Temperature spike (or lull?) resulting in fault (Solved(?))

paniha

EDIT
May have solved the issue.
Jiggling various wires and connectors by hand, I got some noise on the connector between the thermistor and the cable leading to the electronics outside of the enclosure.
It only went open circuit when i applied significant force, but could easily be jiggled to a point where the temperature fluctuated +/- 30-50 degrees.

Seem the connector may have had some corrosion or patina resulting in positions where it would add to the resistance (resulting in a lower temperature report).
Replacing the connector has so far solved the issue, and made it so that the temperature does not change when jiggling the connector.

Hello, I have been reviving an old-ish large printer, and have pretty much gotten it in line.
As of two prints ago, I have however started getting heater faults with the temperature being bumped to 240-250C up from the desired 215.

I thought it might be a loose/shorted thermistor wire, but looking at the temp plot in web interface, it looks like it's under reporting the temperature for a while (causing the heater to, well, heat) and then finally catches up when the temperature is way too high.

The error seems sporadic, first time happened after around 10 hours of printing, second time after less than five minutes.
Repeating the second print after resetting the fault did not result in another fault in the 15 minutes i let it run.

Thought it might be a voltage drop from running steppers and heater, but heating up and running a G1 moving X, Y, Z and extruder at the same time does not change the temperature plot to a noticeable degree.

The machine is big (1200 x 600 x 680 mm print volume) so wiring is a pain, thus I wanted to hear if anyone had some wisdom before i blindly start replacing likely culprits.

I'm running a Duet 2 WiFi with RRF 3.4.5 and DWC 3.4.5
The machine is running a volcano hotend with E3D 40W heater and thermistor cartridges.

dc42

@paniha yes a poor connection to a thermistor can cause it to under-read. A poor connection on a PT1000 sensor causes it to over-read.