How do thermoelectrics work?

[hartmantc]

I understand that a current is generated when electrons move from an n type to a p type semiconductor. However, is this due to heat flowing through the system, or a temperature difference on either side of the thermoelectric generator?

I know at first these sound like the same thing, but I am working on a design where heat flow would be diverted around the thermoelectric in a thermal conductor, effectively "shorting" the system. This way almost no heat would flow through the TEG, but there would be a temperature difference between the sides of the system. So, will a thermoelectric generator with limited heat flowing through it but with a temperature difference between its hot side and cold side generate electricity?

Thanks!

 

[anorlunda]

There are pretty goo descriptions on Wilipedia. Have you read those?

https://en.m.wikipedia.org/wiki/Thermoelectric
https://en.m.wikipedia.org/wiki/Thermoelectric_generator

 

[jerromyjon]

So, will a thermoelectric generator with limited heat flowing through it but with a temperature difference between its hot side and cold side generate electricity?

I'm not sure what you mean by "limited heat flowing through it" but I am pretty sure the electricity generated would be relative to the difference of temperature "allowed" to flow through the generator. The temperatures on either side are the only relevant value for electricity generated, maintaining those temperatures are relative to maintaining a current.

 

[hartmantc]

Yes I have read those, but they are a bit ambiguous. They consistently mention a temperature difference being the driving force. But then the comparison is made between thermoelectrics and a heat engine, which required heat flow to do work.
I'm also confused by what you mean by temperature to flow through the generator. Temperature cannot flow, only heat, where temperature is the result of heat flow.

To try to clarify what I mean by limited heat flow, imagine a thermally conductive piece of metal with a thermoelectric encased in it. Thermoelectric are thermal insulators. Heat therefore would flow through the path of least resistance around the insulator and through the conductive metal. However, a temperature difference between the sides of the thermoelectric would still be established.

 

[anorlunda]

Yes I have read those, but they are a bit ambiguous. They consistently mention a temperature difference being the driving force. But then the comparison is made between thermoelectrics and a heat engine, which required heat flow to do work.
I'm also confused by what you mean by temperature to flow through the generator. Temperature cannot flow, only heat, where temperature is the result of heat flow.

To try to clarify what I mean by limited heat flow, imagine a thermally conductive piece of metal with a thermoelectric encased in it. Thermoelectric are thermal insulators. Heat therefore would flow through the path of least resistance around the insulator and through the conductive metal. However, a temperature difference between the sides of the thermoelectric would still be established.

Heat engines also require temperature differences. Indeed, heat flow requires temperature differences. Do you think heat flow and temperature differences are independent of each other?

 

[hartmantc]

Yes I understand this. In the design, we will have a thermal resistor, the thermoelectric, in parallel with a thermal conductor, a metal. If I draw an equivalent circuit, we will have a large resistor (TE) in parallel with a small resistor (metal). Heat will flow primarily through the small resistor, and there will be limited heat flow directly through the thermoelectric. That being said, there will still be a temperature difference between the two sides of the thermoelectric due to heat transferred through the metal.

 

[anorlunda]

So, the I don't understand your question. Using the electric analogy, if you have a voltage across a resistance, the current will be I=V/R. If you put a smaller resistance in parallel with that, the current doesn't change. It is still I=V/R.

 

[jerromyjon]

a temperature difference between the two sides of the thermoelectric

What is the purpose of the thermoelectric device? Is it to generate electricity or is it's purpose to sense the temperature difference of the hot and cold sides?

 

[JBA]

Generally, it is viewed as an electric generating device but if a voltmeter is placed in the circuit and calibrated correctly it might be used for a temperature differential measurement but I am not aware of any specific cases where this has been done (but I only have a limited exposure to the current applications of this device). This device is more commonly being used as a thermoelectric cooling device.

 

[jerromyjon]

This device is more commonly being used as a thermoelectric cooling device.

That was my only thought that made sense, as in perhaps intercooling high demand lithium cells?