Electricity from heat - Seebeck effect and Peltier devices

[sunshinekid]

Hi All

I've not posted for a while, but just found an old peltier device I bought some time ago and thought I would try and make some leccy with it.  I sandwiched it between two round 6 inch steel plates along with some heat sink compound and a high temperature soldering mat in the middle to protect the 'cool side' from heat.

Then I plonked it on the woodburner which was running around 150 degrees C and placed a baking tray on top full of ice cubes, so it looked like this:

    ttttttttttttttttttttttttttttttt
    dddddddddddddddddddddd
mmmmmmPPPPPmmmmmmmmm
    dddddddddddddddddddddd

where: t is the tin, d is the metal disc m is the solder mat and P is the peltier device.

The results weren't very good - around 1 volt at 1.5 amps so a miniscule 1.5 watts.  After checking round the net, it seems that some Peltier devices are optimised for heating/cooling and some are optimised for generating leccy.  I've found a company that sells the latter so I'll order a couple and have a play.

If people haven't come across these devices before, they are normally used as heat pumps for CPU cooling in laptops.  When you apply a current, one side cools and the other heats up as heat is pumped from one side to the other.  The reverse is also true.  If you can heat one side and cool the other, it will generate electricity.  It's how the ecofans work, although they are a bit pricey!

Onward and upward.....

Steve

[wyleu]

You can run a mobile phone at night off this sort of power and that initself is an object worth having. It all goes to make you appreciate the power that is so instantly available nowadays from the wall sockets. It helps if you have a specific load that you would like to power in this particular way because that means you are designing for an objective rather than just trying to strive against an undeclared but generally unsatisfactory target. It ain't going to drive the fridge but a little bit of LED lighting in a power cut would be mighty popular.

[Bill H]

Hi Steve,

we kicked this around a bit early last year.

http://navitron.org.uk/forum/index.php?topic=1029.0


Seems like a good way to make a few watt-hours  - the peltier devices are getting cheaper.   Would be great if someone would build a unit to just sit on the top or sides of a woodburner....

I think the heatsink is the weakest link. 

Bill

[Gary T]

Not sure of the link, but I am sure that P.V devices have in the past been made and optimised for infrared. I seem to recall one such which could sit inside a wood burner and generate around 100W.

[Ivan]

I think there'd be a lot of interest in a 100W unit runing froma woodstove - please post the link if you find it.

[billi]

 i dreamt of a field arround my house with horses eating the grass and have jackets on , with peltier elements .......feeding my battery bank

the prices for them in relation to the output are not payable in the moment   


end of the dream...


but  is there a wake up ?

[sunshinekid]

Hi All

Looks like the rumours are true....  The power-producing Peltier devices I ordered have just arrived so I did a quick 'bodge test'.  I've run out of heat sink compound, but managed to scrape bit off the last (rubbish) peltier, popped some more ice cubes in my wife's cake tin and popped it on the stove.

The stove is currently running around 200 degrees C and the new Peltier is producing around 6 volts with a short circuit current of 600mA.  Much better.  This quick test is enough for me to try and improve the experiment.  I'll buy some more heat sink compound, some domestic cooker thermal insulation and clamp the Peltier between the two plates with a little force to improve the thermal conductivity.

Then I'll get some low wattage 6v bulbs and check it under sensible loads and report back.

My guess is a single device would be enough to charge a mobile phone.

Just a quick reality check.  When sitting on top of an 11KW stove, and producing a few watts of power, the efficiency couldn't be called high, but I'll push on and see what I can do.  Hopefully I'll be off-grid by the weekend :-)

Just for interest, the devices I bought are HT devices.  These have a higher temperature tolerance than normal Peltiers.  Normal ones are rated up to 175 Degrees C and these are rated up to 220 degrees C and also have teflon coated connecting wires.  Ideal for a woodburner.

The really nice thing is that they don't really use heat, just transfer it.  Obvioulsly with my example, I'm wasting heat by melting ice cubes, but if I built a small water-cooled unit, copper pipes could remove the heat from the Peltier which could be dissipated in a small radiator/heat exchanger in the kitchen, which is the coldest place in the house.  So simply moving heat from one place to another, you can make leccy.  Num num num.

Of course the project could always collapse under it's own weight before it's started! lol!

