High temps with a new H100i V2 and i7-6700k

G

Guest

Guest
Hi,


  • Motherboard: Asus Z170i Pro Gaming
    Processor: i7-6700K at stock speed
    CPU Cooler: Corsair H100i V2 (with 2x Corsair SP120 Quiet PWM)
    Memory ram: Corsair Dominator Platinum 16GB 2x8GB 3200Mhz
    PSU: Corsair AX860i
    SSD: Samsung 850 EVO 1TB
    GPU: none yet

I installed the cooler today and the temps I have got surprised me.
Why they are too high?.


  • The CPU is not overcloked
    I'm using the thermal paste that came with the pump

Please see photos attached below...

What can I do?
Temps should be 25-30 degrees at idle.

571fc501aee72_Capturadepantalla(1).png.e9d48bfd413b969dabd66ae6840d609f.png


571fc50227160_Capturadepantalla(2).png.deb580252813ae34d5518ae2a80d6c2d.png


571fc5028e611_Capturadepantalla(3).png.fa1d07b50ca2bd6a2036a5f6334c5487.png


571fc502f39bb_Capturadepantalla(4).png.c6c6d93b215bcac3eb9f2e3d9a82705e.png


571fc503ce16d_Capturadepantalla(5).png.c65d1980c6434a585a6cb7b376add886.png


IMG_2606.thumb.JPG.197a2f06f894a8fd62949269c42fbb0a.JPG


IMG_2624.thumb.JPG.931b13cd3216eba96b7b307ee6e872c3.JPG
 
Solution
I think if you add an exhaust fan you will see a difference. an exhaust helps push the warm air out of the case, whereas without one that air is just trapped inside of your case.
Peeps have the misconception that "faux water cooling" aka CLC type units like the H100i, will give performance equivalent to a custom water loop. They don't. The reality is that 2 x 1200 CLC type water coolers do not cool as well as much cheaper air coolers while the CLCs can be up to 12 times as loud. When you turn down the fan speed to cut noise, they cool even less.

b2.jpg


Still... those temps are still high.

As a start, I'd switch from quiet mode to performance mode where the emphasis is on cooling rather than noise. And we won't know much about how it's doing till we see it perform under load ...

Also ditch the multiple utilities as they can interfere with one another. Would recommend testing with RoG Real Bench and HWiNFO 64 (Run sensors only) and record max temp on each core and max voltages during the test

http://dlcdnmkt.asus.com/rog/RealBench_v2.43.zip
http://www.hwinfo.com/
 

Quixit

Reputable
Dec 22, 2014
1,359
0
5,960
Those temperatures at totally safe and you don't need to worry. You would get better performance with faster, louder fans like the stock fans. Corsair SP120 fans are less powerful than the stock ones. But if sound level is what you're concerned with you don't need to do anything.

You'd also get better performance with the fans pulling in air from outside the case.
 

tical2399

Distinguished
Nov 20, 2009
1,081
0
19,460


Sounds like a straw man argument to me. Where are these peeps that were under the impression that a clc cooled as well as custom loop. As far as air coolers doing as well a clc? Well me and my Miami summers call bs on that statement. My H100i does way better than the world famous 212 evo that I had prior.
 
G

Guest

Guest


I installed the pump with a screw driver, maybe that is wrong?, should I use just my fingers?.
 

tical2399

Distinguished
Nov 20, 2009
1,081
0
19,460
A screwdriver is fine as long as you're not tightening it too much, like using alot of force when its almost totally screwed in. Also, I never use the included thermal paste. I'm sure it's fine but you cant go wrong with a small tube of arctic silver
 
G

Guest

Guest


I have here a tube of Noctua NT-H1, I will use it
 
G

Guest

Guest


Using Asus Real Bench

2iqzjwj.jpg
 
G

Guest

Guest


Using Asus Real Bench

2iqzjwj.jpg
 
G

Guest

Guest


Not right now, I have the case open, without the panels.

Should I remove my 200mm fan at the top and install two 140mm fans?, I have two here, Corsair AF140mm Quiet Edition.
or one 200mm is better?
 
