Are these test results a sign i need more cooling?

[Divinenite]

Hello all,

This is my first post here 😀. I recently built a mid grade gaming comp 1 month ago and finally got around to stress testing my rig.

I used Prime95 (blend test) and ran it for 6 hours (at night, temp in room was roughly 60 degrees F). I also moved the Case to an open area. The results concern me though.

After 6 hours, my max temp recorded (though HWmonitor) was 98 C and averaged 80 C. For the first hour, max temp was 78 and idled around 68 C.

My Rig is as follows:
Asus Z87-Plus mobo
I5-4670K intell cpu
16 GB Corsair Vengeance (1600 MHZ)
Asus Geforce GTX 660 TI
HAF X 942 (case)
Fans for case currently (currently all stock for now):
1x200 mm side fan (~650rpm) intake
1x200 mm top fan (~650 rpm) exhaust
1x230 mm front intake fan (~640 rpm)
2 x 120 mm exhaust (radiator fans)(~2300 rpm)
H80I Corsair hydro series (radiater+ fans are mounted in the back)

Now i do have another 200 mm after market CM fan and 2x corsair 120 mm after market fans (to replace stock radiator fans)

I do plan to eventually replace all stock fans with after market.


Is there anything i could/should change here? I do want to overclock in the future, but if a stress test is pushing 100 C, then i feel i need to lower that temp first.

Any help is very appreciated!

 

[Buzz247]

Unless you have an octopus mess for wire management, case cooling is not the problem. HAF 942 has great air movement design.

Much more likely you have an issue with your heatsink block making proper contact with the chip.

How much, what type, how applied of thermal grease?

Are you sure pins are completely seated on the cooler? Should be tight, no wiggle

 

[Divinenite]

Unless you have an octopus mess for wire management, case cooling is not the problem. HAF 942 has great air movement design.

Much more likely you have an issue with your heatsink block making proper contact with the chip.

How much, what type, how applied of thermal grease?

Are you sure pins are completely seated on the cooler? Should be tight, no wiggle

That could be my issue, the h80i had thermal compound pre applied. I have some artic silver 5 but didn't use it. The connection was indeed snug, no wiggle room. Do you feel the issue lies with the h80i then?

 

[Buzz247]

I'd try unmount, clean with rubbing alcohol n qtip, apply bb sized artic silver to middle, apply sink flat - dont lay at an angle- test again.

Possible h80 is just too light duty for the wattage of that processor.

This is at stock speeds right? Not overclock?

 

[Divinenite]

I'd try unmount, clean with rubbing alcohol n qtip, apply bb sized artic silver to middle, apply sink flat - dont lay at an angle- test again.

Possible h80 is just too light duty for the wattage of that processor.

This is at stock speeds right? Not overclock?

Completely stock, which i would be amazed if it would be too much for the h80 o.o... but it could be possiable, i will def remount and try again, anything else you would recommend? I just have a hard time believing the h80 would struggle this much on stock settings.....

 

[Buzz247]

H80 and 100 perform at the same level as high end air - well, except for really high end like noctua or phantek.

But ya, at stock you should still be seeing 60s. Silky question - what speed are your fans at? Low med or high?

 

[Divinenite]

H80 and 100 perform at the same level as high end air - well, except for really high end like noctua or phantek.

But ya, at stock you should still be seeing 60s. Silky question - what speed are your fans at? Low med or high?

I had them running on high around 650 rpm for each.

 

[Buzz247]

Well grats! Found your problem! 650rpm isnt even the low end of quiet profile. 900 is. At high performance profile (all 3 white led) 1900rpm. Turn up your fans! And just in case - should be plugged into the pump block, not your mobo. Make sure plugged in right spot too.

