1 Fan vs 2: Chapter 2 - Push vs Pull - Overclock.net - An Overclocking Community

Forum Jump: 

1 Fan vs 2: Chapter 2 - Push vs Pull

 
Thread Tools
post #1 of 8 (permalink) Old 08-09-2012, 09:14 PM - Thread Starter
Retired Reviewer
 
ehume's Avatar
 
Join Date: Dec 2009
Location: New Jersey, USA
Posts: 10,070
Rep: 1372 (Unique: 855)

OK. So some of you said that with mild overclocks a pull fan might give better temps than a push fan.

 

So I added some pull-only results to my prior. The setup:

 

i7 860 @4GHz, Vcore set to 1.28750v.

Megahalems with one or two Slip Stream fans.

 

The results (click to get a bigger pic; right-click the bigger pic to see it full sized):

 

1000

 

So a low RPM pull fan does not cool as well as a low RPM push fan. But when the revs go up the pull fan cools better than a single push fan.

 

Hmm. I wonder what happens when the system gets hot? I pushed it to 4004MHz with Vcore set to 1.31250v:

 

1000

 

The slower fans won't cool this chip at that OC, so I could only use the fastest three pairs of fans, but the results are clear: at high overclock the pull-only fan does not cool as well as the push-only fan.

 

And 2 fans does get you a little better cooling.

ehume is offline  
Sponsored Links
Advertisement
 
post #2 of 8 (permalink) Old 08-09-2012, 09:40 PM
New to Overclock.net
 
Join Date: Jan 2010
Location: Florida
Posts: 8,266
Rep: 238 (Unique: 214)
so confirmed my testing in my push vs pull vs push pull

so at lower clocks/temps, pull is better, although I would generalize and say that pull is usually better because normally the cpu isn't stressed that far, so because the temps are lower normally, pull is indeed better for general use.

However I would like to point out, I think at least, ehume has his rear grill cut out, which helps when it comes to the pull. With a rear fan however, would pull still be better ehume? I don't know if you have a case with a rear grill still attached to test this.
chinesethunda is offline  
post #3 of 8 (permalink) Old 08-09-2012, 10:31 PM
New to Overclock.net
 
airdeano's Avatar
 
Join Date: Feb 2012
Location: little ark, rockansas
Posts: 1,735
Rep: 127 (Unique: 98)
where'd my comment go?? *sniffles*
i even dropped a + rep..

airdeano

airdeano is offline  
Sponsored Links
Advertisement
 
post #4 of 8 (permalink) Old 08-10-2012, 05:42 AM
New to Overclock.net
 
doyll's Avatar
 
Join Date: Aug 2011
Location: UK
Posts: 25,124
Rep: 1641 (Unique: 920)
Good information as usual. Thanks. thumb.gif

Phanteks Owners Club Ways to Better Cooling
i7 980 @ 3.55GHz =PH-TC14PE w/2x TY-143 fans =Crucial Ballistix 3x4GB =GA-X58A-UD5 =ASUS GTX580 DirectCU II =Enermax Modu84+ =Define R2 w/3x TY-140 case intake fans; all PWM controlled by CPU fan socketPhanteks Enthoo Primo MoBo Rampage III Extreme CPUi7 980X
@ 4.0GHz =R1 Ultimate w/2x TY-143 fans =Dominator GT 6x2GB =Rampage III Extreme =ASUS GTX580 DirectCU II =TX850 =Enthoo Primo w/ custom castor base
doyll is offline  
post #5 of 8 (permalink) Old 08-10-2012, 10:44 AM
New to Overclock.net
 
nleksan's Avatar
 
Join Date: Dec 2011
Location: Ohio
Posts: 5,505
Rep: 660 (Unique: 521)
This is very useful information, thank you!

I wonder if there is any gain or loss if you were to use a different-sized fan for pull than you did for push, in a two-fan setup? I ask because my first build was an AMD FX51 +SK8V + X800XT PE and the CPU was cooled by a Thermaltake Silent-Tower HSF. It used 4 heatpipes and had an aluminum shroud around the fins, supposedly to act as an air chamber (to keep air from leaking out), and came with a 92mm octagonal fan in a "push" position. Interesting enough, the mounts for the opposite side (for adding a pull fan) are both 92mm and 80mm spacing, with the recommendation being to use an 80mm fan if you choose to use a second fan.
Now, this computer was built back in 2004 (!) but it's something that I have always wondered about, as I did run the HS with the stock 92mm fan pushing, but about a year after I built it I wanted to push it a bit harder so I replaced most of the case fans with 30-50% stronger ones, and added an 80mm fan to "pull" air through the heatsink and hopefully augment the 92mm and allow higher overclocks and/or better temps.
The result was that I was able to push the chip to 2510mhz from 2200mhz stock, and the RAM was also able to go a bit further as well despite being registered/ECC Corsair XMS3200RE (slightly worse latency than regular RAM, but the stuff was indestructible!). Prior to changing all of this, I would get stuck at about 2350mhz, give or take.
Also, the temps were only a degree or two higher despite the significant increase in overclock, and the OC'd X800XT PE dumping heat into the case faster than it could be removed.

I am sorry I don't have any more recent examples, and this is something that has always bugged me... I figured that the difference in fan size and speed may have contributed to an increase in air turbulence, thus increasing the cooling potential of a given quantity of air? Also, before I forget, there was an exhaust fan about 2-3inches behind the 80mm "pull" fan (exhaust fan was 92mm) and I never could figure out if they were mutually beneficial or harmful...

