Many have heard about the "magic" Peltier elements - with the passage of current through them, one side cools and the other heats up. It works in the opposite direction - if one side is heated and the other is cooled - electricity is generated. The Peltier effect has been known since 1834, but to this day, innovative products based on it do not cease to please us (we just need to remember that when generating electricity, like solar panels, there is a maximum power point, and if you work far from it, efficiency generation is greatly reduced).

Recently, the Chinese have put pressure on the Internet and flooded the Internet with their relatively cheap modules , so experimenting with them no longer takes too much money. The Chinese promise a maximum temperature difference between the hot and cold sides of 60-67 degrees. Hmmm ... And what if we take 5 elements, connect in series, then we should have 20C-67 * 5 = -315 degrees! But something tells me that everything is not so simple ...

Brief theory

The classic “Chinese” Peltier elements are 127 elements connected in series and soldered to the Al2O3 ceramic “printed circuit board”. Accordingly, if the operating voltage is 12V, then each element accounts for only 94mV. There are elements with a different number of consecutive elements, and accordingly a different voltage (for example, 5V).

It must be remembered that the Peltier element is not a resistor, its resistance is nonlinear, so if we apply 12V, we may not get 6 amps (for a 6 ampere element), the current may change depending on the temperature (but not too much ). Also at 5V (i.e. less than the nominal), the current will not be 2.5A, but less.
')
The amount of heat transferred is proportional to the current. But besides this, there is parasitic heating from the flow of current, and parasitic thermal conductivity - all this makes the Peltier element at least somewhat effective in very narrow conditions.

In addition, the amount of heat transferred is strongly dependent on the temperature difference between the surfaces. When the difference is 60-67 - heat transfer tends to 0, and at zero difference - 51 watts for 12 * 6 = 72 watts element. Obviously, this already does not allow to simply connect elements into a series - each next one needs to be smaller than the previous one, otherwise the coldest element will try to give more heat (72W) than the next stage element can pass through itself at the desired temperature difference (1 -51W).

Peltier elements are assembled with low melting solder with a melting point of 138 - so if an element accidentally remains without cooling and overheats, then it will be enough to solder one of the 127 * 2 contacts to throw the element into a landfill. Well, the elements are very fragile - both the ceramics and the cooling elements themselves - I accidentally tore 2 elements “along” because of the dry thermal paste that was dry:

We try

So, a small element - 5 * 2, a large one - 12 * 9. Heat pipe cooler, room temperature. Result: -19 degrees. Strange ... 20-67-67 = -114, but it turned out a measly -19 ...

The idea is to bring everything to the frosty air, but there is a problem - the cooler on the heat pipes cools well only if the temperature of the “hot” and “cold” sides of the cooler lies on opposite sides of the gas-liquid phase transition of the tube filler. In our case, this means that, in principle, the cooler is not able to cool anything below + 20C (since only thin walls of heat pipes work below). We'll have to go back to basics - to the whole-copper cooling system. And so that the limited performance of the cooler does not affect the measurements - we will add a kilogram copper plate - a heat accumulator.


The result is shocking - the same -19 with both one and two stages. Ambient temperature is -10. Those. with zero load, we barely squeezed a measly 9 degrees difference.

We roll out heavy artillery

It turned out that the cold storage plant # 7 was not far from me, and I decided to drop in with a cardboard box. He returned with 5 kilograms of dry ice (sublimation temperature -78C). We lower the copper structure there - we connect the current - at 12V the temperature instantly starts to rise, at 5V - it drops by 1 degree per second, and then it quickly grows. All hopes are broken ...

Conclusions and video for sweet

The effectiveness of ordinary Chinese elements Peltier rapidly drops at temperatures below zero. And if you can still cool the can of cola with apparent efficiency, the temperature below -20 cannot be achieved. And the problem is not in specific elements - I tried elements of different models from 3 different sellers - the behavior is the same. It looks like cryogenic stages need elements from other materials (and perhaps for each stage you need your own element material).

Well, with the remaining dry ice, you can do the following:


Ps. And if you mix dry ice with isopropyl alcohol, you get liquid nitrogen for the “poor” - flowers and the like are frozen and broken in it just as merrily. That's just due to the fact that alcohol does not boil when in contact with the skin - getting frostbite is much easier.