if someone could take a look and see if there are any underlying communication or power issues!
My situation most closely resembles the problem solved in this thread:
As a result, I have ordered a replacement power supply and a new USB cable. I plan to test both, and it’s always useful to have a few extra cables on hand.
It didnt work, it ended up so that mainsail could not get connection at all with my printer at all. Stuck at “prusa startup screen” with the blower fan going full speed.
As soon as i removed the tape, plugged back in, the printer got connection and the fan winded down again and it looks normal.
So i cannot even try if the voltage on the board is what is messing me up.
On RepRap about EinsyRambo under “powering” it says:
“Powering
Power must be connected to the Primary Power input for the board to operate. The USB connection is isolated from the rest of the board and cannot be a power source.”
I dont know if that is relevant, as i dont know if prusa has modified it and made it more modern or something. Im to stupid to know these stuffs.
I do, yes. I used the 3B before, with the fan, and I thought that cooling might be an issue, so I tried the 3B+ which has some improved cooling capabilities and added the fan from the 3B.
Hi Fillipooo, have you been able to fix this issue? it appears to be the same I my issue. The RPI3B+ is freezing during the first or second layer. I have rebuilt and reflashed everything but the issue persists. I have also tried a different RPI 3B+ and stressed tested the RPI and it passed / didnt crash.
I about to try a RPI3B as was using 3B until a couple of weeks ago to see if it fixies the issue.
“Temps rarely go over 40°C. It seems to be hovering around 39°C during printing.”
I thought I remembered temps as high as 39°C, and I was semi-right it turns out.
As I was testing, it ran for 32 minutes and I saw stable 36–37°C temperatures.
I wanted to video it, I guess for my own sake—if anyone can come up with something else than a USB cable and power cable, as I’m itching to print with Klipper :D.
You can see in the video that when it crashes/freezes, the RPi reports 36.5°C, but when it boots up again it reports 39.7°C — kinda weird.
This video is of my latest attempt as the crash happends. I sped up the entire printing and normal speed when the “crash” occurs, you can hear it.
It got further than last time. This video is from when I pulled the logs as soon as Mainsail h
ad restarted.
Video: https://youtube.com/shorts/IP33FcFjg30 (rlly bad quality, but i hope its enough to get the idea)
It is not solved and I’m still waiting for two different high-quality USB cables and a, hopefully, high-quality power supply, as I went with an original Pi power supply for the 3B+.
The Einsy Rambo, and several other boards made by Ultimachine, have an electrical isolator chip between the USB port circuitry and the rest of the board. This is why it wasn’t recognized when you put tape on the 5V pin, and is also why that tape is not necessary.
The USB communication components are powered by the 5V on the USB cable. The rest of the board is powered by the main power input screw terminals. Data goes through the isolator chip to the main MCU. This prevents ground loops and prevents the computer or Raspberry Pi from powering the board when the printer is turned off.
So when you put tape on the 5V pin, the USB components on the board had no power.
I replaced the power supply and tried two different USB cables (including a 90-degree version), but the issue persists. Even though the hardware wasn’t the fix, the old power supply was low-quality and needed an upgrade regardless.
That leaves the SD cards. I’ve tried three so far, but there is a chance they all come from the same batch, they might share a manufacturing defect or simply not work that good. To rule this out, I’m buying a SanDisk Extreme 128GB from a reputable retailer.
Compared to my current Transcend cards, the SanDisk offers double the read/write speeds. While I’m not sure exactly how much throughput Klipper requires during a print, switching to a high-performance, reliable card is my next logical step i think.
on a hunch I installed the legacy version of RPI-OS (12 bookworm). Instead of 13 (trixie), and at the moment I am 45min into a print and it’s hasn’t crashed, where previously it all ways crashed within 10mins.
I will update tomorrow if it completes. But might be worth giving it a go to rule out a software issue.
