Thermocouple interference

Phil333

Hi there,

I am using a 220V AC silicon heatbed with an included K-type thermocouple.
For the Duet I am using the thermocouple daughterboard cersion 1.1 which is MAX31856-based.

I pretty sure I am getting interference because I am getting this message:
Error: Temperature reading fault on heater 0: sensor short to other wiring

The power and thermocouple wires from the silicon heater are about 1.5m and goes into a small box with two SSR that turns the heater on and off.

From there I have another 4-pol wire (about 2m) that connects the SSR box with the "mainboard box".

2 of these wires are to control the SSRs (24V and GND) and the remaining 2 wires are for the thermocouple.

This error only happens if the bed is beeing heated and is about to reach the target temperature.
So I think its because either the "heating bed" output of the Duet is turned on/off or of the SSRs.

During the heating I get spikes of 2000°C readings.
I set the bed to be heated to 100°C and at about 98°C I get the first error.

I know the wire lenght is on the upper side with about 3.5m but I can not reduce it.

So what would be your best guess on what is causeing this problem exactly and how to solve it?

Problematic points are:
I read something about a "2 wire twisted" cable, but I dont have something like that here. Will another shielded cable help/solve it?

Is the main problem coming from the 4 wires in that one cable (2 thermocouple + 2 for 24V and GND for the SSR) or might it more likely be coming from inside the SSR box, with a relative close distance (about 4cm) to the SSRs that is causing this?

Thanks in advance for the help

dc42

The wire between the thermocouple and the daughter board must be twisted pair thermocouple wire of the correct type; for example, type K wire for a type K thermocouple. Search for "thermocouple extension".

Check that the SSR you are using is a zero-crossing type. Most DC-AC SSRs are.

If you still get interference when you use the correct cable, adding series resistors (e.g. 1K) at each thermocouple input on the daughter board may help.

It's very unusual to use a thermocouple to measure the temperature of a heated bed.

Phil333

Thank you DC42 for your fast and competent help.
I will buy a different cable then.

The last sentence is a bit funny tho
Tell this to the manufacturer of that silicon heat mat that its uncommon.
I would also have been happier if it had been a regular thermistor

Phil333

I just want to finalize this:

I bought a quite expensive (~30€ for 3m) twisted and shielded k type thermocouple wire from a german specialized manufacturer. The wire was tested by the manufacturer before shipping.

In the meantime I made my own twisted wire with 2 0.22mm² cables I had around.
Since then the temperature readings were good but every now and then when I touched the 3D printer, the failure happened again.

After the expensice wire came, I also tested this one.
Result is unchanged.

Nearly everytime I am touching the 3D printer, the thermocouple reads 2000°C and I have to shut the power off and on.
So I can not touch the 3d printer while printing or it will abort.

The reason for that is definetly the electrostatic charge I give the (metal case) 3d printer, that accumulates while walking on a carpet floor.

As the case is already earth grounded I think there is nothing else I can do to avoid this.

So I will order a thermistor and try to put it in between the silicon heating pat and the aluminium printbed.

Its a bit sad, that I spend 60€+ for the themocouple daughter board and the k type twisted wire for no benefit

But at least I learned something.

If anyone has an idea that might solve my issue, you are welcome to help.
Otherwise I will just use the "dirty fix" with the thermistor.