TEC build log start
TEC build log start
to insulate the reservoir, i'll begin by covering it with foam insulation, and possibly adding clay on top if freezing occurs nearby.
they offer excellent fittings today for converting from G1/4 to soft copper tubing. the tubing will also be finished with clay.
http://www.performance-pcs.com/new-koola...m-1-2.html
power consumption :lol:
there are many jokes here, it's pretty high. I'm planning to use five power supplies for this project—one 1250W PS for the motherboard and graphics cards, another 600W PS for the four water pumps, plus all the fans, controllers, temperature sensors, and flow meters. There are three 12V 45A PowerMax PS units for the TECs.
My PC is really powerful!
Why are you concerned about the box mobility? You won’t be sending this PC to LANs unless you have a flatbed and a forklift truck. XD
I was thinking about that clay on the tubing. If the tubing shifts even slightly, there will be a gap, right?
Will you insulate the motherboard? I noticed the vacuum pump and thought it was interesting, but what if something leaks or moisture enters, causing an issue?
This is really great, keep going with the updates. It’s something I haven’t seen anyone tackle before and I’m excited to see the results.
Good luck!
Thanks but essentially I’m creating a basic industrial vacuum chamber for electronics testing. The main change is that these chambers can become warm, and they also come with a higher cost.
If I were using only clay to seal the chamber, a leak would be quite likely. Instead, I’m using high-quality silicone sealant and applying flex seal spray on the rubber parts.
When I determine the required thickness of clay to avoid condensation or freezing on the cold side blocks and tubing, I’ll wrap everything in plastic wrap and coat it with flex seal to keep everything intact.
Next, I plan to construct a submarine using an old truck, relying solely on flex seal.
I encountered a few challenges while setting up this system. Cold coolant is beneficial for hot parts, but rubber seals become stiff in the cold, which can lead to small cracks and leaks. To fix this, I took several steps: swapped all the rubber O-rings for ones rated for -65C, applied gas line Teflon tape to every connection, used pipe thread sealant on all joints, and after securing the connections, cleaned them with a Q-tip and alcohol. Then I applied a gasket sealant suitable for -65C at the seam of each connection.
It might be excessive, but I wanted to prevent any leaks.
Freezing of the blocks isn’t a major concern for the coolant inside, but when the system is turned off, the ice melts into water and leaves a puddle under the blocks. During testing, I placed a pie pan beneath the blocks to catch the water, but for the rig I’m trying to avoid ice formation altogether.
What kind of fluid temperatures are you observing? I understand you might not have connected any components yet, but have you hit the thermal threshold of your cooling system yet? (Sorry for my wording, by this I mean: what is the lowest temperature your loop has reached, beyond which it can't drop further? Again, sorry, but in my defense, all my classes are in Italian)
Anyway, since you're still in the testing stage and you're comfortable with power demands, why not install a dehumidifier in your workspace? That way you'll avoid excessive ice accumulation during testing.
This is great, I'm looking forward to seeing the project finished.
I used a lot of fittings for this project, but now my spare parts box is nearly empty. The problem is I still need around $500 more in fittings to finish, which is really discouraging given the cost. Once completed, I’ll prepare a detailed cost estimate, but right now it seems quite high. By the time I finish this rig, it will be about three years since I began saving for these parts.
the TEC's I am using have a 70C delta T meaning it can reduce temperatures by 70 degrees from the ambient, and it approaches that level when cooling just the water, without any components running on the loop; the water reached -35C, which is extremely cold.
with the CPU active, adding heat to the water loop, one TEC kept the water at 0C under load, while the CPU was at 38C under load, especially at a 5Ghz clock. this is the extent of my testing when I proceeded to construct the full cooling system for the entire setup.
And you're adding another 2 TECs to those figures, right? That changes things significantly. It means you could actually reach around -25°C, depending on your surroundings (assuming you're sitting about 25°C). What kind of coolant are you using? I think anti-freeze is what you're considering, but I'm not sure which specific brand. How is the clay coming along? I imagine it will look really cool once you finish (did you get the rainbow version as linked, or just one color?). This build is impressive—thank you for sharing. Are you planning any cosmetic touches, like a custom paint job?
Update: Just reviewed your plan again. Why not run your peltiers at full power? You have plenty of amps available. I guess (though I'm still learning electronics and you're building satellites), you could run them at maximum wattage and still have a good safety buffer. I recall you mentioned adding a 25% safety margin, but you're actually planning for a 55% margin. Is this meant to align with the most efficient point of your power supplies (I know they peak at 80% load, but I'm not sure how much it changes with load)? Did I miscalculate something?
yes, you're correct about the TECs being too powerful. I prefer not to run my PS at full capacity, so it might be a bit excessive. The TECs operate at 16V and 26A at their maximum, while at 12V they perform just over three-quarters of their max at 19A. I also hope they won't burn out quickly under this voltage setting.