Peltiers get hot on one side and cld one the other. They cannot go
below 0C. You need to heatsink them well and maybe cool the hot
side )with fan and heatsink)., Maybe having a fan runnning over it,
and the other side of the Peltier gets cold.They change polarity (heat
and cold) depending on the way you wire then up. I used 4mm perspex tp
make a decent heat exchanger. It is worth the effort.
Peltier devices, are effectively electrical heat pumps. They only get cold
on one side, if the heat being generated on the other is removed. This
heat comprises the heat being pumped, and the heat generated by the
Peltier itself. They can go way below 0C. There is a balance between how
much heat they can move, and the temperature difference between the two
sides of the module. Typically modules can achieve up to about 70C delta,
but at this, they are able to move almost no heat. Conversely, depending
on the size of the module, there will be a point where the temperature
difference reaches zero, for a given amount of heat to move (Qmax).
Typically a Peltier will draw about double the power, that they can move.
So, in the example, the element is drawing about 72W. This would be
typical of a module able to move perhaps 40W of heat at a 0C delta. Add a
heatsink, with a temperature rise of perhape 0.1C/W, fan cooled to the
'hot' side, and this will need to dissipate about 112W of heat in this
'worst case'. It'll then run up to about 11C above ambient. If you have
something that only generates 15W of heat, and attach this to the 'cold'
side, expect the delta C to be perhaps 45C, and the component to cool to
about 34C below the ambient. The actual delta will depend on the module,
but this would be a typical value for a module this size. The 'lowside'
temperature can be well below zero. There is nothing 'magic' for a Peltier
module about 0C. For even higher delta's, cascade two modules. Remember
though, that the 'hot' module, has to pump not only the original heat, but
the heat being added by the first module...
I have one sitting here on the bench, which is currently holding a couple
of semiconductor junctions at -40C. They are only generating about 1W of
heat, but to maintain this temperature (currently 62C below ambient), with
the two stage cooler attached, over 100W of power is being fed into the
cooler module, and the heatsink size reflects this...
The best way to 'characterise' the module, would be to measure the
temperatures on both sides of the junction, with known amount of heat
being applied to the cold side (resistor fed from a controlled voltage
source). You can then measure the delta T, for different power inputs, and
build your own graph for the behaviour of the module.
Best Wishes
