Thermoelectric generator for uncertain times

cranky wrote:

It occurred to me that cooling systems from a laptop could be used for cooling. Maybe the efficiency could be improved, and the elements are free, easily available and ... terribly unsightly.

By increasing the temperature difference rather only the efficiency can be increased. If a burner is used, sufficient cooling will be provided by the flow of air to the flame. Elements are not cheap because copper oxide slowly reacts with moist air to form copper carbonate, commonly known as patina. To do it in a human way (that is how it is done) you would have to buy a constantan wire. The cheapest spool I've seen online is Wed. = 0.3 mm, length = 100 m. So for PLN 300 (you need to add a protective tube) you could make a source composed of 1000 thermocouples with a diameter of approx. 20 cm and a potential of approx. 55 mV/°C.
The permissible temperature of this system with protective tubes is 500 °C. So 55 mV•500 = 27 V. Anyone can calculate the internal resistance. The resistance of 1 m of 0.3 mm wire is 7 Ω. Add the resistance of the copper wire and we know everything about the efficiency of such a source.
A cheaper solution is the T thermocouple system (NiCr-NiAl) which can work without a sheath. You could do it here for 200 PLN. I do not count the costs of "crocheting".
If someone has a fantasy and cash, they can buy 1000 ready-made thermocouples in bulk, they will not have to weld the ends.

hmmmm. I have a constantan tape somewhere, 20mm wide and about 1m long, from old stock (Bomis). I have to look for. What shape would you need to cut?

Let me just remind you that such thermoelectric generators put on a kerosene lamp were produced in the USSR somewhere in the 1950s and powered a tube radio:
https://www.radiomuseum.org/m/metallamp_su_en_1.html

stachu_l wrote:

If in series, then 0.6A 48V and in parallel, then 6A 4.8V, assuming the same operating parameters in both cases - temperatures of the hot and cold sides of each cell.

A layman`s question: if we get as much as 48 V in a series connection, and the manufacturer of, for example, TEC1-12706 cells states that the maximum voltage for them is 15.4 V, will it not mean that all the cells in the system will burn?

It will not mean burning all the cells - unfortunately, as a layman, you need to familiarize yourself with the series connection of voltage sources and receivers (e.g. resistors) and study what voltages and where they occur.

I'm thinking of attaching a few of them to the radiator.