Thermocouple problems

Veti

have you tried a pt1000?

stienfromarden

pt1000

Nope, don't have any lying around here..

Veti

https://duet3d.dozuki.com/Wiki/Temperature_sensors

unless you need very high temperature the pt1000 is superior to the thermocouple
and its not Susceptible to interference.

stienfromarden

temperature will be pretty high since I intend to print a lot of different plastics. Besides that, I just spend 80 euro's on the two boards and because the construction of the extruder fitting pt100's would be a pain in the *ss...

Veti

@stienfromarden said in Thermocouple problems:

pt100

pt1000 not pt100

stienfromarden

still, I hope to be able to fix this instead of buying different sensors..

Veti

@stienfromarden
have you tried with the daughter board on the duet and extending the thermocouple wiring?

stienfromarden

not yet, but I need to extend the thermocouple wire for about 6 meters... so that means buying 3 (16 euro) thermocouples per heater (9 totall)...

What I do not understand is, why does it work directly onto the duet, the signal after the daughterboard will be 5 volts, I have some more cables in the lenght transfering 5 volt signals with no problem. So the problem is in extending the position of the daughterboard... this is what I really don't get..

engikeneer

@stienfromarden what signals are on the other lines?
Also, have you tried measuring the signal voltages at either end of the cable?

My understanding is that the daughterboards convert the thermocouple into an analogue signal/resistance which can be read by the normal thermistor IO pins in the Duet. This makes it susceptible to noise, particularly if your sensor has a high dynamic range (like a thermocouple).

Likely souces of noise could be your stepper motors/wiring or even just dodgy contacts/crimps in your connectors. I'd double check the crimps/wiring and try routing them away from the stepper cables. Maybe even test with the extruder motor unplugged? It looks like your extension cable is shielded - can you ground that at one end to help reduce interference?

There could also be a voltage drop along the cable which means the daughterboards are dropping out or the signal is attenuated?

stienfromarden

The other lines are, the stepper, fans and the 230volt heater bands. BUT these are all turned off while I was testing. It would be surprising if this causes EMI. I also did a test while unplugging the fans and plugging them in, but that didn't give any changes, so I think EMI is out of the question.

The voltage drop is something I will check out tomorrow, that will clear up a lot I think.

Thanks!

stienfromarden

BTW I also use the breakout board, could that also be an issue?

engikeneer

@stienfromarden I was wrong on the daughter boards comms - they use an SPI interface to the main board. A quick google suggests that this has a useable distance limit of 5-10m so my guess is that's your culprit (anyone who knows more feel free to correct me if this is wrong!).

Fair point on the EMI, though I've always found in an elusive beast. Could even be something else in the room? With the steppers, were they definitely off, or just in Idle mode? although the current is low in idle mode (default is usually 30% in RRF, but you may have set different), it can still be something (and particialry with long wires cause issues!). I was meaning actually unplugging the stepper wires from the Duet to be sure/because I'm a mech-y and don't trust electronics...

I don't see how having the expansion breakout board should impact things (as you're connecting the daughterboads to the main board), but it can't hurt to unplug it and try?

stienfromarden

The SPI signal could be the problem, but I am reading 10 meter maximum and I run 6 meter...

I don't think EMI is the problem because when I tested the thermocouples on the daughterboard directly on the Duet, although the tc was running through a lot of wires in the electronicsbox there wasn't any EMI. Or maybe it is in the 10 wire cable in between the daughterboard and duet..

stienfromarden

What could be an option is to extend the thermocouple wires... but that is an expensive guess....

https://nl.rs-online.com/web/p/thermocouple-extension-wire/3706228/

stienfromarden

ok there are no voltage drops at the beginning and end of the cable, I meassure 2,98 at both sides.

I figured out that moving the connections a little changes the reading, sometimes to the ambient temperature, I will receive some plugs to fit the daughterboard and duet, maybe this will help some!

T3P3Tony

While SPI as a protocol may be ok up to 10m in some situations, the SPI bus voltage and frequency we use is not. I recommend extending the thermocouple wires. Its probably wise to shield the thermocouple wires as well over that length of run.

stienfromarden

I decided to just go with some pt1000 sensors:

https://nl.rs-online.com/web/p/platinum-resistance-temperature-sensors/1005972/

It's just too much trouble for a bit faster response time and accuracy.... extending the cables of the tc will cost more then switching to pt1000 sensors

stienfromarden

Is it an option te reuse the type j termocouple wires to extend the pt1000? just for about 20cm to reach the termination box? these wires are steel braided and can handle some high temperature..

dc42

@stienfromarden said in Thermocouple problems:

Is it an option te reuse the type j termocouple wires to extend the pt1000? just for about 20cm to reach the termination box? these wires are steel braided and can handle some high temperature..

Doing so will introduce a small temperature error. The size of that error will depend on the temperature difference between the two ends of that thermocouple wire. It may be insignificant, I haven't done the sums.

stienfromarden

@dc42 said in Thermocouple problems:

@stienfromarden said in Thermocouple problems:

Is it an option te reuse the type j termocouple wires to extend the pt1000? just for about 20cm to reach the termination box? these wires are steel braided and can handle some high temperature..

Doing so will introduce a small temperature error. The size of that error will depend on the temperature difference between the two ends of that thermocouple wire. It may be insignificant, I haven't done the sums.

@dc42 thank you very much for the answer!

I did not now if these:

https://docs.rs-online.com/f7e6/0900766b8155404c.pdf

pt1000's were ok I don't know much about them but I hope it will work! Can't wait to power up the extruder an PRINT!