Thermoelectric generator - candle and cold water

[Magluvin]

Hey Conrad and TK

I havnt seen oil used with these yet.  I think a nice slug of aluminum with many thin fins, and a glob of heat sink compound does very well. These things ar not very thick from hot side to cold side. The edges are filled with silicone it looks like, white, maybe high temp, and looks just to keep the weather out from the matrix inside.

Literally if you put power to a naked chip, do not have your fingers on either side, because things happen fast. Ive only done it in short bursts and you can feel it happen, one side cold, and one hot. Its a strange experience, as it is not a common one.

I saw a vid where with a naked chip, frost develops on the cold side fairly quick.

Conrad

Maybe you could get 2 PC cpu heat sinks to help get rid of the heat in the water.

Use a dremel, or a jigsaw with a fin laminate blade and cut a square hole in the side of your  plastic water tube. Use on cpu heat sink with fins inside the tube with the water, with the flat mounting side of the heat sink facing out and seal it into the tube with silicone well to avoid leaks. So the tighter the heat sink fits in the cut out of the tube, the better mounting adn sealing will be. Then apply some heat sink compound to the flat surfaces of the mounted heat sink  and figure a way to mount the outer heat sink to that surface. 

Or use an aluminum water container. Preferably square with flat sides that cpu heat sinks can be mounted.

Or, you can just use the heat sink alone on the cold side of the chip. it is the end result in the end. ;] Heat into the air.

Any fan will make a difference in getting rid of the heat from the sink fins. More is better.

I used an aluminum block, 2.5x2.5x3/4 in between the hot side and its heat sink for the SS A/C unit, as a buffer of sorts, to give a larger volume of aluminum to absorb heat from the hot side. Ill post some pics tomorrow.

So my setup, as an AC unit, is running backwards compared to your gen. And the water is your aluminum block. ;]   It might be best to put a small heat sink on the side of the chip thats in the water, giving more surface area 'to the water'. ;]  Chip surface area < heat sink fin surface area.

MaGs

[Tito L. Oracion]

 


This is interesting though old but really interesting. 


i will try putting the cold side in a cooler so that cold will last longer. 


and i will try the iron that heats very long. and since heat is transferable then we can make it more stronger device. hmmmmm 
 


This is more efficient if we add some transistors and diode. hmmmmm 


I salute you contrad    


I think it much more better if we add some salt in the cooler to add extra cool.               
I think liquid nitrogen can be also but    lol

Do you know that to make heat, we only need two dissimilar metal  , why?, it acts like a battery then put some resistor. ok to prove it , try shorting the two hands of a battery wright?

[Magluvin]

This is interesting though old but really interesting. 


i will try putting the cold side in a cooler so that cold will last longer. 


and i will try the iron that heats very long. and since heat is transferable then we can make it more stronger device. hmmmmm 
 


This is more efficient if we add some transistors and diode. hmmmmm 


I salute you contrad    


I think it much more better if we add some salt in the cooler to add extra cool.               
I think liquid nitrogen can be also but    lol

Do you know that to make heat, we only need two dissimilar metal  , why?, it acts like a battery then put some resistor. ok to prove it , try shorting the two hands of a battery wright?

From what I understand, the chips/module cant be driven too hard with higher voltages than around 16v. Im reading about them more as i found them also being discussed in one of my power supply builder books, Regulated Power Supplies  by Gottlieb.  They can be used to help keep PS transistors cool. They use them in microprocessor cooling also.

great book. Even discusses magnet biased cores. ;]  If you pulse the primary of a magnet biased core in one polarity and measure the input/output, then reverse the pulse polarity and remeasure, one way causes more input and more output, and the other causes less input and less output. ;] or instead of reversing the pulse polarity, you can reverse the bias magnet polarity. The difference in in/out depending on polarities is caused by saturation effects, affected by the magnet in the core. ;]



MaGs

[TinselKoala]

Yep. I used that latter effect to improve the output of my early Steorn Orbo replication, Orbette. The only toroids I could find at the time were long cylinders instead of rings, and when I used biasing magnets of the right polarity, the thing really sang, without any increase in applied power. Well, it was a pulse motor of type 3, a core effect motor, so biasing the core makes great sense in that design.
It's also very interesting to play with a strong NdB around the toroid of a Joule Thief.

[Tito L. Oracion]

i think 30v is permissible for other chips and that is enough for us.