So I haven’t been able to complete a print for a few weeks. The pi shuts down mid print and ruins the print. There was one time when the pi reset itself and showed a timer too close error message in mainsail but most of the time it just stops printing and the pi becomes unresponsive. After I saw the timer too close error I ssh’d into the pi and ran top and I couldn’t see any processes consuming excessive resources.
I was running a pi 3 b+ when the problem first started happening intermittently and then it failed one time and corrupted the micro SD card to the point where the SD card couldn’t even be rescued. So I flashed a new SD card with Klipper and that card got corrupted when the pi shut itself down mid print a few prints later.
When the second SD card failed I suspected that the pi might be failing, so I switched to a pi zero 2W and rebuilt the firmware for the printer MCU. That ran fine for a few days and then it began failing mid print like the B+ had done. I purchased a new PSU (an official raspi pi PSU) and then it ran fine during a burn in test where I just ran the bed heater for several cycles but then when I tried to print it failed again before the print had completed.
I’m no expert on klipper logging, so maybe someone else will spot something I’ve missed.
I did have heatsinks on the 3 B+ but no fan Eddy. It was shutting down due to overheating before I put the heatsinks on it. It ran fine like that right through summer when the ambient temp was getting well over 40 deg C, I only started having problems recently and the ambient temp has been around 12 deg C.
So I just completed my first successful print for a few weeks. It seems like it’s solved but I need to do a longer print to confirm it.
I ran the stress tests Sineos suggested in the thread he linked to & everything seemed fine, so I added dtoverlay=dwc2 to /boot/firmware/config.txt and it seems OK for now.
The MCU topped out at 34 deg C and the Pi hovered around 48 to 49 Deg C and went to 53 for a few seconds at one point. So I can only assume it wasn’t cutting out from a thermal shutdown before.
So I was able to knock out a 4 1/2 hour print the other day and I thought the issue was solved but I just got another timer too close error after barely warming it up, so the issue persists. So it’s as reliable as 2 bob watch. Maybe the raspi zero 2w isn’t up to the job, who knows. I’m pretty much over it with that printer now. I can’t waste any more filament or sanity on it, so I’ll just scrap it and replace it with one of the cheap core XY’s that are everywhere these days.
Thanks Sineos, I had a suspicion that EMI interference could be an issue. The printer & Pi are on the same circuit as a fridge, which is not really ideal but that’s unavoidable right now. My ocilloscope died recently so I’m just guessing what could be wrong really. I put a hifi style line filter on the power line hoping that would smooth out any noise on the power supply but that won’t help if it’s RF interference.