[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Peltier Tank cooling

> Date: Sat, 1 Jul 2000 13:05:59 -0400
> From: "Matthew R. Sprague" <mspragu2 at twcny_rr.com>
> Subject: Tank cooling and Peltier devices
> I really like the idea of mounting a Peltier device on the tank glass. Do
> you have any "off the cuff" ideas on the surface area that would be
> Do you think the glass between the the water and the device has sufficient
> heat conducting properties for this to actually work? IMHO, it would seem
> that cool spots would not be a problem if the area being cooled has the
> discharge from a power head or filter outlet pointed at it.

I'm sorry that I can't give you any definitive information about the area
needed for cooling.  Peltier devices vary a lot in how they are built.  You
can stack one module on top of another to increase their cooling capacity,
so they would vary.  I also agree that cooling thru the might would be a bit
inefficient, but then every aquarium heater I've ever seen transfers heat
thru a glass tube.  I would think that the glass itself would act as a sort
of heat sink.  It would be slow to cool when the device was turned on and
slow to heat back up when the device was turned off.  Such a system would
act a lot like a radiant floor heating system.  A potential problem I see is
that in a hot humid environment the glass is going to sweat on the outside.

I believe if I was trying such a system today, I would get some Melcor CP
series thermoelectric modules about 2" square and glue them in a row along
the top back of the tank, leaving 2" clear on either side.  As the water
cooled along the top it would sink and create a slight current, preventing
cool spots.  I would try glueing them to the glass with a liquid metal type
glue, something that is a good conductor.  I would use heat sinks on the hot
side of the modules, but I wouldn't use stacked modules.  I wouldn't want to
create spots of intense cold in a sheet of glass.  I wouldn't bother with a
thermostat until it seemd necessary.  Since Melcor modules can be wired in
parallel, it would be a fairly simple matter to unplug some or add another
row if it became necessary.

On the melcor site, www.melcor.com/cpseries.htm  you can see that the model
CP 5-31-10L is about two inches square.  You can also see a lot of
information on their heat transfer capabilities.  I would have to study it a
lot more to even hazard a guess as to what it would take to cool a tank.
Maybe someone who knows something about these things will read this and clue
us in.