Hello good people, a quick question if I may about thermo-dynamics.
As I understand it, the Peltier effect involves the phenomena of sending electrical current through a device with the net effect of that device lowering it’s temperature relative to the ambient.
Unless I’ve missed something jaw-dropping obvious doesn’t this challenge Newton’s law of thermo dynamics? As I understand it a decrease in temperature, relative or absolute, is an increase in Entropy, if you apply energy to any system it’s results must be quantifiable.
Yours in cat like curiosity in anticipation.
Peter
Peltier devices act like a bunch of itty bitty heat pumps all arranged in a square together. When you run current through the entire assembly, it moves heat from one side to the other. Reverse the direction of current, and the temperature reverses as well (the hot side gets cold and the cold side gets hot). Since peltier devices aren’t superconductors, there is additional heat generated by the device itself.
If the waste heat isn’t dissipated from the hot side, the peltier will get hotter and hotter until it overheats.
They also work in reverse. If you apply a temperature differential between both sides, they generate electricity.
As far as their efficiency goes, they kinda suck. But they can be made small and cheap, which makes them extremely useful little buggers. There is also a maximum temperature differential that they can generate, but they can be stacked to create an even greater temperature differential.
Peltiers cannot suck heat from the environment. Like all such devices, they are heat pumps, and have a hot side and cold side. Peltier units in practical use have heat sinks, water cooling systems, or the like to keep the hot side as close to ambient as possible. The cold side can then dip below the ambient temperature, but the total entropy has still increased.
A Peltier device doesn’t violate the laws of thermodynamics any more than any other refrigerator. Heat can be pumped out of something hot, but you can’t do it without a source of low entropy energy.
P.S. Newton lived long before the development of thermodynamics.
There are important details missing from your description of a Peltier device.
Imagine you form a loop out of two segments of different types of metal. There will be two junctions where the two types of metal make contact with each other to form the loop.
Now you cut one of the pieces of wire (somewhere between the two bi-metal junctions), and you connect the loose ends to a battery, which drives electrical current around the loop.
As the electrons pass through one of those bi-metal junctions, they will take some heat with them and carry it to the other bimetal junction. In the absense of any heat transfer between this system and the wider environment, one junction will get colder and colder while the other will get hotter and hotter.
The battery is supplying energy to this circuit in order to make this process happen; the energy it provides manifests as additional heat at the “hot” bi-metal junction (i.e. the cold bimetal junction gets colder, but the hot bimetal junction gets a LOT hotter). So while the entropy of the two wires is decreasing (as one junction gets hot & the other gets cold), the total entropy of the wires PLUS the battery is actually increasing, in accordance with the second law of thermodynamics (which has nothing to do with Newton).
Once the junctions are at differing temperatures, you can remove the battery and observe a voltage between the loose ends of the wires. You can use this voltage as a proxy to measure the temperature difference between the two junctions. You can even use it to generate electrical power. It’s handy for some applications because there are no moving parts. However, it’s not very efficient, and thermodynamics says you will not get more energy out of it than you originally put into it with your battery, i.e. this is not a perpetual-motion/“free energy” thing.
Why does one side get colder (colder than ambient?) rather than one side getting hot and one side getting hot but slightly less hot? Like say ambient is 70F, will one side get to, say, 200F and the other side to 300F or will one side drop below 70F?
Because you have cooling fins and a small fan blowing ambient air over the hot side. This brings its temperature down close to ambient, and so the cold side is lower than ambient.
Thank you all for your answers, re Zen Beams last point, I have come across the effect wrt PC cpu cooling and it’s advantage of being a totally silent method, ie no fan acting upon the ‘hot’ section of the heat pump. Have I misunderstood the design?
Peter
Heat will flow from the hot side to the environment no matter what you do. Fans and such will make the heat flow more efficiently, but even without, there will still be heat flow, and so the cold side will be some amount cooler than the environment.
I’ve seen peltier coolers both with fans and without fans. I’ve also seen non-peltier completely fanless heat sinks. One of the big things in the past few years seems to be “heat pipes”, which can be used as standalone processor coolers or can also be used in conjunction with peltier devices.
In addition to being silent, things like peltiers and heat pipes are very desirable in industrial computers. Since these types of computers have to work on factory floors and similar harsh environments, fans can be problematic since your typical fan/heat sink combo will get completely choked with dust in a very short amount of time. Fanless technology is therefore very important for those types of environments.
And yeah, some folks just like quiet PCs. That works too.
By the way, peltier devices aren’t just for cooling PCs. They were originally developed for space satellite type uses. In consumer goods, they also show up in those little portable coolers that plug into your cigarette lighter, and also in some dehumidifiers. They even show up in some high end digital cameras, where they are used to cool down the cameras imaging CCD to reduce thermal noise and make a clearer picture.
Thanks for that, one thing has always struck as rather odd surely cooling an Earth Satellite is pretty easy. Admittedly local space isn’t at absolute zero but it isn’t the surface of Venus, or the pressure for that matter.
So… if you need to cool the internals of a Earth restricted Satellite why not just circulate coolant through heat distribution fins, use the pressure on the ‘hot’ side to drive a circulation pump.
As an aside to Computer comp geek, hi there, do you know of any silent closed circulation cooling systems suitable for a domestic situation, I’m thinking of a Liquid/Gas/Evaporation system rather than standard Liquid cooling.
I realize that Server farms cool the ambient temp, and probably humidity as well, but what I’m after is the quietist method of moving heat from the surface of, probably, an Intel i7 CPU, o-clocked to within a Gnats nuts. Sig below should show my current air cooled approach.
P
Cooling satellites is hard because vacuum is a pretty good insulator. There’s no air or other medium to carry heat away from cooling fins. Satellites and spacecraft use radiator panels, which can only shed heat by thermal radiation and need to be fairly large in some cases. In the case of the International Space Station, the cooling system needs active pumping to drive liquid ammonia over cooling plates and then out through the fairly large white radiator panels you can see on a lot of the photos of the station,
Thanks Andrew, wow, if anything is counter-intuitive, it’s the idea that cooling a satellite is problematic, serves me right for taking an atmosphere for granted…
P
As an adjunct afaik conduction, convection and radiation are the only means of heat transfer, is there something particular about an almost hard vacuam that makes thermal equilibrium via radiation complex?
P
Well there’s the Sun. A relatively small spot in the sky that throws out a lot of heat and light, surrounded by otherwise very cold space. Things illuminated by it get hot. Things in shadow get cold. You’re orbiting the Earth, and what’s in Sun and what’s in shadow changes.
I’m not sure what you mean by “complex” here pdunderhill, but there’s also the issue of thermal conductivity within a satellite not being infinitely fast. So the side that points towards the Sun will get pretty hot, even if the average temperature of an object at thermal equilibrium at that point in space is acceptable for your purposes.