Long story short: Board on my 3d printer stopped connecting to Klipper and I couldn’t get firmware to flash. I needed it for a project, so I borrowed the SKR pro from my lowrider. Been using it on my lowrider for 6 months now without any problem. Got it all hooked up, flashed klipper. Everything seemed to be working just fine until I warmed the hot bed up. I set it to just 30 degrees to check and make sure the thermistor was set right, checked the bed with my infrared thermometer, it creeped up to 37, and then suddenly the cpu on the SKR started smoking with a bright red dot in the middle.
I figure the board is shot (which sucks because I’m down a CNC and a 3d printer now), but I’d really appreciate some help diagnosing what went wrong before I purchase and plug another board in and try and light my house on fire once more.
My guesses so far:
The board was defective. It’s been working fine, but since it’s been on my CNC and I haven’t had anything plugged up to the hot bed/hot bed power ends of the board and something was wrong with the mosfet or other parts I wasn’t using before. Possibly 12v got into the chip because there’s a manufacturing defect on the board. I couldn’t find any bad solder joints, or anything else that looked burnt out on the board
Maybe something is wrong with the hot bed or power supply on the printer? The previous board did stop working after all. I don’t think it’s this because I’ve had the printer for 4 years now and had no issues until now…but also it’s a hell of a coincidence to kill a board twice.
Maybe I wired something wrong? I doubt this one, but I won’t rule it out. I was following this guide pretty closely and double/triple checked where everything was plugged in.
I’ve been trying to figure out what happened and I’m stumped. Maybe someone smarter than me has some advice.
That is a tricky one. The fact that two boards died in the same printer seems like a big clue. Coincidences aren’t that common.
But I am stumped to explain it from anything outside the board.
12V going through the processor could definitely make it smoke. But the 12V should be pretty isolated. If it was a short, I would have expected it to go fast.
A short in the bed wires would be bad. It would dump a ton of current through the MOSFET, but it shouldn’t affect the st chip. Unless the MOSFET failed first and then showered to the output pin on the st chip. .
I am not sure. I would definitely check the power supply voltage. I would check the resistance of the bed heater. I would look over every wire and then I would probably try to install a cheaper board and cross my fingers.
Power supply voltage is reading 12.57. I don’t really know much about electricity to know if that’s too high.
Bed heater resistance is 1.2 ohms, which looks like what it should be.
Is there anyway to check if the mosfet failed? Or anywhere else I can poke around with a multimeter?
I could also try getting the old board working, because I don’t think it had the same problem. It died after I ran a bed mesh level a couple of times, and then I wasn’t able to reflash the firmware because it appeared like the bootloader wasn’t working. Which made me think that the eeprom went out. The only thing I have to flash it is an old arduino, and it just seemed like more effort than I wanted to go through to fix a cheap board when I’ve thought about replacing it with at 32bit board anyways. But then this happened on the skr which has be doubting what was wrong with the stock board.
I was also using an external mosfet on the cheap board, and that seems to be intact (I’m not exactly sure how to check it though) which makes me think I had two unrelated problems. I’m kicking myself for not just hooking it up on the skr, because that might have saved me.
The hot bed circuit is pretty isolated from the logic circuits. Because of this, I don’t think that the board was faulty and hooking up the heated bet did it.
So, places where 12V can get to logic circuits:
At the hotend, +12V goes to the heat block and the heat block is near the hotend thermistor. A short in the thermistor insulation can short +12V to the logic circuits.
At the heated bed. Ditto. It might be a bit more difficult, but if you have a PCB heated bed, the thermistor leats might scratch the insulation.
Endstops. Endstop pins are directly connected to CPU pins, and are definitely NOT high voltage tolerant. Heck some of these things can smoke a CPU with 5V. So something like a bad endstop configured to short power to the signal pin could do bad things indeed.
These would be my first choices to check.
Methodology: Disconnect all of the connectors from the mainboard, but not from the printer.
Use a spare Dupont pin to connect to one lead of a multimeter. Check all of the logic level leads for continuity to the heated bed and the hotend heater block. Anything where resistance is not infinite is a problem. In your case, because the problem became apparent when you turned on the heated bed, that is probably the prime suspect for where. I would most particularly check for continuity between the leads of the heated bed and the heated bed thermistor.
At least there’s a good chance that the TMC2209 drivers survived.
This has kinda what my gut’s been telling me since my last post, but when I checked, I couldn’t find any continuity. The hot end heated up fine (tested that before testing the bed) and I hadn’t actually hooked the endstops up yet. So that does leave the bed as the prime suspect, even if I can’t find out how.
I’ll poke around more tomorrow, but I might just replace the bed regardless, just to be safe
Gave the board another look over, and it’s hard to tell because it so small, but the diode that’s connected to the thermistor port looks slightly scorched. Still can’t find continuity between the thermistor and the heater…but maybe something is just loose enough that as the bed heats and the metal expands it connects. Regardless, I’m going to call it case closed.
Just a quick follow up in case anyone was curious:
Replaced the board and the bed. After removing the old bed and getting a look at the wires/solder pads and messing around with my multimeter, I think what might have happened was that heat and wear worked back the wire’s insulation just enough that it came in contact with the edge of the aluminum bed. It’s pretty difficult to replicate the connection, so I’m not sure how it got pushed up against it, but I can make it happen if I push up the wires just right, so I’m pretty sure that’s what went wrong.
Lesson learned and I’m replacing the wire that came with the new bed with some wire with silicone insulation.
New bed and board are working fine. Thanks again for helping me solve this because I probably would have been dumb enough to plug the old bed and fried a third board.
Such a bummer. Don’t feel too bad though I am sure most of us would have done the same thing. That is an odd issue. Do what you can to secure the bed wires somewhere before it makes any motion.
.