Stupid chip cooling question...

[Chunky Egg Foo]

I can't to seem to find this information on the sites I've googled so maybe it's just plain common sense. But on AMD64 heatsinks, is the fan on top blowing air away or towards the heatsink? I've placed my palm above the fan but don't feel any real source of air flowing but on the sides, i feel a little breeze. Either my hand is on the dead zone of the fan or that the air is being reflected from the heatsink as it blows in. The reason for this question is I want to put a 120mm fan on my case and it just happens to be above the heatsink fan. I don't want to negate the affect of the heatsink fan.

 

[frag85]

hopefully noone will flame this, but CPU HSF blow inward.

 

[kronchev]

stock heatsinks (stock as in "not modified") usually blow inward. thats actually worse cooling than sucking off of the processor. if you can see the side of the fan, it might have an arrow indicating the direction of flow. best guess is it blows onto the sink.

 

[Chunky Egg Foo]

So if it blows into the sink, what would happen if I put a 120mm fan near it that blows outward? Negation?

 

[zer0signal667]

Originally posted by kronchev
stock heatsinks (stock as in "not modified") usually blow inward. thats actually worse cooling than sucking off of the processor. if you can see the side of the fan, it might have an arrow indicating the direction of flow. best guess is it blows onto the sink.

This isn't an accurate statement, you can't verify that it's true in all situations. There's been evidence that says each way works best, it all depends on the heatsink/fan combo.

 

[jagec]

Originally posted by zer0signal667
This isn't an accurate statement, you can't verify that it's true in all situations. There's been evidence that says each way works best, it all depends on the heatsink/fan combo.

indeed. Anyway, the difference is only a degree or so, so...

Most fans blow towards the label side (where the sticker is on the center of the fan).

 

[frag85]

IMHO put the outer fan to blow into the case, if its a 120MM, then u will most likely also cool anything in its path (North bridge). i have an 80 mm w/ 7V mod blowing right onto my vid card/NB area and it cools it off pretty well. what kind of case, it must look pretty sick w/ a 120mm fan on the side.

my side case fan blows right over this area.

since then the "ghetto" setup w/ the fan on the vid card has come off. although it did allow another 5% from its max OC to happen. that worked better than the giant 2 i had on there.

edit: since i built this rig(8 months ago) i have been playing around w/ fan setup when i'm bored and not working at a job or on my truck, with the idea of optimizing air flow, it was pretty useless. The only real difference i ever made was haveing the side case fan blow inward at the vidcard/NB chipset.

 

[Ice Czar]

Originally posted by zer0signal667
This isn't an accurate statement, you can't verify that it's true in all situations. There's been evidence that says each way works best, it all depends on the heatsink/fan combo.

exactly, some suck others blow
another primary concern is the airflow (and its temperature) within the case (especially if its sucking air off the mobo)
we always tried it both ways on any given heatsink to see which was better (with a stable ambient and load)

 

[rfw2003]

well on my XP with a SLK-900A I have it pulling air from the HS, I noticed a 3c differeance in temps with the fan pulling the air from the HS than with it blowing the air into it.

R.F.

 

[mayakindaguy]

For my personal rig the setup works best as follows.

Cpu fan blowing at the cpu. Two front fans blowing fresh air into case, bottom rear fan blowing fresh air into case, the one just above sucking air out, chasis fan blowing air in and then there is the chasis fan blowing air in as well. It's kind of weird, but I've benched and tested with all other configurations, and this setup gives me 3 degrees lower cpu temps and higher synthetic benchmark scores. too bad I didn't save screenies of each setup. Don't know if anyone could explain this since I traditionally thought that cool air should blow in from the front bottom and hot air rises and gets sucked out at top. However, the idea that the most important components like the cpu not getting a lot of fresh air seems counter productive. I've seen diagrams of hp workstations and i believe a diagram from intel on airflow and they seem to suggest filtering in from bottom front to top and filtering out on the upper rear of case is optimal, but this is not what I've found to be the case. Also, you would think pumping equal amounts of air in and out of the case would be best, but I actually have more fans pumping air in rather that out. However, it seems that the outgoing fans are actually stronger. Unfortunately, I don't have a sucking cpu fan, so that is the one factor that I have not tested. Well, just my .02 1/2 cents.

 

[emorphien]

I'm going to try it both ways on the SP-94 I'm getting soon.

