Water Cooled: GPUs and CPU gradually warming up
Water Cooled: GPUs and CPU gradually warming up
Here’s a revised version of your message:
Thank you for all the advice.
Considering everything, how should I go about troubleshooting this?
If the problem lies with the radiators, what would be the most effective solution? I think replacing the top radiator might be an option, but then what would be the best alternative? The motherboard could help by limiting the thickness of the radiator at the top. Maybe a longer radiator would work better? If that’s the case, do you have any recommendations? Ideally, I wouldn’t need to install anything outside my case.
On the other hand, if it’s a flow restriction issue, I’d probably need a flow sensor, although they aren’t inexpensive.
As mentioned before, I’ve already re-seated the CPU, but I’ll try again.
I’m feeling really frustrated and overwhelmed with all this. It seems like part of the process is taking a toll on me. Thanks again for your help.
Honestly, I think a larger, more WC-friendly case and an extra radiator would be better. If changing the pump setting from 2 to 5 doesn’t help much, and there’s a lot of water turbulence in your reservoir, restriction seems unlikely to be the issue. This might be the most costly choice to make, but you’re trying to manage a significant amount of heat from many components, even without overclocking. I’m using San Ace 38mm thick 120mm fans (model 9G1212H1011) or something similar on my 480mm radiator setup. I had two 140mm NoiseBlocker PK-3 intake fans and one 140mm exhaust fan (same model), plus the radiator fans. My configuration probably has a lower TDP to keep things cool compared to yours even if you OC your CPU. Many experienced users here usually have setups that are either CPU-only or CPU + GPU. For your build, though, you’re handling two high-end GPUs with your CPU. That puts a huge cooling demand on the system. That’s why I believe more cooling capacity will be needed to lower temperatures while keeping fans quieter and running at lower speeds.
For now, I’m considering letting it rest. Another radiator seems like a better fix than buying a flow sensor right away.
I’ve adjusted the FPS and temperature limits via Afterburner, which has helped maintain water temps between 38-40°C.
Maybe later I’ll upgrade to an external radiator and check if that resolves the problem. Any recommendations on suitable models? The link I checked looks promising for a model with a microprocessor-controlled pump.
If I connect it to my second radiator, would that work? And would a stronger pump be necessary then?
For now I believe it's best to let it rest – I think another radiator would likely resolve the problem, and I'm hesitant to spend $40 on a flow sensor just yet.
Through Afterburner I've established an FPS cap and a temperature threshold of 60C. This has at least maintained a water temperature between 38-40C.
Maybe in a few months I'll consider purchasing an external radiator and assess whether that would help. Any recommendations on external models? I was considering this one:
http://www.au.aquatuning.com/water-cooli...12-v-pumpe
Would that be sufficient if I connected it between my second radiator and the rest of the system? Would a more powerful pump be necessary then?
You might not have extra funds at the moment, but you do have the time to follow what I suggested.
I'm not trying to offend you, but from what you showed, your current configuration seems quite limited. Your front intake is blocked by the HDD cage, and the lower radiator is drawing air into the case, which heats up as the system warms – this is reducing fresh air flow.
Your rear exhaust isn't generating enough air to effectively pull cool air from the case. Changing its orientation could improve both airflow and cooling for your motherboard voltage regulators. Without proper cooling, they'll overheat and likely fail, shortening their lifespan.
Right now, your setup is focused on heat dissipation rather than air circulation. Please review your arrangement carefully and make adjustments.
Good luck! Ryan
So it's been a while since I posed this question, but things finally came to a close. The truth is, I was a bit of a dummy, but I’ll tell you a story and TL;DR it at the end.
I heeded the suggestions from that discussion and made a variety of adjustments—added more fans, explored different fan setups, moved the PC off the ground (those tweaks had only minor impact), cleared some blockages, removed radiators, changed radiator positions, switched coolant, adjusted pump speed, etc. I managed to achieve a decent balance, though it wasn’t quite what I expected, but it worked in the short term.
In the end, I replaced my old bulky radiator with an Aquaduct 360XT Mark V external unit. I removed the old one and stored it away.
After setting everything up, temperatures improved significantly. I reached stable readings like:
CPU 68°C
GPU 1: 52°C
GPU 2: 53°C
Temperatures were consistent, and the passive cooling from the case performed well. The fan controller was also a big plus.
Still, the coolant stayed warmer than I preferred, but that’s just life.
During a test of Dying Light (my favorite game), my temps stayed nice and steady, around 144fps.
That night, while shutting down my monitor, I accidentally pressed the Hz button on the screen. A clear “60Hz” appeared in the middle, and everything changed.
In my original post, I hadn’t mentioned that I use a ROG ASUS Swift G-Sync monitor with adjustable refresh rates (60, 120, 144Hz). Before building my new setup, I’d just purchased it and wasn’t heavily using graphics.
After assembling the rig, I wanted to push graphics to their limit. I didn’t realize the impact until later.
So, if I switch to 60Hz or even 120Hz, my temps drop dramatically. The extra tweaks I made were a valuable learning experience, and thanks to everyone for your help. Now I think I can run at 144Hz without overheating much.
The biggest takeaway is that I finally have a solution to my issue—and I’m not crazy, just a bit confused.
That’s the summary. Hope this helps someone else later!