My nozzle temperature started to misbehave

Printers Stuff
#1

Today morning after a whole night print I spotted that my nozzle its showing -15C
I checked all the connections and it eventually jumped to normal but now the temp grows much over the set temperature. I also spotted some places that were molten on top of the print - like a drop of filament

Now when I try to print my temp goes way over the set temperature - thats what I can see in octoprint
Screenshot_2020-11-22 OctoPrint
Screenshot_2020-11-22 OctoPrint(1)
The display also shows such values - one is during the print and one after I stopped it
20201122_124327 20201122_124458
if I would not know that there is no relay I would think its there and its sticking
Im using currently 4.0.6 - I flashed it yesterday to disable bltouch/before I had the bltouch firmware and in the meantime I tried klipper because I had problems with starting the prints with bltouch - the first layer was not really sticking - I had to raise the bed a bit during the initial skirt printing
another thing I spotted on the diagnostics screen when the temp was going down it was showing around 700 and was raising but it often jumped to around 980 if I remember correctly and then back to 700 (977 its showing now when its in room temp)
Any suggestions ? Im thinking to solder some caps because it seems to be lot of noise going on on the lines - but heating to 290 and showing it on the screen is weird anyway because the cpu knows the temp? or is it separate part of memory - the heating temp and the display part ???

#2

This means that the board was reading 0. So looks like you have either a wiring problem or your thermistor is near dead.

Not sure to understand you sentence about “relay” but yes, there are some kind of relays in this printer: the MOSFETs. Could also be involved in your problem but the -15C reading points toward the thermistor.

I will move also your topic since it is not in the right category (your question is about your printer, not ADVi3++)

Edit: after all, your images were not so big.

#3

I did now more testing - I disconnected the ribbon cable and looked at the diagnosis data with the printer powered off - power just from my raspberry pi so there should not be really any noise from the motors but
with unconnected thermistor it shows 20-25 on the pf1 pin value but it sometimes jumps to 970 (like every 10th reading) without touching the machine at all so I don’t think its a wiring problem - the same happens now with the machine powered on with the cable connected but it jumps to 990
I may try a different version of the firmware and see what happens
Also by tracing the pcb I found that there is a capacitor inside for this thermistor

#4

Looks like you have to change your thermistor.
You made tests with the thermistor disconnected, right?

#5

If so, the problem is at the level of your mainboard.

#6

yeah - must be something there - but I cant see anything that may be a problem - the pcb trace goes from the pin header to the avr and the only thing connected is the capacitor- it all looks ok - I measured it and seems fine to me - Ill try reflashing and see how it goes

#7

This kind of things happened to others: a short at the level of the thermistor that “burn” the input pin (inside the Atmel microcontroller).

#8

but I can calibrate the pid without any problems can be seen on the octoprint screenshot - even now after flashing the firmware the calibration looks fine. I may search my electronics stuff for a temp probe and I add a second one.
Thanks for your suggestinons

#9

This clearly means that you get a problem at the level of the temperature measurement. At this point of time, your board received 0 on the temperature pin (0 Volts or open circuit). Something which is normally not possible.

So looks like your temp measurement (i.e. thermistor itself, the wires or the MCU pin) is sometimes working, sometimes not.

#10

I think I found the issue: When I was moving the printer and hit it a bit the heater led by the mosfet was turning on - I was thinking that its driven from the microcontroller pin but I think its driven from the mosfet along with the heater.
and my issue was that I accidentally caught one of the heater wires under the radiator when I was taking it apart and when the gantry was moving it was shorting it probably to the ground and then the heater was heating.
when the gantry was not moving during pid tuning everything was ok.
My adc readings still sometimes are fluctuating in the diagnostic tab but I think it was like that always and now the nozzle temperature is stable during printing.

Just a suggestion: I was thinking to add the leds to my front panes to see when its on but its a bit of soldering so not everyone can do this - maybe instead we can change the bed or nozzle icon color instead in the firmware?

#11

This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.