Jump to content
Corsair Community

H115i Noise


Eaton

Recommended Posts

I recently did an advanced RMA for my H115i because of disappointing temperatures. I received my new one from Corsair a few days ago and immediately noticed it was much noisier than the old one. I figured it could be air bubbles so I have waited a few days, but nothing has changed yet, so I was curious if I should wait a few more, or if this noise is something else. It's installed the exact same way as the old one.

 

The Quiet pump mode is OK, but still a bit noisier than I would like - the old one was dead silent.

The Performance pump mode is unbearable. The old pump could not be heard over the fans, but this new one can.

 

Here's some videos. Please note that all fans in the system are turned off to capture pump noise only.

Quiet: [ame]https://youtu.be/FugkShwawbg[/ame]

Performance: [ame]https://youtu.be/_SAvB6VHaLE[/ame]

 

The temperatures are a little better, but I still don't think they were what they were on day 1 of my build. AIDA64's FPU test immediately brings all cores to 75C+, is that normal for a 6700K and an H115i? Idle temps are great at the low 20s, but stressing causes some pretty high temperatures. Also, the water temp is usually always around 30C.

 

Let me know if I should ask for another RMA now, or wait a few more days. Thanks in advance!

Link to comment
Share on other sites

It's hard to accurately capture sound and even more so to differentiate loudness on a recording. It's a little concerning if you can hear the pump over the fans. However, this has been a consistent complaint regarding this cooler since its original launch. If you can't hear the pump in quiet mode, then you at least have a workable arrangement. You don't need performance mode unless you run renders/encoding all day long. Flow rate is almost never the limiting factor in AIO coolers and my older H110 is another 500 rpm below your quiet mode. Water temperature has not been an issue on either of the overclocked HW-E CPUs I've used on this system. I would register your observations on the RMA ticket. You are not obligated to action by doing so and this creates a record of it. I do not think it will improve on its own.

 

Yes, an FPU only test on AIDA is going to bring high temps right from the start. On my 5930K 4.5GHz@1.275 and 4.0/1.17v cache, it will be in the 75-80 range immediately. This is not a good test of your cooling performance because of the extreme stress on the CPU side of the equation. The cooler carries heat from the CPU to the water and then out of the case. It can't prevent what happens to the silicon when you apply extreme duress. Unless you have a specific need to scrutinize the FPU, run all CPU/FPU/Cache/memory tests together. This is effectively a "blend" test, more useful for general stability and for testing the cooler's ability to manage heat. You will have considerably lower temps on the blend than FPU only. Also, we need your frequency and voltage to make sense out of the results.

 

Idle water temperature is largely based on room temperature and ultimately ambient case temperature. 30C is pretty standard for anyone is a Summer climate or with room temps above 25C. Water temp should be relatively slow to change, both up and down. That is the variable you can affect with your fan speed. The fans won't do anything for the instant on spike when you start a stress test. That is all down to the amount of voltage and the rate of heat transfer through the CPU, TIM, and cold plate.

Link to comment
Share on other sites

It's hard to accurately capture sound and even more so to differentiate loudness on a recording. It's a little concerning if you can hear the pump over the fans. However, this has been a consistent complaint regarding this cooler since its original launch. If you can't hear the pump in quiet mode, then you at least have a workable arrangement. You don't need performance mode unless you run renders/encoding all day long. Flow rate is almost never the limiting factor in AIO coolers and my older H110 is another 500 rpm below your quiet mode. Water temperature has not been an issue on either of the overclocked HW-E CPUs I've used on this system. I would register your observations on the RMA ticket. You are not obligated to action by doing so and this creates a record of it. I do not think it will improve on its own.

 

Yes, an FPU only test on AIDA is going to bring high temps right from the start. On my 5930K 4.5GHz@1.275 and 4.0/1.17v cache, it will be in the 75-80 range immediately. This is not a good test of your cooling performance because of the extreme stress on the CPU side of the equation. The cooler carries heat from the CPU to the water and then out of the case. It can't prevent what happens to the silicon when you apply extreme duress. Unless you have a specific need to scrutinize the FPU, run all CPU/FPU/Cache/memory tests together. This is effectively a "blend" test, more useful for general stability and for testing the cooler's ability to manage heat. You will have considerably lower temps on the blend than FPU only. Also, we need your frequency and voltage to make sense out of the results.

 

Idle water temperature is largely based on room temperature and ultimately ambient case temperature. 30C is pretty standard for anyone is a Summer climate or with room temps above 25C. Water temp should be relatively slow to change, both up and down. That is the variable you can affect with your fan speed. The fans won't do anything for the instant on spike when you start a stress test. That is all down to the amount of voltage and the rate of heat transfer through the CPU, TIM, and cold plate.

 

Thank you for taking the time to write a detailed answer!

I am glad to see the FPU test is expected to show high temps. The only remaining issue then is the noise. What is making the noise exactly? You can hear it pretty clearly in the Performance video - is it air bubbles or mechanical?

Link to comment
Share on other sites

It's mechanical. The faster you go, the louder it will be and the more precise the build and parts need to be to keep it smooth. There has been a lot of variability between different units at that higher speed. It's up to you if that is acceptable or not. Since you just went through a replacement and you have not truncated the original warranty, you may want to let it ride for a little while to see how things go. You can RMA it again if the situation becomes intolerable or otherwise irritiating.
Link to comment
Share on other sites

It's mechanical. The faster you go, the louder it will be and the more precise the build and parts need to be to keep it smooth. There has been a lot of variability between different units at that higher speed. It's up to you if that is acceptable or not. Since you just went through a replacement and you have not truncated the original warranty, you may want to let it ride for a little while to see how things go. You can RMA it again if the situation becomes intolerable or otherwise irritiating.

 

Thanks - that is what I will do.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...