Thermoelectric generator for uncertain times

Krzysztof Kamienski wrote:

This is a semiconductor Peltier, designed for cooling, i.e. working at relatively low temperatures, i.e. with very low efficiency as a generator.

It seems to me that replacing the air heat sink of the cool part with a water cooler would effectively lower the operating temperature of the device, and thus increase the efficiency. I would suggest using a copper heatsink on the hot side with the thermal paste replaced with liquid metal (the paste will dry quickly at over 100 ° C, and due to thermal expansion and "working" of the system under the influence of temperature changes, it will quickly start to crumble). On the cold side, a traditional paste might even suffice, but as a heat sink, use the old CPU cooler, made in such a way that the heat pipes would receive heat from the Peltier, and they would have aluminum fins immersed in the water. Thanks to this, the operating temperature of the cold side, even with boiling water, will not exceed 100 degrees, and in practice (with water changes) it will be even lower.

For higher powers, yes, but the device will also be larger, and will have wearing parts. The one above is small, pocket-sized, and won't break by itself.

gulson wrote:

Charging the phone during a candle power failure is a very important matter and such a solution should be stored in every apartment in the event of W.

But why look for such extreme applications? Once upon a time I saw a device that looked like a mug (in which a fire was lit) charging a connected device. For longer camps as he found.

First of all, it should be remembered that a thermocouple is an extremely simple device that works on very simple principles.
Its voltage depends only on the temperature difference between the cold side and the hot side, and when loaded, additionally on the resistance of the entire circuit. More cells connected in series means higher voltages. The resistance is fairly constant (ignoring the temperature coefficient). However, how much electricity can be drawn from the cell?

First of all, we need to look at such a Peltier module as pairs of cells:

Let us assume that the "connection plates" are such a teleport. They do not take part in the release of heat, but connect individual elements. The energy "disappears" or "appears" not in the connection plates, but at the junction of two semiconductors. If the current flows from a P-type semiconductor to an N-type semiconductor, the temperature at their junction increases, and if from N to P, it drops. And the only thing that connects both parts is the way the electric current flows. We can imagine it as having 2 separate Peltier modules connected to each other in series. We put one of them in warm, the other in a cold refrigerator. A module placed in the heat generates electricity, and the latter consumes it and produces heat.

So we come to something like thermal conductivity. It is the amount of thermal power that flows through an element made of a given material and with a given cross-section and thickness, at a given temperature difference. In the case of Peltier modules, this conductivity is variable. An unloaded module will have a specific conductivity, resulting from the conductivity of the silicon it is made of (and the minor materials), but still relatively high. This conductivity is a loss, and it results in the efficiency of the module. When a module is short-circuited, the thermal conductivity will increase sharply (see comparison to two separate modules). When a voltage is applied to the module, which causes the current to flow in the same direction as when the module is short-circuited, the conductivity will increase even more. On the other hand, if this voltage is applied in the opposite direction, the conductivity will be ... negative? Contrary to appearances, yes, because the heat will magically pass from the cold side to the warm side. It's a bit like heavy objects suddenly start being pushed off the ground instead of falling onto it.
So, we load the Peltier module working as a generator with a receiver. Current flows through it and the thermal conductivity suddenly increases, so the temperature difference drops, which in turn reduces the voltage. This can be easily seen in the oscillogram, which of course I don't have. After connecting the receiver, the voltage will drop in a fraction of a second - this is a decrease resulting from the electrical resistance of the cell. The voltage will then begin to drop slowly downward until it reaches a certain level - a decrease resulting from increased thermal conductivity, i.e. a decrease in the temperature difference.

But I wrote a letter. I'm going to solder

I think I go to bad campsites, but where on earth do I get a packet of tourist gas cylinders in the middle of the forest? Mr. Sticks, this is it

Krzysztof Kamienski wrote:

Iron / steel wire (for fences) is cheap and so is Constantan wire (for radiators).

"L-type thermocouple (Fe-CuNi) is made of iron (Fe) in connection with constantane (CuNi). The range of application of the J-type thermocouple is typically -40... + 750 ° C, and the sensitivity is - 55 uV / ° C. J-type thermocouples (Fe-CuNi) are designed for temperature measurement in an inert, reducing and oxidizing atmosphere, as well as in a vacuum. "
~ Source unknown / stolen.
That is, with a temperature difference of 300 ° (flame) -50 ° (warm water), you will get a staggering 250 * 55 uV = 13.75 mV. What will you do with so much tension?
This is why in the device you wrote about, there were hundreds of these thermocouples connected in series.