G

Guest

Guest


Should I remove my 200mm fan at the top and install two 140mm fans?, I have two here, Corsair AF140mm Quiet Edition.
or one 200mm is better?
 


Don't ya think most who have bought CLCs did so because they **thought** that they did something better than air coolers ? Those that I have spoken to just "assumed" that they were actually similar to custom water loops but were pre-assembled at the factory. But most do not realize that:

- CLC's do not cool as well as air coolers in their price range (air usually 10%+ cheaper)
- CLC are typically 8 - 12 times louder than air coolers in their price range
- CLCs use cheap aluminum radiators with low thermal transfer co-efficient.
- CLC's defy the first rule of water cooling and mix copper blocks with aluminum radiators .. that's a bad thing
https://martinsliquidlab.wordpress.com/2012/01/24/corrosion-explored/
- CLCs defy the 2nd rule of water cooling to provide 1 gpm of flow throughout the loop.
- CLCs must use high rpm fans to offset the poor thermal transfer efficiency of the aluminum rads and the low flow (0.11 gpm pumps). Changing the fans to slower rpm units significantly gimps performance
- CLCs do this
http://www.tomshardware.com/forum/281843-29-corsair-exploded

One argument made is that they are relieving the 2 pound force that the large heat sink places on the MoBo ... so says they guy who just applied 60 pounds of clamping force while tightening down the water block screws.
http://www.innovationcooling.com/key_to_performance.html

Many peeps who think that these two statements mean the same thing

"The Hyper 212 is a great budget cooler"
"The Hyper 212 is a great cooler"

As for your BS, numbers don't lie.

The hyper 212 is a great cooler ... for the money. No argument there; it is not however a great cooler when compared with the likes of the Thermalright Silver Arrow, Phanteks PH-TC14-PE, Noctua NH D14 & D15, Cryorig R1 ... all of which are cheaper than the H100i

So, yes the $100 H100i is better than the $25 Hyper 212 costing 1/4 as much.... but isn't that a bit like saying nVidia is better than AMD cause the GTX 980 Ti is better than the R9 270 ?

And I didn't present any data (see image above) on the Hyper 212, did I ? The the only "strawman" created was the one which argues that CLCs are better than air coolers because a $100 CLC outperforms a $25 air cooler.... not exactly apples and apples is it ?

So lets go back to apples and apples and data I **actually** presented for the the Noctua DH15 and Cryorig R1 which as we can clearly see in the graphic above:

- The $89 Noctua DH15 air cooler thermally outperforms the $100 H100i CLC by 2C
- The H100i CLC is 12 times louder than the D15 while it loses thermally to the air cooled D15

- The $90 Cryorig R1 air cooler thermally outperforms the $100 H100i CLC by 3C
- The H100i is 8 times louder than the R1 while it loses thermally to the air cooled R1

So, no ... no strawmen there ... the graphic shows that:

1. The referenced air coolers are, in fact, cheaper ($10-$100) than the H100i, like I said
2. The referenced air coolers do, in fact, cool better (2-3C) than the H100i, like I said
3. The referenced air coolers are, in fact, 8 - 12 times quieter than the H100i, like I said

Given the above test results ..

Wouldn't your Miami based PC be better served if your CPU was 2-3C cooler ?
Wouldn't you be better served if your CPU PC wasn't 8 - 12 times as loud ? ?





 
G

Guest

Guest


Looks this video that I have uploaded too:
http://es.tinypic.com/r/357gadl/9
 

john_346

Commendable
Apr 25, 2016
17
0
1,520
The way I setup clc is usually air coming in from front or back through radiator and exhaust fan right next to the radiator up above to pull the hot air out. I'm usually around 28-32 during idle and hit 67c under torture test and that's with a cheaper clc than you have.
 


