Thermoelectrics

[Jaffah]

Hey, I am basing my project off https://www.sciencebuddies.org/science- ... p031.shtml. I am investigating which combination of different conductive materials and different temperatures between two junctions will produce the highest voltage. To test the concept first and get everything set up, I started by forming two junctions with copper and steel galvanized wire. Both have have a 28 gauge. When one junction is put directly on a flame, with the voltmeter set to 2000uA, the voltage fluctuates too much to where I can't even tell what the average voltage is. Currently the junctions are about 1cm long and are just formed by coiling one of the wires around the other. The setup is: 10cm steel galvanized wire - 20cm copper wire - 10cm steel galvanized wire. Is there anything I can alter to help decrease the range of fluctuation like using different gauges, soldering the wires, using shorter or longer wire? Solutions/suggestions would really be appreciated as once this problem is solved, the rest of the experimentation is pretty straightforward.

Marc

[kgudger]

Hello and welcome to the forum! From looking at the experiment you mentioned, and reading up on the other web sites, it appears that you should be able to get this to work by just twisting the metal wires together. Have you also seen this website? http://en.wikipedia.org/wiki/Thermocouple.

From your question, I noticed a problem, but it may just be a typographical one. You said

with the voltmeter set to 2000uA, the voltage fluctuates too much to where I can't even tell what the average voltage is

. If you are trying to read the voltage output (as the experiment suggests) your answer says that you are using the wrong scale on the multimeter. You are trying to measure current (uA) instead of voltage (V). The experiment suggests

Set your voltmeter to the most sensitive DC voltage (usually 50 or 200 mV)

. Please let us know if this helps. Thanks.

Keith

[Craig_Bridge]

The setup is: 10cm steel galvanized wire - 20cm copper wire - 10cm steel galvanized wire. Is there anything I can alter to help decrease the range of fluctuation like using different gauges, soldering the wires, using shorter or longer wire?

I don't think you fully understand your setup. The leads to your meter are probably not galvanized wire so you have more dissimilar metal junctions: meter lead material to galvanized wire, galvanized wiere to copper wire, copper wire to galvanized wire, galvanized wire to meter lead material. Galvainzed wire only has a thin plating of tin/lead over some base material (usually copper or steel). It is easy to scratch this plating. If you are using galvanized copper wire, they you may have ended up with a copper to copper connection.

with the voltmeter set to 2000uA, the voltage fluctuates too much to where I can't even tell what the average voltage is.

With your multi-meter set to the 2000uA scale, you are measuring current on a 2 mA scale (1000 uA = 1 mA). Measuring current in this kind of an experiment will be subject to changes in resistance of the connections as well as the thermocouple voltage. This tells me that you probably need to be twisting the wires together tightly using a pair of flat nosed pliers and something to hold the part of the wires not being twisted. You want at least 4 complete twists and you want both wires to be twisted, not just the softer wire twisted around the stiffer wire. You also need to make sure that the connections to your meter leads are also tight. If you can use a 20, 50, 100, or 200 mV DC scale, small changes in junction resistance will not be as significant to the voltage measurement unlike how it affects the current measurement.

[Jaffah]

Thanks for your replies. Your advice was really helpful and I have gotten my project to work successfully. I will definitely come back here if I have further questions. I really appreciate your time.