Read this http://www.corsair.com/blog/understanding-the-hydro-series-h80-and-h100-cooling-performance-profiles/

 

[Divinenite]

Well grats! Found your problem! 650rpm isnt even the low end of quiet profile. 900 is. At high performance profile (all 3 white led) 1900rpm. Turn up your fans! And just in case - should be plugged into the pump block, not your mobo. Make sure plugged in right spot too.

Read this http://www.corsair.com/blog/understanding-the-hydro-series-h80-and-h100-cooling-performance-profiles/

I believe i wasn't quite specific enough, my apologies. I had the fans on the radiators running at 2300 rpm. My case fans seem to max out at 650 rpm

 

[Buzz247]

Damn, was hoping a knob turn or plug chk was it lol. Try the remount. Remember, bb size, put on flat not angled

 

[Divinenite]

Damn, was hoping a knob turn or plug chk was it lol. Try the remount. Remember, bb size, put on flat not angled

Yeah, but i may have found my problem. I do believe that i installed the outter most fan as a intake and the inner most fan and exhaust (meaning that both would be blowing towards the radiator). Could this cause the temps i was seeing?

 

[Buzz247]

Yes! No air is moving they are fighting each other. Mount both to blow out of case with rad sandwiched between

 

[Divinenite]

Yes! No air is moving they are fighting each other. Mount both to blow out of case with rad sandwiched between

Im just really at a loss now.... I re positioned my fans to now correctly being 2 exhausts fans sandwiching the rad. I also unmounted the pump, cleaned all the factory compound off both pump and cpu and then applied artic silver 5. I placed the pump square (even pulled up to make i sure i had a good spread.

20 minutes into the stress test i have already peeked at 89 C....... Idle is around 38 C. I feel like at this point, the facotry HS would be better then this pos lol. Do i have a deffective unit?

 

[Buzz247]

Pulling up ruined the contact, possibly trapping air. What did you use to clean?

Unmount
Clean with rubbing alcohol and qtip - both surfaces
Apply bb size dot - less is more with tim
Apply heatsink flat - do not pull up- usetwist n slide if minor adjust needed.
Tighten screws evenly little at a time on each to keep flat

Pc should be on its side for whole process. Do not do while verticle. Makes flat and keeping seal impossible

 

[Buzz247]

If tubes, rad, and air get warm your h80 is not defective

 

[Divinenite]

If tubes, rad, and air get warm your h80 is not defective

I really appreciate all the help man! 😀

I am removing it as we speak.Also, should i notice a decent temp change on the rad within 20 minutes? I only ask as i ran prim95 for 20 minute, cpu temps climbed to 87 C, but the rad and barely luke warm

 

[Buzz247]

Properly seated cpu prob about 65-70s on prime and rad air will feel warm as in more than blech luke warm. Be noticably warmer than ambient. Speaking of which, what is your room temp?

 

[Divinenite]

Properly seated cpu prob about 65-70s on prime and rad air will feel warm as in more than blech luke warm. Be noticably warmer than ambient. Speaking of which, what is your room temp?

Room is around mid to high 60's (F)

 

[Divinenite]

Im 15 minutes into prime95 4 worker test, and my temps have spiked up to 89 C. The first series of tests, it held wonderfully at 70 C. However, as soon as the next test started, i saw the spike. I had my fans at 2000 rpm, and i increased them to 2500 once i saw the spike. Currrently, i cant seem to get it below 85 C........

 

[Divinenite]

30 mins in, temp has knocked back down to 68 C, i also backed my fans down to 1.95k rpm. Holding steady. Is it normal for cpu's to spike to the upper 80's or low 90's with prime95?

 

[Buzz247]

given u r running an h80 on a 4670 - not surprising. Stress test is exactly that - a worst case scenario - not norm usage

Why did u back your fans down?

 

[Divinenite]

given u r running an h80 on a 4670 - not surprising. Stress test is exactly that - a worst case scenario - not norm usage

Why did u back your fans down?

I backed down the fan speed as the temps had normalized
And I wanted to see how the temps would handle the performance settings instead of Max settings.