1. Read the manual. Corsair recommends (consistent with all custom loop installs) that for best cooling, fans should be used as intakes. Expect a reduction of up to 50% if you don't do this. And yes, we all learned in 8th grade earth science that hot air rises ... but not when faced with a high pressure fan. What matters here is thermodynamics and here's how that goes like this:

Ambient Air = 23C
Inside Case Air = 28C
Coolant Temp = 33C

Amount of cooling is proportional to Delta T:

Intake: 33C - 23C = 10C
Exhaust: 33C - 28C = 5C

In the above example, using outside air to cool your radiator is twice as effective as using inside the case air. The next statement usually is related to the effect of the air warmed by the radiator on other components. I am not aware of any component that is affected in a negative way by 28C air.

http://www.corsair.com/~/media/corsair/download-files/manuals/49-000175_rev_ab_h100i_qsg_web.pdf

Your best bet would be to:

a. Install two (2) 140mm case fans as intakes (push) in the top of the case.
b. Install two (2) 140mm case fans as intakes (pull) on the bottom of the radiator.
c. Link all fans to a single MoBo header (provided that all fans amperage combined does not exceed 1 amp). Alternately you could use Corsair Link but I have to confess to a complete absence of knowledge there. We do not install or recommend CLCs so our experience is limited to custom water loops and OLCs like the H240-X
d. Phanteks F-140SPs would best, but as you already have the Corsairs just be aware the Corsair PWM fans are known to lose speed control if too many are ganged on a single channel. You should be OK with 4 tho, its at 5 or 6 that the problem normally arises.

2. As for the case, figure one 140mm for every 100 watts of power .... 75 watts if ya want quiet

980 Ti - 250 watts (300 OC's)
CPU - 90 watts (135 OC'd)
MoBo - 30 watts (40 OC's
All else - 40 watts (40 OC'd)

So I'd use 5 fans as a minimum, 7 to run at 850ish rpm and dead quiet.

Ya have 4 on top.... 2 intakes in front (or bottom) and 1 on rear will do it as it seems you are shooting for quiet.


3. Again, there's far to many utilities running at one time so as to rule out them interfering with one another. Run RoG Real Bench w/ just HWiNFO loaded (run sensors only).


4. Your fans are running very slow.... under RoG Real bench they should be up around 850 rpm or higher if temps are hitting 80s. At stock settings, I am at 48-53 under RoG Real Bench
 

Gamer1985

Reputable
Dec 19, 2015
622
0
5,360
My H100I keeps my 6700K at 22-28C at idle, with an ambient of around 21C. I have positive airflow with 7 fans in my case. My CPU never goes above 58C under load for long periods of time. I didnt buy my AIO because I thought it would be better then air nor did I in thinking it would be as good as a custom loop. I bought it because I knew it was effective and it kept a nice sleek aesthetic profile going throughout my build. I think the Corsair series AIOs are great and mine is not loud at all and the fans are set at performance/turbo. Maybe because my corsair case does a good job of sound dampening but I hear my GPUs more then my pump or radiator fans.

OP yours is a little higher but make sure you also have a combination of good airflow in your case as well. You have nothing to be concerned with as far as your temps though. They really arent bad.
 
G

Guest

Guest


My case is a Phanteks Mini ITX with a 200mm fan at the Front, this fan came with the case.
At the back I don't have a fan because I will install on it the radiator of my GPU in the future.

I have tried Performance mode and Balance mode with a program to stress the CPU and I get the same temps, no matter what settings I have using Corsair Link.

I remove the pump and install it again with a new thermal paste from Noctua and nothing, the same temps.
I had better temps with a low profile cooler from Noctua.

Remember that my CPU is not overcloked..

:(

2mdjkut.jpg
 

john_346

Commendable
Apr 25, 2016
17
0
1,520
If it was me and this just my opinion that case is not well suited for clc but if I had to work with it I would face the fans on the radiator to push air out of the case even though it's not recommended and same when you put in another radiator for the gpu and make sure you have as much air as possible coming in through the front and as long as your max is under 70 I wouldn't worry about it. Also make sure your pump is running at full speed or close to it.
 

tical2399

Distinguished
Nov 20, 2009
1,081
0
19,460


Seems you dont know what a straw man argument is. A straw man is you shooting down an argument that nobody made. "many peeps believe". Where are these peep? Do you have links to any posts where they specifically said they though that their clc would cool just as well as a custom loop? If so please link.
 

Pubudu_1

Commendable
Oct 5, 2016
2
0
1,510