Error caused by thermocouple-connector junction

[BrunoARG]

Hello guys

Not to create another topic, I will request your help here.

I am making my own soldering station, and I want to know if I have to make any kind of junction for the connection of the thermocouple metals with the DIN-5 (iron, I think) connector. Every different metal junction will generate a Seebeck potential, and that's going to add an ambient temperature dependant potential, which I am going to compensate with a diode as actual sensor.

The matter here is... what's the total "offset" generated by this thermocouple-connector junctions? It's a K thermocouple soldered to the DIN male connector. Will it be the same 41uV/ºC, or another ?

Thank you very much.

 

[betwixt]

Re: How to interface a K type Thermocouple with microcontroller?

The voltage will be different for each type of metal-metal junction but you can either accept a fixed offset caused by other metal-metal junctions or you can link another two pins of the DIN plug together, measure the potential across them and subtract it from the thermocouple voltage.

I notice my own commercial soldering station reads 25C without having the iron part plugged into the base at all. I suspect it uses a 'nominal' workspace temperature without actually measuring it then adds the uV/C from the thermocouple to it to get a close approximation of tip temperature.

Brian.

 

[FvM]

Inserting a different metal (e.g. connector, terminal strip) into a thermocouple wire will only generate an additional thermo-electric voltage if both end junctions are at different temperatures. This can of course happen, if the connector is mounted in an enclosure wall with different in- and outside temperature. For this reason, dedicated thermocouple connectors are made. The error is rarely larger than 1 or 2 K and e.g. mostly irrelevant for a solder iron.

Please consider that the cold junction reference thermometer (e.g. diode) must be placed at the junction between thermocouple wire and measurement circuit. It can be at the connector.

 

[BrunoARG]

I know that the heating element is 4 wires: 2 for the heating element, and 2 for thermocouple, and there is a PCB where they are soldered to join longer wires that go to the connector.

Here a little diagram:

Copper is from the PCB. The wires could be iron or copper, doesn't matter because they are the same and at the same temperature.


There is a very low voltage (but actually exists) that is generated between Chromel and copper, and Alumel and chromel. Is there a way to calculate it to avoid measurements? (I know there is, but I don't know how).

If the junction between copper and the thermocouple is right in the handle, there's no way to compensate it with temperature but with a constant voltage. Will that affect so much?

Thank you for your replies.

 

[c_mitra]

The voltage between Chromel and Alumel is the one of interest. There will be a voltage between Cu and Chromel AND another voltage (different) between Cu and Alumel. The difference between these two voltages is the cold junction potential- i.e., the voltage between Chromel and Alumel at 40C.

In your diagram, you will need to apply cold junction temp correction for 40C and NOT 28C.

 

[Warpspeed]

The usual way to do this would be to get a length of "K type" thermocouple wire, (which is chromel alumel), and use that to extend the thermocouple wires already in the soldering iron, back to the temperature control system.

There will almost inevitably be a plug and socket at that point, where you have a junction of dissimilar metals, and where an error is introduced. But it will be well away from any heat producing part, and room temperature can be assumed.

Room temperature will then be subtracted from the soldering iron temperature due to this introduced error voltage.
Many ways to get around that, but the simplest of all might be to just know what room temperature is, and add that mentally to the temperature you see on the readout.

 

[c_mitra]

I am not sure how much accuracy is needed for the temperature measurement of a soldering iron. The moment you touch the tip to the solder, or to the copper on the board, the temperature may drop 5-10C (guessitmate value). band gap based devices can measure absolute temp directly (I do not know about the accuracy; but they will surely go upto 200-300C; if you want higher ranges, you can certainly use SiC devices)

 

[BrunoARG]

Thank you for your help!

Well, it seems that as there is not a very high room temperature variation (maybe 20ºC in the whole year) I can just correct a value equivalent to 30ºC. Why 30? Because the thermocouple - copper junction will be located right at the handle, and as it's pretty near to the heating element and it's holded by a hand (36ºC), it would be the most appropiate value between it and 25ºC.

Thank you for replying again, salute.

 

[c_mitra]

If the most likely used temp in your soldering iron is expected to be 250C, you can manually calibrate for that one temperature. Inside the room (and your workplace) the temp variation during the year is expected to be much less.

The heating coil is fed a voltage that us PWM'd. Essentially you are feeding the heater with a variable power source. The heater attains a steady temp when the input power equals to the power loss by conduction and radiation.