Jump to content
Corsair Community

I need help with CPU Temperatures and Fan speed with the Corsair H100i Pro


bazikid

Recommended Posts

Hey. So I recently cleaned my computer of all the dust on the dust filters using a can of air. I also used the can of air on the radiator of my corsair h100i to get rid of any dust that might be sitting on the radiator. As soon as I reassembled my computer, the idle temps for my cpu were fluctuating between 50 C and 80 C. My cooler fans were running at max speed trying to cool them off, but nothing was working. Eventually, I realized a screw was missing on one of my fans. I found the screw and put it back in with the washer, and the temps went back down to 28 C on idle. The next day, the idle temps go crazy again. I checked my thermal paste and there was barely any there, so I reapplied it and it worked. The next day, the idle temps are the same, however when I am playing games, the temps go crazy. The highest I've seen was 88 C. From then until now, the temps have somewhat calmed down, with idle temps being around 30 C and the temps while gaming are around 75 C. However, the fans on my cooler are running at max speed. I set them in Corsair Link to only run at 70% speed, but anytime the cooler reaches 40 C or the CPU reaches 70 C, the fans run at max speed. I have had my computer for almost 4 years now, and I have never encountered the fans running at max speed until I cleaned my computer. I've also never paid attention to the temperatures until now, so I assume my temperatures were normal until this point. Any idea what the problem is? I want to get this fixed as I feel like a temperature ranging from 70 to 80 C is not normal for gaming, as well as it is annoying to hear my fans running at 3000 rpm when playing games, especially since the cpu load is only around 40% and I have noise canceling headphones, which means it is very loud with the headphones off. I will really appreciate any help I can get.

 

Sorry if I wrote a lot, it's just what I have been experiencing over the past week. I don't know much about building computers since I bought a pre-built computer, but I haven't had any problems until I decided to clean it.

 

EDIT: I have a Corsair H100i V2, not a pro. Accidently put pro in the title. My bad!

Edited by bazikid
Wrong Title
Link to comment
Share on other sites

The fan blast is normal for pre-programmed curves when the coolant temperature (H100i Pro Temp) reaches 40C. While your coolant to CPU temp is largely determined by voltage and CPU type, as you've noticed that is about where CPU temps start to become an issue. What is likely different is the coolant temp is reaching 40C and that is its job to regulate.

 

Let things cool down naturally or test this after sleep or cold boot. Your idle coolant temperature should be about 4-6C above your room temp or more or less the same as the internal case temperature. You can compare this your motherboard temp sensor, non-C drive temps, etc. If you power on and it starts climbing up to a lot more, that likely means there is an issue. A normal CPU only stress test should go about +6C for 6700K. Use something mild like Intel XTU or even the CPU-Z "stress CPU" from with the Bench tab. You don't need to run it more than 10 minutes max. If you see the coolant go up +10-15 in short amount of time, stop immediately. Again, the coolant temp should drop 2-3C within 60 seconds of finishing and most of the rest within 3 minutes.

 

Also, can you confirm you really have a H100i Pro and not a H100i v2/GTX? For recognition purposes, the key is the block/pump head shape and the v2/GTX will have a silver plate instead of the RGB square.

Link to comment
Share on other sites

Okay, I will try that out. And yes I do have the H100i V2/GTX with the silver plate. I do not have the RGB Square. My bad, I just realized I put pro in the title. Edited by bazikid
Link to comment
Share on other sites

Yeah, everybody tacks “Pro” onto the end of everything these days, including Corsair. Unfortunately the behavior you’re describing and even the apparent trigger are a familiar story for the v2/GTX. We’ll see how the tests turn out, but you likely have a developing flow problem with the cooler.
Link to comment
Share on other sites

In response about the test results, I completed the test with CPU-Z. My starting temp on the coolant was 29.8 and after 10 minutes, it was 39.7. It gradually increased over the 10 minutes, and did not instantly shoot up to 39.7, as it reached the 7-8 minute mark, it was at 39 and by the 10 minute mark it was 39.7. After finishing the test within 60 seconds it came down from 39.7 to 36.2 and after 3 minutes it was down to 34 C. Are these good results? Throughout the test my cpu was 75 C with 100% load, however, when I play some games my cpu is that temperature with only 40-50% load. Anyway, let me know what you think.
Link to comment
Share on other sites

So the most likely scenario is the H100i v2 is starting to slip a little in performance, possibly due to a flow restriction inside the cooler. This seems to be the number one cause of premature termination for that model. It is not so bad yet that it is a clear failure, but performance is likely to continue to decline over time. You can operate pretty close to normal in your current condition.

 

However, the part I do not like is the sudden change between normal and high CPU idle temps. It suggests the coolant temp was unnaturally high as the result of some kind of flow stoppage. This could have been from bubbles in the wrong place after the cleaning, it could have been a momentary power twitch, it could be a physical obstruction. This is something to watch out for. When it starts to do the 40-50C coolant temp at idle, then it's fine 20 minutes later, then something is wrong and it's time to change it. Your gaming experiences are worse than the stress test because the GPU is heating up the case and everything else along with it. This is very common for most users and they will see gaming coolant temps higher than a straight CPU stress test. Case temp plays a strong hand in the total coolant temp since it is the baseline minimum.

 

1) Your test results seem normal, but to put the starting point in perspective you need to know the room temp. 29C starting coolant temp in a 23C room is just fine. If it was 17C, then I am more concerned and you should find a way to assess your in case temperature with the motherboard, drive, or any other sensor data.

 

2) H100i v2 - MAKE SURE your BIOS fan control for the cooler lead is set to disabled/100%/Full Speed depending on whether you are using the EZ BIOS or Advanced Monitoring Tab for fan control. This is where the power comes from. A clear indication this is not the case is the pump speed. It should either be 19xx rpm on the low setting or ~2900 rpm on the high. If its at 1500, you know you are choking power to the unit.

 

3) Get off the pre-programmed Quiet/Balanced/Performance curves. This were designed to operate in a 20-23C environment. Most people cannot maintain that case temperature when gaming. Make your own custom curve with fan speeds you can live with. +-200 rpm will have no meaningful effect on your temperatures. It often is meaningful for sound.

 

4) How does the pump sound? Bubbles will often have a random, static pop like tone. A restricted pump might start to accentuate the high pitch hum that is normally there. Normally, you would not see any performance difference for your CPU between the low/high pump speeds. If you switch to Performance and coolant temps suddenly drop 2-3C in a matter of seconds, the flow pathway is restricted.

 

5) Keep an eye on things. Getting hot under gaming/case temp increase is one thing, but high fan speeds at cold boot or on the desktop (not after gaming) is something else. As soon as this starts to happen, then reliability is gone and you need to make a change.

Edited by c-attack
Link to comment
Share on other sites

×
×
  • Create New...