Overclock.net › Forums › Cooling › Air Cooling › Fans and cooling, did you know?
New Posts  All Forums:Forum Nav:

Fans and cooling, did you know?

post #1 of 13
Thread Starter 
I was fooling around and testing my OC, and I decided to try and reseat my CPU coolers fan on the top so that instead of pushing the air to the heat sink, it will pull the air in the heat sink.
Before I did that, I read a lot about it on the internet and everybody was under the "logical" impression that pushing the air would be better than pulling the air.

I didn't see how that was logical at all, of course this depends what kind of heat sink you have, and what case for that matter. My prediction was that the cooling performance would be the same, but I I didn't expect to see this to be honest.

The tests were performed with OCCT CPU test, I let it run for 10min to warm up.

These are the results with the CPU cooler pushing the air.



These are the results with the CPU cooler pulling the air.

(Note that the temperatures are rounded, in fact in the previous
test the CPU surface temp. was close to 48°C Core temps are what matter anyway..)


I was amazed, apparently my CPU cooler prefers the fan to pull the air in the heat sink, I think this has something to do with my case airflow, I even double checked and confirmed the results.
After reseating the fan, I had more free space and was able to fit an additional fan on the side panel blowing air on the motherboard itself, pulling cool air from outside.

These are the results with only one additional case fan.


I was even more amazed by the difference only one additional case fan made for my results.
I wasn't really expecting this much difference with all the tests I did, I'm running all my case fans on low for noise reasons, I wonder how cool my system would be with all 8 fans running at full power.

Conclusion: Experiment with your airflow and fans, even seemingly unlogical fan placement can improve temperatures, you need to consider everything, CPU fan direction, heat sink, and case airflow and whatnot.
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 965BE Asus M4A89GTD Pro/USB3 HD 5870 4GB 1600 CL8 
OSMonitorKeyboardPower
Windows 7 x64 24" 1920x1080 Logitech G11 Antec TPQ-850W 
CaseMouseMouse Pad
Antec 1200 Razer Imperator Steelseries QcK 
  hide details  
Reply
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 965BE Asus M4A89GTD Pro/USB3 HD 5870 4GB 1600 CL8 
OSMonitorKeyboardPower
Windows 7 x64 24" 1920x1080 Logitech G11 Antec TPQ-850W 
CaseMouseMouse Pad
Antec 1200 Razer Imperator Steelseries QcK 
  hide details  
Reply
post #2 of 13
+rep
BBK
(14 items)
 
  
CPUMotherboardGraphicsRAM
INTEL CORE I5 2500K @ 4.5GHZ INTEL DP67BG  XFX GEFORCE GTX 275 896MB G.SKILL 8GB 1600 DDR3 
Hard DriveOptical DriveCoolingOS
WESTERN DIGITAL LG INTERNAL BLU RAY BURNER COOLER MASTER V10 WINDOWS 7 ULTIMATE 64-BIT 
MonitorKeyboardPowerCase
ASUS 22" LCD MONITOR LOGITECH MK710 THERMALTAKE 750WATTS COOLER MASTER HAF932 
Mouse
LOGITECH M705 
  hide details  
Reply
BBK
(14 items)
 
  
CPUMotherboardGraphicsRAM
INTEL CORE I5 2500K @ 4.5GHZ INTEL DP67BG  XFX GEFORCE GTX 275 896MB G.SKILL 8GB 1600 DDR3 
Hard DriveOptical DriveCoolingOS
WESTERN DIGITAL LG INTERNAL BLU RAY BURNER COOLER MASTER V10 WINDOWS 7 ULTIMATE 64-BIT 
MonitorKeyboardPowerCase
ASUS 22" LCD MONITOR LOGITECH MK710 THERMALTAKE 750WATTS COOLER MASTER HAF932 
Mouse
LOGITECH M705 
  hide details  
Reply
post #3 of 13
Awesome testing. One thing to keep in mind though is that your CPU temp sensor is probably only accurate to within 5C or so. Don't take it as gospel.
Kathis
(16 items)
 
