Sign in with
Sign up | Sign in
Your question

Fan placement in regards to system cooling: a case study

Last response: in Overclocking
Share
August 22, 2011 7:57:02 PM

So, I opened a thread a week or so ago about the placement of my aftermarket cooler and additional fans to get the best cooling performance, and got a bit of a range of responses.

The cooler arrives today, and over the next week or so I plan to run a few tests on different configurations, to see which is best for different goals:
1) CPU cooling
2) Overall System cooling

I will have access to a temp probe/monitor to get average ambient temp over the duration of the test to account for differences. I plan to use Prim95 for CPU and EVGA OC Scanner for the GPU.

If there is better test programs or additional temp logging systems, I would appreciate it if you could suggest any. Also, if you can think of other test configurations than what follows, suggestions would also be welcome and appreciated.

1) Control (Stock cooler, side exhaust sealed)
2) Control Side (Stock cooler, fan on side exhaust)
3) Aftermarket/vertical/single fan (side exhaust sealed)
4) Aftermarket/vertical/single fan (with side exhaust fan)
5) Aftermarket/horizontal/push-pull (side exhaust sealed)
6) Aftermarket/horizontal/single fan (side exhaust sealed)
7) Aftermarket/horizontal/single fan (with side exhaust fan)

Thinking I may eliminate 3 & 4 just because I think that will have minimal difference, and I'd also like to save on the Arctic Silver since I'm edging on poor.

Setup being tested on:
Mobo: MSI P67A-GD53 (B3)
CPU: Intel i5-2500k
RAM: G.SKILL Ripjaws Series 4GB (2 x 2GB) DDR3 1600
GPU: EVGA GTX 560 Ti FPB (Fermi) 1GB 256-bit GDDR5
PSU: CORSAIR Enthusiast Series TX850 V2
HDD: SAMSUNG Spinpoint F3 1TB 7200 RPM 32MB Cache
Case: Antec 300 Illusion
Cooler: COOLER MASTER Hyper 212 Plus
Thermal Compound: Arctic Silver 5
Extra Fan: Scythe DFS123812-3000 "ULTRA KAZE"
a b K Overclocking
August 22, 2011 8:15:13 PM

Testing is always interesting if you can control the variables... There should not be a need to change the heatsink, if you're just interested in case airflow.

IME front in airflow and back out works best for most PC cases - but the PSU location has an effect on where the heated air exits also. Side fans may improve or actually hurt airflow if they cause turbulence.

Keep in mind that to get accurate data you would need to run a specific configuration for awhile under load for the components and case temp to stabilize. Ambient room temp will also affect any temp deltas.
August 22, 2011 8:35:45 PM

beenthere said:
Testing is always interesting if you can control the variables... There should not be a need to change the heatsink, if you're just interested in case airflow.

IME front in airflow and back out works best for most PC cases - but the PSU location has an effect on where the heated air exits also. Side fans may improve or actually hurt airflow if they cause turbulence.

Keep in mind that to get accurate data you would need to run a specific configuration for awhile under load for the components and case temp to stabilize. Ambient room temp will also affect any temp deltas.


Oh, should have clarified. My interest isn't so much in case airflow, but the best use of a second fan (basically push pull vs side exhaust), however I figured while I was at it, check it against the Stock cooler. So the focus of comparison would be between 5, 6, and 7. I figured it would be interesting to compare to stock while I was at it.

Temps will be rather hard to control directly (no AC) so I will have to look at temps over ambient.
Related resources
a b K Overclocking
August 22, 2011 9:15:54 PM

Other dependable software for temp measuring/reading is:

1. HW Monitor
2. SpeedFan
3. RivaTuner

This sounds like a very interesting experiment. Keep us posted on your results. Also, if you have the time, I'd be very interested in seeing results with side panels removed; just for shits and giggles.
August 23, 2011 7:58:59 PM



Interesting, but noticed the hardware was a bit older, be interesting to compare against new hardware, seeing as (hopefully) thermal designs have improved.

Secondly, I noticed the 'All Fans' was never far behind the top, maybe a degree C or two off, which most people I've heard refer to that as negligible in part due to ambient differences. They say ambient was within 1.5 degrees C range overall, which could account for that difference.
!