*Case Fans were as follows: 1x 92mm front-lower intake, 1x 92mm top exhaust, 2x 80mm side intake, 1x 92mm rear exhaust, 1x 80mm PCI slot exhaust-blower two slots below the X800, 1x 60mm South-Bridge fan (added by me), and the top PSU which had 2x 92mm fans and exhausted air outside (1x fan on the bottom pulled air thru the PSU from the case, and the second one blew it out the rear of the case).


I decided a while back that I wanted to get this bad boy up and running again, as it "died" a few years ago but I couldn't ever figure out what happened. So I have the entire thing taken apart, and I have been slowly cleaning every bit of every part, lapped the HSF and CPU, cleaned the GPU'S heatsink, meticulously washed and re-lubed every fan, and so forth. I miss the days when overclocking was a nerve-racking experience, the BIOS was a confusing DOS screen, and the FX51 was the absolute fastest consumer CPU ever and the first time that AMD had beaten Intel not only in the consumer market but also the Enthusiast market.
I have also stripped the case (a Tt Xaser V Extreme) down to the bare aluminum and I am trying to figure out how best to modify it... I have cut out the fan grills, but I really want to see how modern air cooling affects the performance of a system built 8 years ago. To that end, I am changing the top exhaust by cutting out the existing hole to 120mm and adding a second exhaust hole for a 92mm fan, the front intake will be cut out to fit a 120mm and a second front intake will be added by fitting a 120mm fan into the spare 5.25 bays, a 120mm fan will be fitted to the HDD rack to help "pull" the air that the front- lower intake is pushing, the rear exhaust fan will be upped to 100mm, the bottom of the case will have two holes cut out for 120mm fans, the side panel fans will be switched from 2x80mm to 1x120mm, and a high-speed 92mm fan will be mounted to the inside of the rear PCI slots with the covers removed in order to exhaust the GPU heat. Lastly, I am going to be cutting my own CPU backplate opening in the MB tray, cutting the behind-the-MB-side side-panel to support an 80-92mm fan and use it to exhaust hot air from behind the CPU (seen it done in modern cases, figure it can't hurt).
Coupled with some very much improved MB cooling (VRM/MOSFET Heatsinks, a much-larger SB Heatsink, direct airflow from a duo of 60x15mm fans, and thermal paste with a drop of super glue on 2 corners instead of thermal pads), the newly-lapped HSF and CPU, two new CPU fans (1x 120x38mm PMB-MAT 2600rpm Panaflo 138cfm/8.3mmHg SP pushing and a second 120x25mm Scythe Slipstream 1900rpm/110cfm fan pulling with a 25mm thick plexiglass shroud between it and the Heatsink allowing it to dump air directly out of the case), Antec's Nano 7 Diamond TIM (whatever it's called, but I know it's the 7), and Arctic Silver Ceramique applied to the GPU and VRAM/VRM/MOSFET on the card in contact with the heatsink, and the wires hidden as best as is possible behind the MB tray courtesy of a few holes I will be cutting for cable routing but with IDE drives and a non-Modular top-mounted PSU, it's definitely not going to be pretty.

Anyway I apologize for the very long post, I have been stuck in a Dr's waiting room so I am rather bored.

I do very much appreciate the time and effort you put into doing these tests, and the information is great. Thank you!

"The world breaks everyone and afterward many are strong at the broken places"Ernest Hemingway
sniper.gifS.T.A.L.K.E.R FAN CLUB aaskull.gif

~*~ THE TAO OF BACKUP ~*~

PLEASE DOUBLE-CHECK ANY ANSWERS I MAY PROVIDE, AS I AM HUMAN AND QUITE FALLIBLE
Warning: Spoiler! (Click to show)
Samsung Galaxy Nexus - Rooted & Running JBSourcery FINAL & LeanKernel Experimental
nleksan is offline  
post #6 of 8 (permalink) Old 08-10-2012, 05:05 PM - Thread Starter
Retired Reviewer
 
ehume's Avatar
 
Join Date: Dec 2009
Location: New Jersey, USA
Posts: 10,070
Rep: 1372 (Unique: 855)

Whew! Do post a link to your build log when you do this, showing old and new. It will be very interesting.

 

I don't know if you have used a nibbler before, but it makes many things easier (the link has instructions on how to use the tool).

 

The testing was done on an open motherboard -- no case in sight.

ehume is offline  
post #7 of 8 (permalink) Old 08-10-2012, 05:37 PM
New to Overclock.net
 
Join Date: Jan 2011
Location: a
Posts: 1,129
Rep: 58 (Unique: 53)
Hmmm, so you get better gains from/push pull at low RPM fans than you do from high RPM fans. That was similar to what martin210 got.

Good work thumb.gif

a





















































































































a
mam72 is offline  
post #8 of 8 (permalink) Old 08-10-2012, 06:10 PM
New to Overclock.net
 
nleksan's Avatar
 
Join Date: Dec 2011
Location: Ohio
Posts: 5,505
Rep: 660 (Unique: 521)
Quote:
Originally Posted by ehume View Post

Whew! Do post a link to your build log when you do this, showing old and new. It will be very interesting.

Will do! I agree, it will be quite interesting to see the absolute best of everything from 2004 vs (very close to) 2012!

"The world breaks everyone and afterward many are strong at the broken places"Ernest Hemingway
sniper.gifS.T.A.L.K.E.R FAN CLUB aaskull.gif

~*~ THE TAO OF BACKUP ~*~

PLEASE DOUBLE-CHECK ANY ANSWERS I MAY PROVIDE, AS I AM HUMAN AND QUITE FALLIBLE
Warning: Spoiler! (Click to show)
Samsung Galaxy Nexus - Rooted & Running JBSourcery FINAL & LeanKernel Experimental
nleksan is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off