TEC/Peltier based chilled water cooling system
TEC/Peltier based chilled water cooling system
This idea is intriguing but doesn't fully capture the potential cooling benefits of the Peltiers. The main concerns are lower voltage limitations and integrating it with a radiator, which may not support the Peltier's performance. Despite these challenges, your experience seems to point to effective solutions. What exactly are you trying to cool?
Hey there. I really appreciate your post. I’m considering setting up something similar with my rig, but I’m having trouble with water cooling—it seems to be getting too hot, around 50-60°C. I’d prefer near ambient temperatures, aiming for a small temperature difference.
I wanted to ask a few questions about your Peltier system. You mentioned you didn’t choose the water cooling option quickly because it would cool the hot side too fast. Would a bigger radiator help more? I’m not sure if it really improves things.
Also, from what I understand of your test results, are you looking for a high-flow water block with very little turbulence inside? I was thinking that using solid copper to soak up some heat from the cold side might be better than some acrylic-top water blocks. Maybe something like a 1/4" thick copper base with a 1/4" or 3/8" copper top, soldered or brazed together, and a large hollow cavity in the middle? Or perhaps a design similar to the early water blocks you mentioned? Something without a pin matrix, would be great!
Thanks for sharing your thoughts—I really value your insights and the inspiration you bring.
Hey there. First of all, I really appreciate your post. I’m considering setting something similar with my rig, since the more I research, the less I like water cooling. If I invest 500 dollars in a water cooling system, I don’t want to end up with temperatures around 50-60°C—I aim for close to ambient, with a temperature difference almost equal to the environment.
It’s frustrating when you spend a significant amount on water cooling and it doesn’t meet expectations because the extra heat from overclocking wasn’t fully accounted for. The biggest error with water coolers using radiators is that they often underestimate the cooling needs; they usually require 2 to 3 times more radiator area than necessary before overclocking even starts.
Then comes the question: What’s the issue? It leads to a nagging feeling in your stomach when you realize all your money was spent. I’ve been there before.
So I thought maybe there’s a better approach—something with full control, not relying on radiators that depend on ambient air. Peltier or Tec cooling can get you well below ambient, even down to zero if you want. But that’s not what this thread is about.
kermdawg :
I wanted to ask a few things about your Peltier configuration. You seemed to skip the water cooling setup for the hot side quickly. Do you think a bigger radiator would help? I know it might seem counterproductive, but it could make a difference.
Back in the day with Peltiers, the cold side was directly on the CPU die, which required special motherboard insulation. The CPUs were small, the Peltiers were compact, and they used low voltage. Those hot sides were water cooled, but the ones I’m using now get too hot for water cooling—the pump isn’t built for that temperature. An all-metal heat pipe or air cooler could handle it better.
A bigger radiator won’t help much when trying to cool around 180°F. You’d need to touch it and risk burning your hand. My Peltiers max out at 125°C—well below the boiling point of water, which is 100°C.
The Peltiers today are much more powerful than they were before, handling a lot more power.
Secondly, from what I understand about your testing, you’re looking for a high-flow water block with minimal turbulence. I was thinking maybe a solid copper base with a thinner copper top—like 1/4" or 3/8", soldered or brazed together, and a hollow cavity in the middle. Or perhaps something like the early water blocks with a wide maze design? No pin matrix needed.
Thanks for sharing your thoughts! You’re inspiring others to explore new ideas. The cold pickup block is crucial, and both of my blocks have been modified for better flow. The copper base doesn’t need to be thick—just enough to absorb heat quickly.
If the cold stays in the block too long, it can freeze, restricting flow and causing the Peltier to keep working until the ice expands and opens the block. That’s not ideal.
Some ideas about the Peltier seem logical, but they have their limits. This thread is sharing what actually works and can be replicated. I’m currently running this setup, with only one Peltier active. My water temp is 14.8°C, room temp is 22.2°C—about 7.4°C below ambient. CPU is idle at 18°C. If I switch to radiator cooling, it might help. My graphics cards are already radiator cooled; I think that’s the best option for full GPU coverage.
Thanks for your feedback!
That's the reason I haven't gone out and bought my setup yet. Everyone I talk to about it on forums and whatever says "Oh, a 120x1 or a 120x2 will be plenty if your just doing your cpu!" I want my cpu to stay running at 70C all the time. I need it to be running at 40-50C while I'm stress testing. Absolute maximum temps. That's why I was looking at a giant phobya 160x9 rad (or whatever the car radiator looks like). Not only would I have enough for a cpu, but I could also fit a couple of GPUs too, and never have to worry. Not be one of these clowns with 3 tiny radiators trying to piece it together because they didn't have the foresight to plan ahead.
But, why bother with all that BS if you can just get one of these neat little peltiers?
This might seem like a dumbass question, but did you think about running some antifreeze in the water cooling loop? I'd think it could give you a bit more leeway in your temperatures since then you wouldn't have to worry about ice forming.
If worse came to worse, you might even run some heat trace on the water block itself to prevent it from freezing...as a backup in case something went wrong or you forgot to turn the fans on. The heat trace would keep it just warm enough to stop ice from forming. It could even help you fine-tune the water loop temperature so you could get as close to the dew point as possible without needing extra insulation.
I know, it kinda defeats the purpose, but...I dunno.
Also, I don't know about some other folks, but maybe you could share some of those lessons you learned the hard way if you don't mind. I know it's a lot to type up, but it would help save someone like me from making some of the mistakes you already might have made.
Just got one of these to cool my GPUs with, a Watercool MO-RA3, I believe the MO means monster, it’s massive! No, it wasn’t thought about before, because in my past water cooling experiences I used anti-freeze and a leak was a huge problem—learned the hard way, anti-freeze can damage wood finishes or floors. It’s possible but not essential; you can manage temps by adjusting fan speeds on the heat sinks to cool the hot side.