Sucking or Blowing
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- Posts: 346
- Joined: Tue Dec 12, 2006 5:09 pm
- Location: U.S.A. Massuchusetts...... *Folding For SPCR*
- Contact:
Sucking or Blowing
in a tipical case with no ducting is it better to have the HSF blowing on the heatsink or sucking off the heatsink?....IMO i dont think it matters because the air souce is gonna be the same temp weather its being pushed though or pulled though the heatsink but thats IMO... and does any one think differntly?
IMO blowing is probably very, very slightly better because the airflow coming out of the fan is turbulent and this breaks up the boundary layer which forms around the heatsink fins and impedes heat transfer.
One thing to look out for that affects both methods is the "airflow dead spot" caused by the fan hub; on some heatsinks this can cause the centre of the heatsink, which is often directly over the CPU core, to receive no airflow at all. This can be prevented by locating the fan a couple of inches away from the heatsink with a short duct.
One thing to look out for that affects both methods is the "airflow dead spot" caused by the fan hub; on some heatsinks this can cause the centre of the heatsink, which is often directly over the CPU core, to receive no airflow at all. This can be prevented by locating the fan a couple of inches away from the heatsink with a short duct.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
Re: Sucking or Blowing
It would not matter IMHO if the fan were symmetrical. It isn't. Go grab any 25mm-thick fan; I'll wait...PopCorn wrote:in a tipical case with no ducting is it better to have the HSF blowing on the heatsink or sucking off the heatsink?....IMO i dont think it matters because the air souce is gonna be the same temp weather its being pushed though or pulled though the heatsink but thats IMO... and does any one think differntly?
OK. Now take a close look at how closely the blades come to the edge of the case on the intake side. Turn it over; the blades don't come anywhere near as close to the back edge of the case. So, on a fan mounted to a HS, the closeness of the blades depends on whether the fan is blowing or sucking. Madshrimps, for example, often reports getting better cooling results in "suck mode". YMMV.
-
- Moderator
- Posts: 7681
- Joined: Mon Dec 09, 2002 7:11 pm
- Location: Maynard, MA, Eaarth
- Contact:
Hello,
I think it depends on the HS -- some definitely work better is "suck" mode, like the Alpha, and some of the Thermalright models.
You see, pushing air requires velocity, and things get slowed rather quickly due to friction. But, pulling air causes a lower pressure, and since air "abhors" a vacuum, the nearby air moves in to equalize the pressure, and then that pull the air from farther away, and so on. It causes a "chain reaction" that can move air much farther away from the fan than blowing can.
The nature of air makes it easier to pull, than it does to push.
I think it depends on the HS -- some definitely work better is "suck" mode, like the Alpha, and some of the Thermalright models.
You see, pushing air requires velocity, and things get slowed rather quickly due to friction. But, pulling air causes a lower pressure, and since air "abhors" a vacuum, the nearby air moves in to equalize the pressure, and then that pull the air from farther away, and so on. It causes a "chain reaction" that can move air much farther away from the fan than blowing can.
The nature of air makes it easier to pull, than it does to push.
some time ago i had scythe kamakaze rev. b on my athlon xp - together with a custom hole in the side panel of the case almost directly over the hsf it worked quite well. just out of curiosity i once flipped the fan (80 mm) - temps were higher 2-3 degrees. i guess with a duct to the side the whole contraption would work even better. the heatsink was all copper, quite tall, had a little shroud, and had pins (almost 500) instead of fins.
-
- Posts: 89
- Joined: Tue Sep 07, 2004 8:24 am
- Location: MA, USA
Blowing would blow air directly at the surrounding components which need cooling.
In a sucking configuration, the incoming airflow will be from all directions (and hence slower), rather than the down/out forced flow. It'll take the path of least resistance, which would avoid the mobo components.
... I think.
In a sucking configuration, the incoming airflow will be from all directions (and hence slower), rather than the down/out forced flow. It'll take the path of least resistance, which would avoid the mobo components.
... I think.
In my particular example, the Papst fan on the XP-120 in suck mode results in uniform temperatures for CPU, voltage regulators (which run hot on this MB) and northbridge. All peak at ~47C under sustained load. The CPU and Vregs are measured with MB sensors, the NB with an infrared thermometer.
If the Papst is flipped into blow mode, the CPU temp doesn't change, but the Vregs increase by 7C while the northbridge drops by 7C.
I use suck mode, although the temps would be acceptable either way.
If the Papst is flipped into blow mode, the CPU temp doesn't change, but the Vregs increase by 7C while the northbridge drops by 7C.
I use suck mode, although the temps would be acceptable either way.