Investigating Cooling Below Ambient Water Temperatures
Investigating Cooling Below Ambient Water Temperatures
I would personally employ heat exchangers with a closed loop, placing the cooler as a water/slush box with the rads inside. The WC loop would stay closed, but maintain either a coolant or growth-controlled water in the cooler, using the ice jugs similarly. The internal growth in the cooler must come from condensation collecting and then harboring crawlies; otherwise, I doubt anything would develop in the coolant created with those microbial inhibitor tablets. You've mentioned this approach many times before, even though I first shared this cooling solution over 11 months ago. Have you ever tried your own idea?
I recall once, a few years back, using a bucket, ice, and a heater core to test how it would function on my P4 system when it was moving toward its afterlife. I'm just sharing some thoughts—someone who hasn't read through your configuration might find this useful for a similar project.
Adjusted the water temperature, which led to increased growth. I wasn’t sure if giving up was the right choice. Maybe a few more radiators before it reached the water? Or adjusting the jugs more often? Or simply using a growth inhibitor.
I kept pushing forward; adding extra radiators didn’t fix the issue. What I required was a dual-loop setup, since GPUs don’t actually need ice cooling—they’re fine with a closed-loop radiator system.
The mildew was growing because condensation built up under the lid, and the warm, humid air in the large space above the water became suitable for growth once the temperature hit around 78°F (26°C).
I hadn’t been using ice when the radiators were fully cooling the system, though I attempted that method. Unfortunately, the radiators reversed the cooling process by heating the chilled water faster.
Even when the water temperature matched the standard loop setting for my rig, my GPU load remained 40°C below the stock air cooling solution. That’s the same cooling I was expecting with my Rad Box setup.
Taking them out of the CPU loop would have brought the system back to ice-water cooling, allowing me to achieve my original aim of high multiplier overclocking flexibility with my 2500K.
I was curious about whether this situation was accurate. The surrounding air would probably play a bigger role because of the combined CFM of air and ambient heat in watts being exposed to the rads, compared to what the cards would be sending into the loop and requiring cooling.
Taking into account it was performed years back and was mostly for observation, it wouldn't be suitable for current hardware requirements without adjustments in heat exchanger surface area/volume and revised heat loads. The primary reason I mentioned it was the project's similarities.
I noticed condensation forming on the CPU block, which meant I had to keep strong airflow around it and use some paper towels to absorb the moisture. I don't remember having a significant amount of it, not enough to cause drops falling, but it was definitely noticeable. I don't think I had any temperature monitoring tools available at the time, to be honest—maybe ASUS Probe was the only one, but it wasn't very reliable. I’m not sure if there were many good temperature monitoring options back then.
The main reason I no longer use this method for cooling now is mainly due to practicality and mobility. I occasionally visit LANs (though much less recently) or even play games at friends' houses instead of being fully remote. I also prefer not to keep ice on hand just to cool the water in the slushbox. I used a 5-gallon bucket and a bag of ice from the grocery store for my experiments back then. I realize a different approach would be needed for a long-term setup, but I don’t want to constantly manage ice for this purpose. Sure, there are other alternatives for continuous operation besides this one, but I’d rather have 24/7 cooling without the inconvenience of needing ice when needed.
I’ve never doubted that your cooling solution wasn’t effective; my intention was simply to expand the possibilities for other forum members who might read about your build. The original discussion seems to have been deleted or lost somewhere in the forum, so while I’ve mentioned the heat exchangers several times, not everyone has seen the old thread to notice them.
The DIY aspect and the results clearly demonstrate how well your system is functioning. That’s undeniable.