Can I use thermoelectric generators to generate heat from electronic components?

[maryammagdy]

I'm looking for a renewable source of energy to conserve power beside my main battery. I was thinking of why we can't use the heat generated by the chips in the circuit as a thermoelectric generator? Does this make sense? I haven't seen anyone using that or mentioning it. Discussions are welcome of how and why I can or cannot achieve this. My application is mainly a GPS tracker with accelerometer.

 

[Nigel Goodwin]

I suggest you try googling 'perpetual energy' - and you'd probably use more power trying to harvest the heat from the chips.

 

[Pommie]

Most systems try to cool things with fans etc. Trying to harvest that heat will stop the heat escaping.

Mike.
P.S. don't do it.

 

[dknguyen]

It's wildly expensive and wildly inefficient for what you get out of it because the temperature gradient is too low.

Look up what a radio thermal nuclear generator is and note it's efficiency when using the fuel that it uses and running at the temperatures that it does.

You're better off just trying to power it off movement.

 

[Morgann]

The energy produced by the electronics itself is not too much.
Machines used to collect heat may consume more energy and the energy lost during the collection process is greater.

 

[dknguyen]

Thermocouples do this. But are so inefficient that the power produced can only be used to sense temperature, and even then it takes a lot of amplification and signal processing circuitry.

The only power-producing application I know of that what is called a radioisotope thermal generator (RTG). They surround a block of radioactive material that produces heat as it decays with thermocouples. They use this to power deep space probes like Voyager 1 and 2. No moving parts. Long lasting and very reliable, but very very inefficient. The only other application I know of is the USSR powering lighthouses with them.

Oh yeah. Efficiency increases as temperature gradient increases so with the already bad efficiency of thermocouples, you need to run the thing much hotter than the surroundings to recoup any efficiency you can.

It's a bit like asking why we don;t use geothermal energy everywhere. It's just not feasible or practical in most cases.