How to build your own portable thermoelectric device with a quadcopter motor.
This project might not have an electrifying impact to the world but it's a baby brother to my first instructable (Thermoelectric Fan) and just had to be built. It's much smaller than most similar projects and the top power device can be detached from the lower heat source base and be used as experiment in other areas. What does it do? It converts heat energy into kinetic energy, looks nice and sounds awesome!
It uses a thermoelectric generator module (TEG) to power a motor with a fan. The heat source (candle) heat up the lower aluminium plate => creates a temperature difference over the module => generates electricity to the motor => increases air flow through heat sink => increases temperature difference => more power. Without the fan it would eventually stop generating power since the heat sink would get almost same temperature as the plate.
Almost any heat source with enough temperature can be used to power the motor and this makes this small device easy and fun to experiment with. It's very sensitive to temperature differences. I could run it on water with only 20C difference (23C air temp and 43C water temp). As seen in the video it's powered from fire, hot water, food, and even used to cool a computer processor. The latter is an idea I had when i created my first build in 2013 and people said to me it cannot be done. It can and it's stable for normal use but with further optimizations it could probably be even better. I don't think it can replace a regular CPU cooler but as it uses no electricity from the computer it's somewhat environmental friendly. More information on last page.
It require few parts, easy to build and you do not need any knowledge about electronics. You just need some mechanical tools for cutting, drilling and polishing the aluminium. If you want to build it, please read further.
If you like my projects and want to contribute for future instructables, feel free to use this bitcoin address (or QR code image): 1BouwowuprgQrtUYgyzYnNvHyRYbLceqHg
Step 1: Materials and Tools
I think you can find everything on Ebay. Prices are estimated. Aluminium and bolts not included in price since it's basically scrap metal.
Step 2: Construction
Wire (Used for fixating the heat sink)
If you use a 8mm motor you will need a 8mm centered hole in the heat sink.
Bottom Aluminium Sheet
Top Aluminium Sheet
Step 3: 3D Printing (Optional)
For design purpose you can print this special part if you have a printer and transparent filament like T-glase.
I used a wanhao duplicator 4x printer and simplify 3D software. If you want it lower or higher you can simply scale it along Z.
Link to model:
Step 4: Assembly
Step 5: Self-regulating CPU Cooler
The module could be used in a computer to replace a conventional CPU cooler. It uses only heat from the CPU to power the fan, which in turn transfer the hot air away from the CPU. It might not have enough efficiency for a modern CPU but this is more or less a proof of concept. It does not require a sensor, electronics or motor controller as the fan speed is self-regulated by CPU temp.
Currently tested on an AMD Athlon 64 3800+, 2400MHz single core. Normal computer use was stable but I also did a 5 min stress analysis both with and without the motor fan compared to stock cooler. Clearly the fan helps to cool the CPU. The stock cooler is a 80x70x70mm heatsink (402g) plus a large fan compared to my small device (85g with motor & fan):
After the test the CPU went back to stable 37C and fan RPM went to normal. Also be aware that I did not apply any down force on the device, it just rested on top of CPU and if you're a computer builder you know that is not very efficient. The device is also way smaller than stock cooler and a larger TEG and heat sink would probably do a better job. Would be great to see someone else experiment with this! Though, I will not be held account for any CPU damage. This particular CPU has a rated max temp of 69C so I would not recommend the build as it is.
Update 1 (Second picture):
Also tested with added small heat sinks on the bottom plate for better cooling from left-over air. I did this to show that parameters can easily be optimized further. Just to be sure, I removed the extra aluminium during the test and temp increased to 75C as before.