Odd PT-100 behaviour after firmware update
I recently updated all my firmware and immediately began getting errors from my temperature sensor. Basically it went from working perfectly to reading 2000, which could be rectified by sending M305 P1 X0, then M305 P1 X200, then M562. I found that wriggling the cables near the effector platform could induce the error.
I tried to nail down the cause: I reconfigured the sensor to be a four wire setup using twisted pair wires, then checked and replaced electrical connections, then finally replaced the PT-100 sensor. It still happens. I think it may be the result of the PT-100 cables being close to one of the other signals coming to the effector platform (hot end fan or IR height sensor) and having some signal induced in them, since I can cause the error to appear by manipulating the cable bundle near the effector (I have checked for damaged insulation and shorts). The problem there is that this happens when the printer is at standby, not just when the hotend and fan are working. I'm about to start disconnecting wires at the Duet to see if I can isolate a set of wires somehow causing interference.
Can anyone suggest to me:
1.) A good way to diagnose the cause.
2.) Why this only started happening after a firmware upgrade?
Something probably got loose when you last touched the 3D printer. There may have been no correlation to the firmware upgrade. You may have thought the upgrade caused it because it happened around the same time.
If you wriggled the cable and it shows up, most likely you keep wiggling until find the location of the intermittent connection. Most likely cause is due to a bad crimp in the connector of the PT100 sensor. I know the E3D style of sensor is very prone to breakage. It is only a single wire that is crimped to a contact. I ended up soldering the wire to longer set of stranded wire and shrink sleeve over the connection. I then zip tied the leads to make sure any bending does not flex the connection.
I had to recrimp new contacts three times until I used the method above. The single strand of wire breaks too easily when its directly crimped to the contacts. There is no stress relief for the single wire at the crimp location. Any movement of the wire would stress the crimp until the wire breaks.
It does seem to have been a mechanical issue after all, thanks