Seebeck measurement tool

[zenman]

I need to measure the seebeck coefficient for some research I am doing on thermomaterials but the cheapest measurement tool I can find costs $9000! Here is a link:  http://en.jouleyacht.com.cn/product/20.html

So I have decided I will try to reverse engineer it. My plan is to combine a soldering iron, a microvolt meter, and two thermocouples connected to an ESP32 microprocessor. Then dump the measurements into a google spreadsheet and calculate the seebeck value.

I plan to use:
 ESP32 mcu
 AD620
 MAX6675 + therocouples
 60watt soldering iron

I am kind of guessing with my design as I have only watched someone else use a seebeck measurement tool and have never even operated one let alone know the detailed specs.

It is ok if my tool is not perfectly accurate. I would be ecstatic if I can get 80-90% accuracy. That would still move my research forward. Is there anyone here that has experience with seebeck measurement  tools and can offer some advice?

[mvs]

My plan is to combine a soldering iron, a microvolt meter, and two thermocouples connected to an ESP32 microprocessor.

You can try to use two soldering iron tips from cartridge based soldering station as probes, like Hakko T12 or JBC C210. They have integrated heater and thermocouple.

[zenman]

That’s a good thought. I was thinking about building the thermocouple in so I could output the data to a microcontroller. But maybe working the other direction is simpler and use their physical design for the housing of the thermocouple.

[Kleinstein]

Measuring the Sebecke coefficient is rather tricky, especially measuring the actual temperature at the contact points. Much of the difficulty is likely in preparing the samples. It can be relatively easy with large samples of low thermal conductivity material, but rather tricky with brittle, high conductivity materials.

The general idea would be measuring thermal EMF and temperature difference. In the simple form this could be with 3 material junctions on both sides. To form 2 thermocouples:  one could be something like copper and sample and the other could be the same copper and constantan.

The setup really depends on the samples - likely no one version fits all solution.

[doktor pyta]

https://www.ebay.de/itm/Messplattform-fur-Seebeckeffekt-v-Fraunhofer-Institut/352183775462?hash=item51ffc9f0e6:g:sToAAOSwB-1YyYbq:rk:2:pf:1&frcectupt=true