I'm trying to build a DIY working model of TEG/TEC generator similar to the exhaust pipe designs that recover energy from extremely hot exhaust pipe or light those antique candle powered radios. I have 400 (1 inch) pieces of copper wire and 400 (1 inch) pieces of paper clip steel wire. I will twist each one of each together then solder so that one end is copper with the other end steel having 400 pieces with a 1/4 inch solder joint in the middle. I plan to then then take a large PVC pipe (4" only 1 end open, the bottom has a sealed cover, to make a cylinder bowl, insides paint with high temp flame resistant paint) and drill (or melt with soldering iron) 800 holes spiraling up the outside (closely spaced, like 1/4 inch apart to insert the copper/steel ends into from the inside so that the 1/4 inch solder point remains in the middle of the tube but not touching. I can then twist/solder each of the remaining 400 copper/steel wire connections together on the outside of the tube to create an alternating series of copper/steel twists with solder spiraling up on the outside as well as the inside. I'll close off the each hole with high temp auto sealant glue from the outside and place the 4" PVC cylinder bowl into another 6" PVC cylinder bowl and fill the external surrounding cavity with ice to cool the external solder points. I can then place several candles in the middle cavity to heat the internal solder points sticking out in the middle and cover as necessary to trap the heat inside. I'm not expecting to get much power from the positive/negative terminals (maybe 200mV.) The question I have is, which is better to be exposed to the hot/cold? Is it the 400 soldered joints or the 800 alternating copper/steel wires? Peltier modules look like steel cubes joined in a series alternating on the top and bottom by copper leads. Am I right in that they just 'look' deceiving, where the tops and bottoms are being joined by a 3rd soldered metal, and that the cubes are actually 2 different alternating types of metal? From what I can see that's what the diagrams show. If so then my design is correct, the soldered joint is supposed to be exposed to extreme temperature differentiation for better efficiency. Is this correct? Thanks.