Wild bed temperatures

Hi all

I have a clone I3 that has been running for a number of years with no issue, then one day, out of the blue, the bed temperatures started to jump wildly. Thinking that it was the thermistor, I ended up buying a new heated with integrated thermistor. This worked fine for a while, and the same thing started to happen again. I switched the sensor pins on the ramps board and uploaded the modified Marlin firmware. This helped for a while, but it wasn't long and the issue returned. Attached are the thermal readings from Octoprint. What is odd, is that when the temperature falls below a certain value, the reading is ok again. This isn't an issue so much when printing PLA as I just heat the bed for the first few layers, but its going to be an issue when I print ABS. Also, when the the thermal issues started, I noticed that the overall printing quality went down. Thinking it was a supply issue, I switched to a newer more powerful supply, without resolution. I have tried many variations of sensor configurations, again, no fix. Right now, I suspect that the ramps board is slowly but surely seeing its demise, however, I would love a fresh opinion or two on the matter. Any help will be appreciated.

Thanks all

Gavin

That's quite a baffling graph. Are you sure the heated bed cools down? When the read out jumps up for a short time you'd expect the temperature to drop but it doesn't. It could be a communication problem between octoprint and the ramps board. The temperature may be correct and the pid loop may still be adjusting it while octoprint receives garbled data.

Do you have a volt meter? Do you observe the same wild read out?

---

--
Kind regards
Imqqmi

NFAN CoreXY printer:
[reprap.org]

Hi imqqmi

What I can tell you is that its not a communication issue. If I don't turn off the bed, soon enough the firmware halts the print with a temperature error. I will however try looking at both the voltages on the bed and thermistor. Thanks for the idea.

Cheers

If its a system with polyfuses, check they are not getting hot and cutting out.

I see:-
small spike at -12 which may be the start of the issue.
At -9 the temp jumps up about 20C. This is wrong as temperature can't rise that fast. Can't be a loose thermister as temperature would drop. Can't be open circuit thermister as it would go to max temperature.
30secs later the temperature drops to the correct value. I think imqqmi's point is valid but unclear as the temperature drop in 30 sec is quite small.
At -7 it goes high again (similar level), stays high but does seem to track the bed temperature which would now be turned off.
As its a thermister the higher temperature must be caused by either a lower resistance or a higher voltage at the ADC. In either case it seems to be intermitent but related to temperature and time. Its also a repeatable offset. That makes me think its not the thermister circuit directly but may be a parallel circuit switching in and out. By that i mean that resistors in parallel produce an overall lower resistance. So suddenly connecting to another resistance will cause the resistance of the temperature circuit to drop - and it will be a defined amount as the other resistance is a defined value.
So i would expect it to be a joint or component on the control board.
Hopefully some one who knows the Ramps circuit can identify possible checks and solutions.

Oh just thought - temperature related short on the heat bed. Have to think through the permutations for that but there is no sign of the bed heater afecting the temperature so it can't be shorting to the 12v feed side or the earth side. I just have visions of the thermistor joints shorting through the bed material.
Then im probably over thinking the whole thing.

How is the thermistor attached? Are you using thermal paste?

Thermal paste is conductive, and not all of these are really well insulated, so the thermal paste might have an intermittent short across the thermistor. It won't be like a dead short, but it might be enough in parallel with the thermistor to account for a sudden rise in measured temperature.

Another possibility that occurs to me is a poor thermal connection, so the bed temp is actually 15-20 degrees hotter than usually indicated, but something puts the thermistor in good contact with the bed from time to time, showing a spike in indicated temperature. I had something like this happen in a hot-end on me once

---

MBot3D Printer
MakerBot clone Kit from Amazon
Added heated bed.

Leadscrew self-built printer (in progress)
Duet Wifi, Precision Piezo parts

Thank you all for the replies.

@Dust, the polyfuses are warm but not hot, also they are now directly under a fan and so stay very cool
@MCcarman, quite a bit to consider, though I am not sure where to start myself. Though the heat related short is something I will look at
@SupraGuy, its an Anycubic Ultrabase and the thermistor comes premounted in the board. The previous hotbed (where it did the same thing) was a generic cheap and I tried messing with the heat paste as well as a few different thermistors. What you describe definitely happened but a lot more frequently than my current issue. This is why I ended up going for the current hotbed.

I will have a tool around in the next few days and give some feedback to the above suggestions.

Again, many thanks for the replies.

Hi All

I have had family life get in the way, so its been tough trying to resolve this. As an exercise in hopefulness, I stripped all electronics and rebuilt the printer. Right now it seems to be working fine. We will have to wait and see though.

Cheers