  
CPUMotherboardGraphicsRAM
2600k Maximus V Formula EVGA GTX 670 Kingston something 
Hard DriveHard DriveCoolingOS
OCZ Vertex 4 ~ 4 terabytes across 3 drives, not in RAID. 2 ... Water cooled, aquaero 5 LT controller/temp monitor Win7 x64 
MonitorMonitorKeyboardPower
Acer 23.3", 1920x1080 Acer 19", 1440x900 CM Storm Quickfire 700W Cooler Master Silent Pro 
CaseMouseMouse PadAudio
NXZT Switch 810 Logitech G9X Yup.  E-MU 1820m running into a Pyle 120wx2 amp, driv... 
  hide details  
Reply
Kathis
(16 items)
 
  
CPUMotherboardGraphicsRAM
2600k Maximus V Formula EVGA GTX 670 Kingston something 
Hard DriveHard DriveCoolingOS
OCZ Vertex 4 ~ 4 terabytes across 3 drives, not in RAID. 2 ... Water cooled, aquaero 5 LT controller/temp monitor Win7 x64 
MonitorMonitorKeyboardPower
Acer 23.3", 1920x1080 Acer 19", 1440x900 CM Storm Quickfire 700W Cooler Master Silent Pro 
CaseMouseMouse PadAudio
NXZT Switch 810 Logitech G9X Yup.  E-MU 1820m running into a Pyle 120wx2 amp, driv... 
  hide details  
Reply
post #4 of 13
Quote:
Originally Posted by naikee View Post
I was fooling around and testing my OC, and I decided to try and reseat my CPU coolers fan on the top so that instead of pushing the air to the heat sink, it will pull the air in the heat sink.
Before I did that, I read a lot about it on the internet and everybody was under the "logical" impression that pushing the air would be better than pulling the air.

I didn't see how that was logical at all, of course this depends what kind of heat sink you have, and what case for that matter. My prediction was that the cooling performance would be the same, but I I didn't expect to see this to be honest.

The tests were performed with OCCT CPU test, I let it run for 10min to warm up.

These are the results with the CPU cooler pushing the air.



These are the results with the CPU cooler pulling the air.

(Note that the temperatures are rounded, in fact in the previous
test the CPU surface temp. was close to 48°C Core temps are what matter anyway..)


I was amazed, apparently my CPU cooler prefers the fan to pull the air in the heat sink, I think this has something to do with my case airflow, I even double checked and confirmed the results.
After reseating the fan, I had more free space and was able to fit an additional fan on the side panel blowing air on the motherboard itself, pulling cool air from outside.

These are the results with only one additional case fan.


I was even more amazed by the difference only one additional case fan made for my results.
I wasn't really expecting this much difference with all the tests I did, I'm running all my case fans on low for noise reasons, I wonder how cool my system would be with all 8 fans running at full power.

Conclusion: Experiment with your airflow and fans, even seemingly unlogical fan placement can improve temperatures, you need to consider everything, CPU fan direction, heat sink, and case airflow and whatnot.
Great post!
Gizmoto
(18 items)
 
They beat me:(
(15 items)
 
Home Server V1.0
(17 items)
 
CPUMotherboardGraphicsRAM
AMD Phenom II x6 1040t @2.7Ghz ASRock 880GM-LE FX  AMD Radeon HD 4250 integrated graphics Micro Center(ADATA) 8Gb (2x4) 1333 
Hard DriveHard DriveHard DriveHard Drive
Western Digital Caviar Blue 640Gb Samsung SpinPoint F4 2Tb Seagate Barracuda 7200.9 ST3500641AS-RK 500Gb Seagate Barracuda 7200.9 ST3500641AS-RK 500Gb 
Optical DriveOptical DriveCoolingOS
ASUS 24X DVD Burner ASUS 24X DVD Burner Stock AMD Cooler Windows 7 Home Premium 64 bit 
OSOSPowerCase
Linux Mint (VM) Linux Ubuntu (VM) Inland ILG-500R2 500w PSU DiabloTek Evo Mid-tower Case 
Other
Inland IL internal card reader 
  hide details  
Reply
Gizmoto
(18 items)
 