Keep Smilin'

Steve

ps we're having a woodburner with back boiler fitted next week.  Gas up by 10%?  Who cares? :-)

[sunshinekid]

Oops - forgot to mention the price.  The HT (High Temperature) devices were quite cheap.  Here are the prices:

MODEL HT1-12710 (40mm - 127 couple - Power Generation)
1-5 pieces $12.80 each
6-24 pieces $11.60 each
25-99 pieces $10.60 each
100-499 pieces $9.60 each
500+ pieces $8.40 each

MODEL HT1-19802 (37mm - 198 couple - Power Generation)
1-5 pieces $13.10 each
6-24 pieces $11.90 each
25-99 pieces $10.90 each
100-499 pieces $9.90 each
500+ pieces $8.70 each

International Shipments
Shipping and handling charges for International shipments (outside of the US and Canada) are $20.00 for the first 5 pieces and $1.00 for each additional piece.

[Andyr]

Hi,
   the IR PV is still years away from the market, but check out the attached. Lets hope it makes it to market one day.   It may be usable on stoves.   
http://www.inl.gov/featurestories/2007-12-17.shtml

  Also check out this group, again like most high tech development, they may or may not be pie in the sky 
http://www.powerchips.gi/



      Andy

[billi]

sunshinekid

sounds good... and dont get me wrong i like experementing a lot...

i think it would be great to produce power from a free or anyhow available heatsource and then who cares about efficiency   ,  but  <( i understand its just testing and not the end)   if you get 6 v  with 0.6A or about half voltage when charging so about 50 pieces  to produce 100 watt   i guess

not too bad  if heatsource is free.....


all the best   billi

[KenB]

Sunshinekid, Billi,

These devices would be ideal for utilising some of the waste heat that otherwise goes straight up the flue.

If a woodstove is perhaps 70% efficient - then a 11kW stove is losing 3.3kW up the flue.

Several of these devices could be fitted to a radial fin assembly set into say a 12" section of fluepipe, with watercooling to their cold.
sides.

Even if they are perhaps only 3% efficient, you would still get about 100W from the assembly.

There is a company in the US (Hi-Z)  that are developing a 1kW thermopile to fit in a car or truck exhaust manifold.

http://www.hi-z.com/



Ken

[sunshinekid]

Hi All an thanks for your replies.

Bill it's not quite as good as I had hoped so far....  The mistake I made was to measure the open circuit voltage and short circuit current.  A proper test is to produce some voltage and current.  I just bought a pushbike 6v 150 mA bulb and tried the circuit again and with the stove at 200 degrees, it produced 6.2v at 140mA.  With a 500mA bulb, the voltage dropped to 3.5v and 300mA.  Both results show about 1 watt produced.  I've also cooked my solder mat :-)

It was good though, to see a real lamp lit from the heat from the stove.

Does anyone know of some high temperature insulating sheet which would cope with around 300 degrees C?  I'd like to use this to insulate between the hot and cold sides of the Peltier.

Not giving up yet though, I've just ordered a nice big Northbridge CPU heatsink and some thermal paste so I'll make a 'proper job' of it, clamping the sink to the Peltier which should improve thermal transfer.  The ad for the heatsink didn't specifically say, but the thermal dissipation capacity may be an improvement on the cake tin.  Also using the cake tin stops you making cakes :-)

I'll also try using a cooling fan.  It may be that the improvement that gives is greater than the power consumed by the fan.

Best do some real work now....

Cheers

Steve

[KenB]

Sunshinekid,

I used to work for a scientific instrument maker that made Peltier heating plates for heating and cooling laboratory samples. The plate was 4" diameter and could maintain the sample at between -15C and +99C to within 0.1C accuracy.

The cold side of the peltier was cooled by a water cooled heatsink - a block of brass with water channels milled into it.

I friend suggested a novel approach - a square section pyramid shaped block of metal to convey the heat into the thermoelectric module. The TE module clamped to the base of the pyramid.

The apex of the pyramid would be red hot - say 600C but the ever increasing cross sectional area of the pyramid profile would reduce the thermal flux to the safe limit for the T.E. matrix.

I never got around to trying it though.


Ken

[sunshinekid]

Hi KenB

Thanks for the tips.  I quite like the pyramid idea as while I was writing my last post, the woodburner went up to over 300 C and the solder melted on the wire connections and dropped off!  Luckily the edges of the device are filled with silicon rubber which cuts out quite easily, revealing the metal connection lugs.  The wires soldered back on no problem and the device works again, illuminating the room with a 6v bulb :-)

I'll update once the heatsink is here......

Cheers

Steve

[Gary T]

The absolute best insulator if you can get hold of a piece is silica aerogel. This will tolerate a naked flame on one side,while you could put your hand on the other without getting hurt