TR Forums

ok i know i've been on here a while, and have got a fair share of posts, but, well this is probably a really stupid question

but anyways i was reading something that got me thinking that maybe the fan is suppose to pull air off the heatsink not blow air over it, i don't know for sure the correct way, but it seem to me the correct way would be to blow air over it, would appriciate an answer

i really feel stupid for asking this question

thanks

ps please don't be to harsh on me

Those things blow a lot harder than they suck, so blowing over moves more heat off of the HS. If you put your hand right next to the HS when the fan is blowing over it, you'll feel a lot of air coming off of it, and with the air goes the heat the HS took off the CPU.


Hey, there are no stupid questions, only stupid answers;-)

Hi. Well, on the HSF itself it's always best to have the fan blowing down/across onto it as this moves ambiant air around the heatsink to take the heat energy away from what has built up on the heatsink itself. Also, "exhaust" fans are a good thing in the computer case, as this removes warm used up air in the case, and it is also useful to have suck fans positioned in a sensible part of the case (e.g not right next to the exhaust fans!) to draw in unused cooler air that can be passed over the HSF unit rather than circulating the same old warm air over the HSF, as this will simply hinder than energy transfer process between the heatsink and the surrounding air. Hope that lot made sense. Good Luck.

Generally its defitenly best to have it blowing, unless the heatsink is designed otherwise.. Alpha PAL8045 anyone....

cheers

Alpha recommends that you install your fan blowing up, pulling air off of their PAL series of heatsinks. Most independent testing verifies that this does produce lower CPU temperatures than blowing down onto the PAL's pins.
http://www.micforg.co.jp/faqe.html

Alpha NovaTech wrote:

Which type of fan heat sink is better, "expulsion type" or "induction type"?
According to our tests, the induction type showed better results than the expulsion type. This is because we used an intake cover to draw air through the lower part of the heat sink for more heat dissipation from the lower part of the fins. However, if some other heat source is located near the heat sink, heated air would come into the heat sink and affect the thermal performance, so the choice must be made depending on the environment.

For many other heatsink designs, it is better to have the fan blowing down into the heatsink.

Think about it...
All of the air close to the fan moves goes through the blades; the same volume is moved on both sides of the fan! The difference you are feeling by hand is the turbulence that the fan blades cause as the air moves through. Yes turbulent air has more cooling ability because more molecules will have contact with the surface than if the air was in laminar flow (like on the intake side of the fan). The problem is that the tightly spaced fins on the heat sink prevent turbulence, and funnels the air in laminar flow again. Turbulence does not become a cooling factor any more on these heat sinks.

In fact in the center on stock heat sinks, and designs antything like them, there will be an area where hot air will become trapped if the fan is blowing into it, and act like an insulator rather than taking hot air away. This is also where you want the coolest air closest to the processor. Turn the fan around so it is sucking, and cool air coming in from the sides can reach the center easily. This has the added benefit of not blowing hot air over capacitors (which shortens thier lives quite a bit) and over nearby controllers. If hot air is spit from the top of the heat sink, an exhaust fan can then lift it outside of the case before it gets a chance to heat anything else!

Oh, and it entirely depends on the design, but from what I can tell from many I have seen, exhaust is best.

-LS (JustAnotherEngineer)

Test and find out. No one can tell you for certain, as their conditions will always be different, and that can make a crucial difference.

Just remember that temperature of the CPU isn't everything, surrounding components are also effected. Put a mobile probe on the sides of some of the larger capacitors near the CPU and surrounding chipsets to see the REAL differences and run your tests. 1C of CPU saving isn't worth it when the temperature of capacitors is raised 5C. You will just need to use some common sense when choosing.

-LS

well hmm, i think ya it was something about the alpha i was reading that got me thinking but ya the amout of air moved should be the same i would suppose, say i made it though a year and a 1/2 of enginnering for i gave up (as i am quite lazy it seems) and now am 2nd ed major, but you all think that it wouldn't be a drastict increase if any increase, nothing to kill the processor anyway i'm guessing, and might help the other conponets i guess i see the logic in that also

for what it's worth it a v7 (no resistor took it off) (thinking of geting a slk 800 in a month or so) full speed anatec 1080 case
two exhaust fans full speed one side fan and two front fans (blowing air in, as i can't think of the proper word for this ) hooked up to vairable speed ps

thanks everyone for all the replys as i didn't think there would be nearly this much comment on such a question

I agree with Forge that sometimes you have to try both ways and see what is better. For me I used a Globalwin 60 MM 6800 RPM fan pointed towards a CAK38 heatsink and got decent cooling. I used a OCZ Goliath SE with a OCZ 80 MM 3300 RPM fan and got a little better performance with the fan pulling air off the heatsink.