They beat me:(
(15 items)
 
Home Server V1.0
(17 items)
 
CPUMotherboardGraphicsRAM
AMD Phenom II x6 1040t @2.7Ghz ASRock 880GM-LE FX  AMD Radeon HD 4250 integrated graphics Micro Center(ADATA) 8Gb (2x4) 1333 
Hard DriveHard DriveHard DriveHard Drive
Western Digital Caviar Blue 640Gb Samsung SpinPoint F4 2Tb Seagate Barracuda 7200.9 ST3500641AS-RK 500Gb Seagate Barracuda 7200.9 ST3500641AS-RK 500Gb 
Optical DriveOptical DriveCoolingOS
ASUS 24X DVD Burner ASUS 24X DVD Burner Stock AMD Cooler Windows 7 Home Premium 64 bit 
OSOSPowerCase
Linux Mint (VM) Linux Ubuntu (VM) Inland ILG-500R2 500w PSU DiabloTek Evo Mid-tower Case 
Other
Inland IL internal card reader 
  hide details  
Reply
post #5 of 13
These results are only specific to actual CPU temps.

The reason people say "pushing" is better, is due to all the other cooling benefits it has. Such as, pulling hot air away from the RAM, pushing a little bit of air over the northbridge and motherboards VRMs, and a little bit of air on the top of the GPU PCB (depending how close it is). Basically, putting it as a "push" on the front of a heatsink allows better overall case airflow, especially if your cases airflow is poor.
    
CPUMotherboardGraphicsRAM
Intel Core i5 4690K MSI Z97 Mpower MSI GTX 770 Lightning Mushkin Blackline 2x8GB DDR3-2400 
Hard DriveOptical DriveCoolingOS
840 Pro 128GB + WD 640GB Black + WD 1TB Black +... Samsung SH-S203N DVD-RW Phanteks PH-TC14PE_BK w/ 2x SP120's Windows 7 Professional x64 
MonitorKeyboardPowerCase
ASUS VE258Q 25" LED Corsair K70 RGB (Cherry MX Red's) Corsair HX750 Silverstone TJ04-E w/ Window 
MouseMouse PadAudioAudio
Logitech G5 v1 X-ray Thunder 9 smooth ASUS Xonar Essence STX Numark NPM5's + Dayton SUB-1200 12" sub 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i5 4690K MSI Z97 Mpower MSI GTX 770 Lightning Mushkin Blackline 2x8GB DDR3-2400 
Hard DriveOptical DriveCoolingOS
840 Pro 128GB + WD 640GB Black + WD 1TB Black +... Samsung SH-S203N DVD-RW Phanteks PH-TC14PE_BK w/ 2x SP120's Windows 7 Professional x64 
MonitorKeyboardPowerCase
ASUS VE258Q 25" LED Corsair K70 RGB (Cherry MX Red's) Corsair HX750 Silverstone TJ04-E w/ Window 
MouseMouse PadAudioAudio
Logitech G5 v1 X-ray Thunder 9 smooth ASUS Xonar Essence STX Numark NPM5's + Dayton SUB-1200 12" sub 
  hide details  
Reply
post #6 of 13
Quote:
Originally Posted by naikee View Post
I was fooling around and testing my OC, and I decided to try and reseat my CPU coolers fan on the top so that instead of pushing the air to the heat sink, it will pull the air in the heat sink.
Before I did that, I read a lot about it on the internet and everybody was under the "logical" impression that pushing the air would be better than pulling the air.

I didn't see how that was logical at all, of course this depends what kind of heat sink you have, and what case for that matter. My prediction was that the cooling performance would be the same, but I I didn't expect to see this to be honest.

The tests were performed with OCCT CPU test, I let it run for 10min to warm up.

These are the results with the CPU cooler pushing the air.



These are the results with the CPU cooler pulling the air.

(Note that the temperatures are rounded, in fact in the previous
test the CPU surface temp. was close to 48°C Core temps are what matter anyway..)


