Why use carbon for thermocouple junction "welding"?

[cupoftea]

As 5:48 of this video shows..

..welding a thermouple junction is just about making a circuit out of the two paralleled thermocouple wires, and then connecting this circuit momentarily via a carbon block. The carbon block is needed as it vaporises on contact and so produces an arc due to the air gap thus formed...the arc producing the heat needed to weld the two ends together.

We don't have a carbon block, so will instead use the following graphite pieces..

https://www.amazon.co.uk/dp/B0033WLZFG/ref=pe_27063361_485629781_TE_item_image

Since we also don't have the £2500 thermocouple welder, we will simply use the rectified and smoothed mains. We will have a variac to get the right voltage for the best "weld".
Do you know why they use carbon in the video? Its far harder to source blocks of carbon, but graphite is easy and cheap to source.

 

[betwixt]

Looking at the markings on those graphites in the link, they appear to be for artistic rubbings, essentially large sale pencil lead. They may not be conductive enough or withstand the heat of the weld. Pencil lead and artistic graphite is a mixture of graphite and clay, the proportions deciding whether it dark or light. Note the ones in the link are marked 2B to 6B and Koh-i-noor is a pencil manufacturer.

These are more like the ones you need:

High Purity 99.99% Graphite Rods Graphite Electrode Conductive Cylinder Bars | eBay

99.99% Graphite Electrode Cylinder Rod. Material: graphite. Graphite electrodes, carbon rod for spot welding machine, carbon rod machine, butt welding machine. Purity: 99.99%.

www.ebay.co.uk

Graphite or other carbons are used because they arc nicely to produce intense heat but do not contaminate the metals in the junction because that would change the thermocouple characteristics. You might consider spot welding as an alternative bonding method.

Brian.

 

[BradtheRad]

We don't have a carbon block, so will instead use the following graphite pieces..

Carbon-zinc cells may not be readily available. As a boy I disassembled these to find the rod of carbon in the middle. In flashlight batteries and the larger dry cells.

 

[cupoftea]

Thanks, i see the point about the clay in the graphite.
I think we will use these carbon brushes....(?)

https://www.amazon.co.uk/YUMIN-Electric-Drill-Carbon-Brushes/dp/B0CWN8L5M7/ref=sr_1_104?dib=eyJ2IjoiMSJ9.KQReFG2vjMuP5A_7HrkzssbrsWAMnJoskfyZwhS5jEIyBPEUYzb4nOWrz5qIlH5o9nE-ZmBGj5aUIu7i6hDPT8EnX9dYLdpa5R4vkNMgeUfIIFjQF8me3W3Mn0nW7ahqBXnxTPrUGPVBFdASXTm0Tg.ysVx-vORj3_0pfnU5-twLez4fq_oPXehEM_ftz2MV_s&dib_tag=se&keywords=carbon+brush&qid=1725193018&sr=8-104

...they will be especially good as they have a metal bit that we can connect up to.

 

[dick_freebird]

Spot welders tend to have bare copper tongs and might poison the melt, or demand control that might not be had every time, every operator (esp. the untrained - our Rel lab had one guy who always got the task).

Check out replacement motor brushes if you want high current density and low resistance. Cheap enough at Ace.

 

[betwixt]

The brushes in post #4 should be OK but beware that in normal use the current through them is relatively low, the connecting wire may not be able to withstand higher currents and the spring probably isn't attached to the brush but trapped between it and the contact.

Brian.

 

[cupoftea]

Thanks, we now have the brushes and have used one to
do the thermocouple welding. Firstly, charged a 220uF cap
up to 370V and then touched the thermocouple ends to it.
This seemed to have worked, but the junction just didn't look
good enough...not a nice "bead". And after yanking at it for a while it eventually came apart.
We then tried 660uF charged up to 370V..but that was too much...it literally vaporized all the
stripped thermocouple wires back to the plastic insulation.
So we will try something inbetween. Ie, the right combo of voltage and capacitance.
We find that the thermocouple wires must be twisted for the welding.
If they are not twsited together then they just blow apart on contact with the carbon.
This (twisting) is bad because as you know a twisted thermocouple has multiple parallel junctions instead of one end junction.
This then means that you end up reading a lower temperature than where the end bit is.
I suppose we must try and untwist the wires after the weld has been done?