Noise on bed pt100 signal
deckingman last edited by
I use a PT100 on my bed ( just because I had a spare - no other reason ). Recently I've started to see short sharp spikes up to 2,000 degC in the reading. I wouldn't normally notice it except that it shows on the DWC plot as a series of vertical lines. It only ever happens when I move the bed a considerable distance such as during homing. I never see it at any other time and it doesn't cause any fault triggers so it's no big deal. However, I'd like to fix it if I can. I've checked all the connections and can't find anything amiss. I can wiggle the wires around all day long and never see any spikes. I never see them during normal printing either. Only when doing a large Z move.
So I would normally think that it has to be picking up noise from the Z motor wiring. The trouble is, the wires are nowhere near each other. The PT100 runs in a cable chain along with the mains wiring for the bed but I don't see any noise from that - I guess the chip is pretty good at filtering out 50hz?. I do get the noise when the bed is unheated but moving. There is one section of my frame where the Z motor wire and the PT100 run parallel to each other for about 400mm but they are either side of a 2040 extrusion so spaced about 40mm apart.
This has only recently started happening and I haven't re-routed either of the Z motor or bed PT100 wires so I'm inclined to think that it might be something other than the routing of the cables but I'm open to suggestions. The motor wires are twisted but otherwise unscreened. The PT100 is a 4 wire one.
Any ideas of what to check or try would be most welcome.
I was about to say try a 4 wire PT100 connection but you are already doing that. It's really hard to say if it's difficult to reproduce. I would eliminate the possibility of anything on board the chip by swapping to the other PT100 channel to confirm the fault follows the sensor.
What sort of cable are you using for the PT100 wires from the bed?
Is it possible that you have an intermittent connection in one of the PT100 wires, and bed movement is causing it to lose contact occasionally?
Dougal1957 last edited by
This is exactly the effect I was seeing in my Hotend setup however it only manifested itself when I moved to a Flying extruder setup and the wires where in the Same bunch.
I have re-wired the Hotend PT100 now with 4 core screened twisted pair cable but have yet to test it but I suspect it is noise picked up from a stepper somewhere (The Extruder one in my case) The spikes only ever happened when the extruder was actually running?
deckingman last edited by
Cable is just 4 strands of 7/02 - not screened but with nylon braid outer. The wiring for the hot end PT100 on the second channel is the same but runs inside a flexible conduit alongside the cables for all three extruders for about 1 metre, so a much worse case but that one is fine.
Can't find any intermittent connections when I wiggle the wires around. As I said, I can flex the wires and the connections as well as the cable chain that they run in between the moveable bed and the frame, all day long but never make it happen.
The only time I can make it happen is by doing a largish Z move and it's regardless of whether the bed heater is running or not. Never seen it during a normal print. So the only time it happens is when the Z motor is running continuously for a few seconds or more. Which would indicate that it's noise from the Z motor wiring being picked up by the PT100 wiring? I wouldn't have thought it's the Z motor itself because that's at the bottom of the frame, normally about 750 mm below the bed. The PT100 itself is embedded in the aluminium build plate, miles away from anything else (well a couple of hundred mm at least). The closest the PT100 wiring gets to the Z motor wiring is about 40mm where they both run up one of the 2040 legs but at opposite sides. Where the cables enter my Duet enclosure, the PT100 comes in at the top but the Z motor cable comes in at the bottom.
The odd thing is that it's only recently started happening and if it was interference due to cable type or routing, I'd expect to see much worse on the hot end PT100 with it's long parallel runs with the 3 extruder motor cables. Having said that, the extruders don't work as hard as the Z motor when it's lifting my 7kg bed,
I guess I'll try swapping the PT100s on the daughter board as Tony suggested and see if it stays with the sensor or with the channel on the board.