I was amazed, apparently my CPU cooler prefers the fan to pull the air in the heat sink, I think this has something to do with my case airflow, I even double checked and confirmed the results.
After reseating the fan, I had more free space and was able to fit an additional fan on the side panel blowing air on the motherboard itself, pulling cool air from outside.

These are the results with only one additional case fan.


I was even more amazed by the difference only one additional case fan made for my results.
I wasn't really expecting this much difference with all the tests I did, I'm running all my case fans on low for noise reasons, I wonder how cool my system would be with all 8 fans running at full power.

Conclusion: Experiment with your airflow and fans, even seemingly unlogical fan placement can improve temperatures, you need to consider everything, CPU fan direction, heat sink, and case airflow and whatnot.
Have you tried push/pull yet?
Ethylene Glycol
(13 items)
 
  
CPUMotherboardGraphicsRAM
Core 2 Duo E7300 DFI LanParty DK X38 series Nvidia GTX 260 4gb of Mushkin 
Hard DriveOptical DriveOSMonitor
Maxtor 250GB SATA Lite-on DVD rom drive (soon to be upgraded) Windows 7, but can run Mac OSX and Linux. Hitachi CM721F flat screen CRT 
KeyboardPowerCaseMouse
Gateway Sk9920 Corsair TX750 Dunno who made the case, but its modded! HP Voodoo DNA laser mouse 
Mouse Pad
N/A 
  hide details  
Reply
Ethylene Glycol
(13 items)
 
  
CPUMotherboardGraphicsRAM
Core 2 Duo E7300 DFI LanParty DK X38 series Nvidia GTX 260 4gb of Mushkin 
Hard DriveOptical DriveOSMonitor
Maxtor 250GB SATA Lite-on DVD rom drive (soon to be upgraded) Windows 7, but can run Mac OSX and Linux. Hitachi CM721F flat screen CRT 
KeyboardPowerCaseMouse
Gateway Sk9920 Corsair TX750 Dunno who made the case, but its modded! HP Voodoo DNA laser mouse 
Mouse Pad
N/A 
  hide details  
Reply
post #7 of 13
The problem with doing testing is all the variables you must control. For example, always record your ambient temps.

But actually testing temps is the best way to proceed. You always learn things by getting data.

+rep.
alpha updated
(19 items)
 
secundus
(14 items)
 
 
CPUMotherboardGraphicsRAM
i7 4790k Gigabyte GA-Z97X Gaming-7 Intel HD4600 Crucial Ballistix Sport Very Low Profile 8GB Ki... 
Hard DriveCoolingOSMonitor
Samsung 840 EVO 500GB Prolimatech Megahalems Rev. C Windows 8.1 Home Premium 64-bit Acer K242HL 
KeyboardPowerCaseMouse
Dell SK-8110 (PS/2) Seasonic X460 Fanless motherboard tray from CM ATCS 840 Logitech MX 1100 
Mouse PadAudio
Dell Creative SoundBlaster X-Fi MB3 
CPUMotherboardGraphicsRAM
i7 860 Gigabyte GA P55M UD2 ZOTAC/NVidia GTX 650 Ti 2GB 4x4GB G.Skill Ripjaws DDR3-1333 CAS9 
Hard DriveOptical DriveCoolingOS
Samsung 1TB Spinpoint F3 7200.12 Plextor 24x DVD burner Cooler Master Hyper Z600R Win7 64-bit 
MonitorKeyboardPowerCase
Dell E207WFP 20-inch flat panel Logitech Wireless 510 Seasonic X-650 fully modular 80+ Gold Lian Li PC-7FN 
MouseMouse PadAudio
Wireless Mouse WACOM Intuos 3 generic 
  hide details  
Reply
alpha updated
(19 items)
 
secundus
(14 items)
 
