generating power with peltier cells

[Simon]

I'm thinking of playing with some peltier cells and making electricity with them. Is this feasable ? what sort of generation do they have versus their power consumption as heaters/coolers.

What is the max temperature you can heat a peltier to ?

[sacherjj]

Generally, the setup people think of is heating by the sun on one side and not heating the other.  Typical efficiencies are very poor.  Around 5%.  You need to maintain a very cool second side to do much better than that.  $/watt come out to about the same as solar, but you would need to get the surface very hot to generate some real power.  So parabolic mirrors or lenses, etc.

You will also find the voltage is very low, as compared to solar.  Look at the Seebeck coefficient.  It is typically uV/K.  So you need a very large temperature difference to get any usable voltage.  Although I'm not sure what the current levels are.

[Simon]

I was thinking of a small wood furnace, solar just won't do it I realize that

[ejeffrey]

I have seen this done successfully.  The efficiency isn't great, but the application I saw was to use the heat from a small stove to run a circulation fan (basically a 120 mm computer case fan).  The fan blew over the radiator on the 'cold' side of the peltier to help keep delta-T high to keep the generated power up.

I think the maximum temperature is usually limited by the packaging -- especially contact wire insulation and solder.  So you probably have to shop specifically for high temperature devices.

[insurgent]

I've been looking into thermoelectrics recently but am by no means an experts on the subject. From what I've seen, the Peltier devices can be used to generate electricity but will not last very long when used for that purpose.
Apparently for power generation, you want use a thermoelectric generator (TEG) device that exploits the Seeback effect as opposed to the Peltier effect (they are very closely related effects). I'm not sure if there is a difference in efficiency between the two devices.
If you just want to mess around and have Peliers lying around, I'd use them. If you desire to design a project that will last, you should probably look at the TEG's. There are some inexpensive ones on ebay in the $15-$20 range.

[Simon]

there are some very cheap devices on ebay with "TEG" in the title but they only give specs for heating and cooling and nothing about power generating capabilities and limits

[Zad]

From my experiments they seem to produce slightly less than bugger all power. The solder holding the elements in place tends to soften quite early and distort the array.

[Simon]

I did read an essay about a commercial unit that cost over 1000 whatever (I think he was in Canada) and it was quite clever. As i've got wood to burn and potentially these things are cheap now I thought it would be interesting to have a play and see what i can come up with.

[johnmx]

Read the LT Magazine of October 2011.
All you need to know is there.

I’m going to make this circuit. I already have all components, including the LTC3108 and several transformers from Coilcraft. All I need is time to design the PCB. But first I want to finish other projects.

http://cds.linear.com/docs/LT%20Journal/LTJournal_V20N3_Oct10.pdf

I recommend everyone to read the LT Magazines, now called LT Journal of Analog Innovation:
http://www.linear.com/designtools/lt_journal.php

[Zero999]

You could water cool the other side but I doubt you'd get enough power to do more than run a pump for the water.

A Peltier heat engine is very inefficient, you're better off running a small steam turbine, the only advantages of Peltier are cost, weight noise and that there are no moving parts.

[saturation]

This was recently in the news:

http://www.tuaw.com/2011/06/14/hatsuden-nabe-thermoelectric-cookpot-keeps-your-iphone-battery-c/

[Zero999]

A novel idea but look at the price!

I suppose there are other applications. You could use a Pelter to power a pump for a power shower for use on a camp site where there's no access to mains water pressure, just a tank or a portable dishwasher.

[Time]

You could harvest power from a car engine if you directed air flow over one side of the plate.

[saturation]

Here's more, going back > 100 years:

http://www.neazoi.com/technology/thermocouple.htm

[johnwa]

Here's more, going back > 100 years:

http://www.neazoi.com/technology/thermocouple.htm

I think the original source of that article may be over here: http://www.douglas-self.com/MUSEUM/POWER/thermoelectric/thermoelectric.htm. The rest of Doug's site is worth checking out too if you haven't seen it.

It probably wouldn't be too difficult to build something along these designs with access to suitable metalworking facilities. Of course, there are a few constraints to the design: you would need readily available, cheap metals with high thermoelectric coefficients, and a construction with minimum electrical resistance between ends of the elements, but high thermal resistance. Given the quantities involved to generate a reasonable voltage, construction might be a bit tedious, but it would be an interesting project. I might have to put it on the end of the list of things to do...