Jump to content
Corsair Community

H115i Performance Mounted to a 6700K


Troy.Carnahan

Recommended Posts

Hello everyone!

 

I just recently built my new system and installed an H115i. Right out of the box I replaced the two (4 pin) fans with SP140 Blue LED Fans (3 Pin). So far the only overclocking I have done is enable XMP or let the ASUS utilities configure an overclock.

 

I have installed quite a few radiators in the past, but nothing on a skylake chip. The consistency in temps to me does not seem correct when running benchmarks or stress test. My temps have been seen going from 60C under load to over 80C. At idle, my temps fluctuate quite a bit as well. They'll go as low at 30C to 45C.

 

The last cooler I installed on an overclocked 4790K was an H110 GTX and the temps never went above 60C under load. Same case and fan setups, different processor and mobo. I do not have a GPU installed yet (its on back order).

 

I have installed quite a few radiators, I work in the computer service industry. But with that said, I am open to the possibility I installed my radiator incorrectly. I noticed the thermal paste on the H115i is much thinner and circular shaped compared to the H110 GTX which was square and slightly thicker. The other thing I noticed out of the box was a trickling sound in the radiator when I pulled it out of the box, it could be heard the more I moved the radiator around. I've never heard that on a radiator before and I was always under the impression that they were supposed to be airless to prevent thermal pocketing.

 

*EDIT* I understand my Corsair Link is not properly detecting the radiator fans, I have them plugged directly in to the MOBO. The ASUS utilities are reading the fan RPMS. As for why XTU is reporting incorrect info, I have no idea what's wrong there. It shows 4k watt power draw on the CPU?

 

What are everyone's thoughts, defective radiator or bad install? The stress test was performed with all fans maxed and the radiator pump on performance mode.

Temps.thumb.PNG.aa150789a3713cd81d057f7a5ea79893.PNG

1798094540_Temps2.thumb.PNG.a90959725875a4151dfc1744ab35a58e.PNG

Link to comment
Share on other sites

I suspect you are fine. If you had a contact issue, the XTU test would be untenable with a resulting thermal shutdown. Also, the water temp (H115i Temp) is rising, so heat is being transferred into the system. If you look at your VID from HWmonitor, you'll notice it peaked out at ~1.39v. That's getting right up where most overclockers stop on 6700k. This is also normal behavior for the chip on the auto voltage setting. It is made to be accommodating to a wide variety of scenarios and will keep pulling voltage to make itself stable.

 

To address this you can 1)Set a fixed voltage before running synthetic stress tests and then put it back to adaptive for normal use; 2) You most likely can lower your Vcore to something below the stock value. That will take some experimentation, but again the lower specified adaptive voltage shouldn't climb quite so high during normal use.

 

I don't see anything to suggest a cooler issue, but if you can give me some more information, we can be pretty sure.

 

I need your H115i temp right after cold boot.

Let it idle at the desktop for 5-10 minutes and give the H115i temp again.

(I already have the stress test water temp)

The approximate room temperature when you did this.

Link to comment
Share on other sites

I suspect you are fine. If you had a contact issue, the XTU test would be untenable with a resulting thermal shutdown. Also, the water temp (H115i Temp) is rising, so heat is being transferred into the system. If you look at your VID from HWmonitor, you'll notice it peaked out at ~1.39v. That's getting right up where most overclockers stop on 6700k. This is also normal behavior for the chip on the auto voltage setting. It is made to be accommodating to a wide variety of scenarios and will keep pulling voltage to make itself stable.

 

To address this you can 1)Set a fixed voltage before running synthetic stress tests and then put it back to adaptive for normal use; 2) You most likely can lower your Vcore to something below the stock value. That will take some experimentation, but again the lower specified adaptive voltage shouldn't climb quite so high during normal use.

 

I don't see anything to suggest a cooler issue, but if you can give me some more information, we can be pretty sure.

 

I need your H115i temp right after cold boot.

Let it idle at the desktop for 5-10 minutes and give the H115i temp again.

(I already have the stress test water temp)

The approximate room temperature when you did this.

 

The first pick is on a cold boot after having been powered down for 10 minutes. The second pic was taken after 15 minutes of idle time. The room temp is about 23C (73-74F).

 

For some reason when I set the fan speed to auto in Fan Xpert 3 while I had Link open, Link all of a sudden started detecting all the fans in my system, including the CPU fans. Fan 3 was actually idle at the time of this snipet, I guess Link is just displaying the last detected RPM.

642153490_ColdBoot.thumb.PNG.b0a3c329e21983c8071e73c9dff518c0.PNG

Idle.thumb.PNG.88cac3c0db01dc44b8b3d044b05d88e4.PNG

Link to comment
Share on other sites

Your fine. A water temp about 4C over ambient room temperature is about as low you can go without taking drastic power saving measures. There was no meaningful increase in water temp at idle over the 15 minutes. When there is a pump or flow problem, the water temperature will continuously climb. You are good to go. If you want to run a bunch of benchmarks, set your voltage to a fixed value. For normal use, this should not be an issue.
Link to comment
Share on other sites

Your fine. A water temp about 4C over ambient room temperature is about as low you can go without taking drastic power saving measures. There was no meaningful increase in water temp at idle over the 15 minutes. When there is a pump or flow problem, the water temperature will continuously climb. You are good to go. If you want to run a bunch of benchmarks, set your voltage to a fixed value. For normal use, this should not be an issue.

 

Awesome, that's good to hear. I appreciate your help!

Link to comment
Share on other sites

most mobos will overvoltage your cpu.

i did it like that:

adaptive mode with neggetive offset.

for normal boost 6700k at 4.2Ghz you shouldnt be needing over 1.275vcore

just adjust the offset value to neggetive 0.1 and you should get a total of 1.29vcore at benchs and around 1.2vcore at gaming.

hope it helped you out:)

Link to comment
Share on other sites

This is the first mobo I've used with the "adaptive mode". I sometimes saw my voltage exceed 1.4 which I wasn't comfortable with at all.

 

Yesterday I finally received my 1080 and decided to tinker with my overclock some more. Now that the embedded GPU would be disabled, I was hoping I could get my temps to drop and also increase my overclock. I manually set my voltage to 1.315 and increased the multiplier to 47. I managed to get a stable 4.7GHz overclock and with the manual voltage adjustment I no longer see temps above 66C in bench/stress test. Adaptive Voltage seems too overzealous.

 

*Edit* I have EIST disabled, so my CPU stays at 4.7 and the idle temps are around 22-24C.

Link to comment
Share on other sites

Archived

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

×
×
  • Create New...