 
CPUMotherboardGraphicsRAM
i7 4790k Gigabyte GA-Z97X Gaming-7 Intel HD4600 Crucial Ballistix Sport Very Low Profile 8GB Ki... 
Hard DriveCoolingOSMonitor
Samsung 840 EVO 500GB Prolimatech Megahalems Rev. C Windows 8.1 Home Premium 64-bit Acer K242HL 
KeyboardPowerCaseMouse
Dell SK-8110 (PS/2) Seasonic X460 Fanless motherboard tray from CM ATCS 840 Logitech MX 1100 
Mouse PadAudio
Dell Creative SoundBlaster X-Fi MB3 
CPUMotherboardGraphicsRAM
i7 860 Gigabyte GA P55M UD2 ZOTAC/NVidia GTX 650 Ti 2GB 4x4GB G.Skill Ripjaws DDR3-1333 CAS9 
Hard DriveOptical DriveCoolingOS
Samsung 1TB Spinpoint F3 7200.12 Plextor 24x DVD burner Cooler Master Hyper Z600R Win7 64-bit 
MonitorKeyboardPowerCase
Dell E207WFP 20-inch flat panel Logitech Wireless 510 Seasonic X-650 fully modular 80+ Gold Lian Li PC-7FN 
MouseMouse PadAudio
Wireless Mouse WACOM Intuos 3 generic 
  hide details  
Reply
post #8 of 13
Thread Starter 
All the tests were dune under the same ambient temperature, so it should be pretty accurate.
Don't take the benchmarks too seriously though, my point of this whole thing, that such a small thing can make such a big difference, as in one case fan or simply different fan placement.
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 965BE Asus M4A89GTD Pro/USB3 HD 5870 4GB 1600 CL8 
OSMonitorKeyboardPower
Windows 7 x64 24" 1920x1080 Logitech G11 Antec TPQ-850W 
CaseMouseMouse Pad
Antec 1200 Razer Imperator Steelseries QcK 
  hide details  
Reply
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II x4 965BE Asus M4A89GTD Pro/USB3 HD 5870 4GB 1600 CL8 
OSMonitorKeyboardPower
Windows 7 x64 24" 1920x1080 Logitech G11 Antec TPQ-850W 
CaseMouseMouse Pad
Antec 1200 Razer Imperator Steelseries QcK 
  hide details  
Reply
post #9 of 13
I prefer the pull method because it ensures that all the hot air coming off the heatsink doesn't have a chance to diffuse (as it would in a push config through the fins) before being rammed out the rear case fan. Ideally of course push-pull so you get hot air rammed out and fresh air rammed in.

It can also be a toss up depending on the fan you're using on the heatsink. Some fans don't like the back pressure created by the push config, some don't like the intake vacuum created by the pull config.
Edited by dr/owned - 5/5/11 at 4:47pm
post #10 of 13
Gonna try it now.....
Liquid
(18 items)
 
  
CPUMotherboardGraphicsGraphics
3930k ASUS Rampage IV Gene EVGA GTX480 @ 900MHz @ 1.1V EVGA GTX480 @ 900MHz @ 1.1V 
RAMHard DriveOptical DriveOS
4 x 4Gb Corsair Dominator 1866MHz Samsung 830 256GB LG DVD Burner Win 7 x64 
MonitorMonitorMonitorKeyboard
DELL U3011 DELL U3011 DELL U2410 Ducy Shine 2 
PowerCaseMouseAudio
Enermax Platinum 1200W Silverstone FT03 G500 Native Instruments Komplete Audio 6 
Audio
Yamaha HS80m 
  hide details  
Reply
Liquid
(18 items)
 
  
CPUMotherboardGraphicsGraphics
3930k ASUS Rampage IV Gene EVGA GTX480 @ 900MHz @ 1.1V EVGA GTX480 @ 900MHz @ 1.1V 
RAMHard DriveOptical DriveOS
4 x 4Gb Corsair Dominator 1866MHz Samsung 830 256GB LG DVD Burner Win 7 x64 
MonitorMonitorMonitorKeyboard
DELL U3011 DELL U3011 DELL U2410 Ducy Shine 2 
PowerCaseMouseAudio
Enermax Platinum 1200W Silverstone FT03 G500 Native Instruments Komplete Audio 6 
Audio
Yamaha HS80m 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Air Cooling
Overclock.net › Forums › Cooling › Air Cooling › Fans and cooling, did you know?