Weekly 4: Conductive Plying
goal:
Since I have already spent some time working with the resistance of metallic thread, I wanted to see if I could push my skills a bit further this week. So, I set out to create a soft thermocouple.
research:
I understood going into this week that thermocouples are an electric component used to sense a change in temperature, but that was it. In my research I learned that basically — because physics — when some conductors are heated or cooled, they create an imbalance of charged particles. Depending on the conductor, charged particles are either attracted to or repelled by heat (approximately) and this gives an unevenly heated conductor a charge. This phenomenon is called the thermoelectric effect.
So in a thermocouple, two conductors with different thermoelectric properties are paired to encourage the charged particles to flow and create a current when the system encounters a temperature gradient. The propensity for charged particles to move when a conductor is heated is measured with the Seebeck Coefficient. The larger the delta between two conductors in a system, the more likely they are to generate a detectable current when subjected to a temperature gradient.
The key to a thermocouple system, is that the two conductors must be connected at the point where they will be subjected to a temperature change. Then, they must be completely insulated from one another between the sensing node and the electronic sensor that will detect a change in current. I hoped that plying two suitable conductors with a nonconductive material would keep them electrically insulated from one another between the sensing node and my multimeter.
plan:
My first goal was to find two conductors that would be suitable for generating the Seebeck effect with a readable current change. I started by looking at the materials already available in my kit, but I was disappointed to find that silver and copper have the same Seebeck coefficient (meaning that their pairing won’t generate a current). While carbon has a slightly different value, common commercial thermocouples have a combined Seebeck coefficient of between 8 μV/K and 60 μV/K, so a measly 3.5 μV/K likely would not work well as a thermocouple.
So, I started to look at common thermocouple pairings. I was especially interested in trying to source a material with a negative Seebeck coefficient to pair with the copper or silver thread. This process, paired with lots of Amazon shopping, led me to find spools of small gauge chromel wire (nickel chromium alloy, also called Nichrome). Since this material is used as the negative Seebeck coefficient half of K-type thermocouples (the most common thermocouples in consumer thermometers), I thought it was a super promising start and ordered a spool!
process:
1: Test the system
Before plying my chromel wire, copper thread, and non-conductive yarn together, I decided to test the Seebeck effect on the copper/chromel thermocouple first. So, I twisted one end of the copper and chromel together, and connected the separated ends to my multimeter on the mV setting. Then, I heated up the thermocouple end of the system to test the effect.
While the resulting voltage was small (4μV max), I was able to get a reliable voltage change on heating that returned to zero! I tried touching the leads while the system was creating voltage to see if they would lose their effectiveness, from what I can tell the voltage doesn’t change when the system is touched. I did not use my hand to connect the leads. I also tested the voltage change when the leads were different lengths, moving around the voltage tester, and saw no change in output regardless of the length of conductive thread/wire.
2: Ply, ply, ply
I chose to use a basic, medium weight, acrylic yarn as the carrier for the two conductors. It was thick enough to allow for reliable separation between them, was non-conductive, and a high enough contrast to the two colors that I would be able to inspect them to make sure there was no contact between the two leads.
Plying the yarn took forever, and I tried many different methods. I managed to get 12 inches of composite using the drop spindle. Keeping the two conductors at equal tension was key to preventing overlap, but this was challenging because they had different grips in my hand. I definitely think this process would be more efficient and reliable with mechanization from something like a fringe twister.
reflection/results:
It worked! I was able to make a thermocouple that could be used to knit something, which feels like it shouldn’t even be allowed. Making this system more accurate would most likely involve looking for a sensor that operates on the μV scale, rather than the mV scale. I would be interested in a more formal understanding of how the output changes based on the amount of heat being applied to the system, but it would be a waste without a more sensitive computer.
Additionally, I learned too late that thermocouples do not have to be entirely the coupled conductors and can have copper legs. If I was going to build another thermocouple for this purpose, I would use a short system of copper/chromium and attach it to a longer conductive thread set up. Overall, I am super happy with the results and I feel like there is a lot more for me to investigate here!