Peltier Questions

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View

I'm trying to figure out how much power I could generate from a peltier
device from a given temperature difference. With water as the medium on each
side of the peltier, what temperature difference would I need to have to
power a small water pump to circulate the water round my CPU cooler block?

In other words - could a cooling system be designed to actually drive/power
itself from the very heat it is trying to remove? Perhaps convection would
be required to start or subsedise the equation, but do 'we' think this is

Could this be scaled up to drive a water pump for a small pool, driven by a
peltier that is powered by sun energy heating water on one side in a black,
heat absorbing tank and cold water from the pool on the other side?

Re: Peltier Questions

Quoted text here. Click to load it

It seems it would depend on the size of your peltier.  A
single peltier would produce a very low voltage so you would
need several in series, an expensive project considering how
little electricity costs, and quite large unless it were (an
even more expensive) specialized device like this:

Quoted text here. Click to load it

Depends on the water pump and flow rate desired?  If you are
intending to use waste heat from the CPU you have a problem,
that to create the temperature gradient needed you would
have to let the CPU get very hot  or create a sub-ambient
temperature on the cold side, again using externally
supplied energy to pump heat off the cold side.

Quoted text here. Click to load it

Possible yes... on a NASA budget and developed for years.
Not the best solution here on earth.

Quoted text here. Click to load it

No, practically speaking the size and cost of doing it would
negate any real benefit.  Peltiers are not very efficient,
you'd be better off with solar panels which aren't very
efficient either... even for powering the computer water
pump from the slight bit of light in the computer room
instead of sunlight by using more/larger panels.

Re: Peltier Questions

GT wrote:
Quoted text here. Click to load it

There are devices, that demonstrate the kind of temperature differential
and heat flux needed for decent power generation.,104,53221

"Thermoelectrics to replace car alternators and improve MPG" /

But to address the problem, of "powering a cooling system", you
need to compare the efficiency of water cooling to heatpipes.
There is no comparison. Heat pipes are much more efficiency
at heat flux transport, than an equivalent water cooling system.
And it's because of the larger amount of heat per gram of material
involved in a phase change. A heat pipe uses phase change as the
transport mechanism. The only thing I don't know, is to what
distance the capillary effect, can be used to transport
condensed liquid back to the source. Water cooling systems
are used, because they're relatively cheap, and easy to
install. Custom heat pipes, bent to fit your installation,
wouldn't be nearly as easy to set up. The capillary effect
is available, when the inside surface of the heat pipe is
sintered. The condensed liquid can actually go uphill, but
with some loss of effectiveness.

Zalman built a couple computer cases, that use heatpipes to
connect heat sources, to the outer wall of the computer case.
The computer is a giant convection cooler. The case pictured
here, was around $1000.

If you have a heat source and sink, there is always stuff like
this. The Stirling engine allows you to get close to the
Carnot limit, although at the expense of a non-compact solution.
Some day, we may see more of these.


Re: Peltier Questions

Quoted text here. Click to load it

Ahh - might as well forget it I think. If 170 degree difference is only
enough to drive a small, effecient fan, then water pumps are out of the
question for a 30-50 degree temperature difference! Looks like I'd be pushed
to drive a few LEDs with that!

Thanks for the links and info, but I'll give up that idea straight away!

Site Timeline