R and S type thermocouples above 1050C

[Gary 0]

Hey folks,

After I got the previous issue actually reading an accurate measurement from the R type thermocouple I was using, I have come across a new issue. This has occurred with both R and S thermocouples. Essentially when the temperature reaches 1050C, the temperature reading starts to fluctuate by a few degrees, then it gradually gets worse until the swings are plus and minus 100C before an error message is thrown.

I have spoken to the thermocouple suppliers and they have assured me that they should be able to comfortably go well above 1300C. I am using a length of compensated thermocouple extension wire between daughter board and the thermocouple which has a ferrite ring towards the daughter board side. I did try shielding this cable however this caused more issues with the signal to the board.

Does the duet have the ability to read temperatures as high as I'm looking to read? Or am I going to have to find another interface to allow me to read these high temperatures? Is there likely something else I'm missing?

All temperatures have been verified using a high temperature flir thermal imaging camera so I'm confident that the temperature readings are accurate until we hit 1050C. I have attached a couple of images showing the behaviour.

[dc42]

@gary-0 I am not aware of any reason why you should not be able to read high temperatures. I recall at least one user reading up to 1400C in the past.

The way in which the difference between high and low readings gradually increases suggests something to do with connections or wiring to me. Are you certain that the connections between the thermocouple leads and the extension cable is sound? Is the thermocouple junction definitely isolated from the case? Is the hot end metalwork grounded?

[TLAS]

@gary-0
Beyond the potentially grounded tip with interference that DC mentioned (which is likely), you may also want to check your heat source with an independent thermocouple. At high temperatures air currents or other heat-transfer mechanisms can cause oscillations in temperature that are in fact real.

Just buy a k-type and TC reader online to verify it isn’t an issue with whatever you are measuring. K-type dual measurement might also give you some clues to the electrical interference because of the different EMF profiles.

[Gary 0]

@dc42 Thanks for the response.

Ok so it should be possible.

Interference was my first thought. I will go over all the connections however I can say that the thermocouple is fully isolated from the metal work which is also grounded.

I get the feeling i'm going to be chasing a gremlin that will be something incredibly simple in the end.

[Gary 0]

Quick update for anyone following the thread.

I have a thermocouple reader on the way. Decided to splash out a bit more and get one that would handle the S and R thermocouples directly. That along with twisted pair shielded wiring should hopefully solve the issues I've been having but time will tell.

I'll update here with my findings.

[Gary 0]

An update here,

So I hooked up the indepedent reader to the thermocouple and it was reading the same value as the duet board throughout. Think I can safely assume that the duet board and the associated wiring is not causing my issues.

I am now suspecting that the interference is coming from the electric heating elements inducing a current in the thermocouple. They are unfortunately rather close due to design constraints. A small capacitor across the terminals has helped but I think I'm going to have to look at moving it and adjusting the heating characteristics of the elements as the interference drops when I lower the duty cycle of the elements.

[dc42]

@gary-0 my guess is that one of two things is happening:

1. There is some leakage between the heating element and the thermocouple, so that when the heater is on a small current leaks to the thermocouple wire. As the thermocouple output is measure in microvolts per degC, it only take a very small current to affect the reading.

2. The heater current is magnetically inducing a current in the thermocouple wires. Using twisted pair wiring for both the thermocouple (which should already be twisted pair) and the heater, and keeping the two sets of wires apart, will minimise this.

If the reading instability gets worse with increasing temperature and time, then my guess is that #1 is to blame. Can you try a different heater cartridge?

[Gary 0]

@dc42 Hey thanks for getting back to me.

I have experimented with moving the thermocouple and also installing a few capacitors here and there to try and remove the noise but nothing was working. Sadly I can't change the heating elements as they are custom made for our project and rather expensive. You can probably guess from the temperatures I'm dabbling with that we aren't using a 40w heater cartridge!

Well I'm going to tentitively say that I may have solved the issue I was having. I had setup the heating elements with PID control etc. However removing that and just having it setup as bang bang control with no reduced duty cycle or PID has removed the interference I was seeing around 950C. I can only guess that this was causing some signal to be induced in the thermocouple.

So for anyone else experimenting with super high temperatures with elements run off AC power just setup your heater with bang bang control.