Custom loop worse than AIO
Custom loop worse than AIO
For those familiar with water cooling setups:
I own an EKWB custom loop kit, the EK-KIT X240, and an Arctic Freezer 240 All-In-One liquid cooler. I tested them alongside a matching CPU (i7 7700K, no overclocking) using identical thermal paste. My focus was just on CPU cooling.
Using Prime 95 with a small FFT setting, the EK-kit hit 81°C while the Arctic Freezer AIO held steady at 72°C.
I’m quite confused by these numbers. The main difference between the two is the fan count: the Arctic Freezer has four fans compared to the EK kit’s two. However, I’ve heard that fan count doesn’t really matter much. On the other hand, the EK kit features a larger radiator (60 mm versus 38 mm on the Arctic Freezer) and a more powerful pump, which should give it an edge.
I’m curious to understand why the temperature gap is so big or if something is wrong with my EK kit before I invest further in custom water cooling.
Thank you for your feedback. I’m only testing on an open bench, yet airflow might still be the key reason the EK kit doesn’t perform well, even if the pump and CPU are working properly. I’m using the two Vardar F3-120 fans included with the kit. Perhaps they aren’t powerful enough to cool the radiator effectively. What improvements would you suggest? Should I switch to stronger fans? Add another radiator? I’d like to overclock to 5Ghz someday, but the current temperature is too high for that.
On the other hand, the Arctic Freezer 240 is a great AIO and probably the best in its category. I might have to admit this AIO is significantly more optimized than the EK kit when it’s set up normally.
Push/pull generally outperforms simple push or pull methods, so the advice you received was likely inaccurate. Radiator thickness plays a role, but it also relies on fan selection. Thicker radiators may require higher static pressure fans or a push/pull setup. Additionally, the quality of thermal paste and proper CPU installation are important factors. Certain thermal compounds might need to undergo curing or cooling cycles before they are fully effective.
the thicker radiator has its pros and cons. With sufficient air power, a thicker one can help lower the operating temperature. However, you still require enough airflow. if your case lacks good inbound and outbound air connections (like the NZXT H440 where airflow is restricted), then your fans might not perform well. with a larger radiator, you'll need more air across the fins compared to a smaller one. the EK model is excellent on its own, so I think you've made the right equipment decision. it seems like you're missing some extra airflow across the rad for optimal performance compared to your AIO setup.
Thank you for your feedback. I’m only testing on an open bench, yet airflow might still be the key reason the EK kit doesn’t perform well, even if the pump and CPU are working properly. I’m using the two Vardar F3-120 fans included with the kit. Perhaps they aren’t powerful enough to cool the radiator effectively. What improvements would you suggest? Should I switch to stronger fans? Add another radiator? I’d like to overclock to 5Ghz someday, but the current temperature is too high for that.
On the other hand, the Arctic Freezer 240 is a great AIO and probably the best in its category. I might have to admit this AIO is significantly more optimized than the EK kit when it’s set up normally.
I appreciate the feedback. I’m only testing on an open bench, but airflow might still be the key reason the EK kit underperforms, even if the pump and CPU block are working fine. I’m using the two Vardar F3-120 fans included with the kit. Perhaps they aren’t powerful enough to cool the radiator properly. What improvements would you suggest? Upgrading to stronger fans or adding another radiator could help. I’m considering overclocking to 5Ghz someday, but the current temperature is too high for that.
On the other hand, the Arctic Freezer 240 is a great AIO and likely the best in its category. It’s possible this model is far more optimized than the EK kit out of the box. You might be right about the efficiency gap between the two.
I’ve used the Corsair H100i before when I first set up my custom loop, but I wasn’t particularly impressed, especially after experimenting with overclocking. Push/Pull could improve airflow across the radiator, but I’m not sure if that extra effort is worth it for an AIO. For a custom loop, I’d recommend using push/pull.
Additionally, keep in mind the pump speed shown in AIO builds—sometimes it runs a bit slow. In the fan header, I made sure it was set to maximum, which really helped. This seems like one of the main issues with AIOs when troubleshooting. For custom loops, we should check pump speed, radiator flow, and tubing connections. I’m not sure if these steps will be useful, but they might help clarify other problems.
The EK kit fans are now in the push position. I once detached the pump's PWM connector from the chassis fan plug on the motherboard, allowing the pump to run solely on the PSU and maintaining full speed without any temperature change. I'm considering checking the flow rate to verify the pump is functioning correctly. Do you have any suggestions on how to do that?