 

[Chunky Egg Foo]

My case is a micro atx from Athenatech, so space is really cramped. Even the DVD drive touches the heatsink fan (barely). The 80mm intake fan in the front is useless because all the y-cables sit in front of it. Let me try inverting the fan after I molest my side door with a dremel.

 

[titan151]

Don't feel bad I used to assume that all heat sink fans blew out. So I had been using the wrong direction for the past few years. That said, they still work in the opposite direction, perhaps not as well.

 

[kronchev]

Originally posted by zer0signal667
This isn't an accurate statement, you can't verify that it's true in all situations. There's been evidence that says each way works best, it all depends on the heatsink/fan combo.

it cools the heatsink best, period. if your case air flow doesnt work with that, obviously itll perform worse. my statement is still correct however.

 

[hardwired]

Originally posted by zer0signal667
This isn't an accurate statement, you can't verify that it's true in all situations. There's been evidence that says each way works best, it all depends on the heatsink/fan combo.

this was a really big deal back in like 2000 when the serious heatsinks started coming out for the amd tbirds. it took a lot of us many hours of testing to reach this conclusion.

therefore, its the freakin gospel truth. try it both ways.

 

[zer0signal667]

Originally posted by kronchev
it cools the heatsink best, period. if your case air flow doesnt work with that, obviously itll perform worse. my statement is still correct however.

It has nothing to do with case airflow, people have switched the direction keeping all other factors constant and gotten various results. It depends on the heatsink and fan combination. How can you say your statement is correct, have you tried it with all possible heatsinks and fans? How about with even just all heatsinks and the same fan? You can't make a statement like that with no scientific data to back it up, you don't even have theory to back it up.

 

[kronchev]

Originally posted by zer0signal667
It has nothing to do with case airflow, people have switched the direction keeping all other factors constant and gotten various results. It depends on the heatsink and fan combination. How can you say your statement is correct, have you tried it with all possible heatsinks and fans? How about with even just all heatsinks and the same fan? You can't make a statement like that with no scientific data to back it up, you don't even have theory to back it up.

ok lets see

fan sucking off = big blind spot over center of CPU

fan blowing on = no blind spot

I WONDER

thats my "theory" and if you can find some hole in it feel free to let me know

 

[hardwired]

Originally posted by kronchev
thats my "theory" and if you can find some hole in it feel free to let me know

sure.

YS Tech TMD fans and crossflow fans.

edit: they dont have the center spot, eh?

 

[kronchev]

Originally posted by hardwired
sure.

YS Tech TMD fans and crossflow fans.

edit: they dont have the center spot, eh?

they dont, last i checked only one heatsink could even mount a TMD fan, and in fact theres only one kind, in 70mm, the dumbest design ever. lets make it so it doesnt work with what everyone already has

 

[hardwired]

swiftech non-V heatsinks can mount a 70mm. so can the entire thermalright line since the SK7.

i'm sorry to be the fly in your ointment. are we still pals?

 

[frag85]

i keep mine blowing in b/c thats the stock direction on stock cooling, and my cpu doesn't heat up that much even at the crazy FSB speed i run it at, i'd go higher but my ram didn't eat their wheaties this morning.

 

[zer0signal667]

Originally posted by kronchev
ok lets see

fan sucking off = big blind spot over center of CPU

fan blowing on = no blind spot

I WONDER

thats my "theory" and if you can find some hole in it feel free to let me know

Have you extensively analyzed the airflow charactersitics of every possible system? Is it not conceivable that when blowing air onto a fan that there is indeed a dead spot in the center? All fans do have some kind of hub assembly in the center, some of them rather large in fact. Couldn't that create a dead spot in the center? If air is flowing through the middle, which direction is it going, since the fan/heatsink model is basically axisymmetric?

You're free to have your own theories, but you seem to have a problem accepting the fact that something you say might not be the truth. And I have a big problem with saying that ANYthing is absolute, especially when there is little to no evidence.

 

[jagec]

Originally posted by kronchev
ok lets see

fan sucking off = big blind spot over center of CPU

fan blowing on = no blind spot

umm...there IS a "blind spot" either way...the fact of the matter is that if the air is entering straight down and leaving in multiple directions, there will be a "dead zone" with little to no airflow.

Once again, it really, really depends on the heatsink design, some heatsinks work better blowing, some sucking.

 

[frag85]

Originally posted by zer0signal667
Have you extensively analyzed the airflow charactersitics of every possible system? Is it not conceivable that when blowing air onto a fan that there is indeed a dead spot in the center? All fans do have some kind of hub assembly in the center, some of them rather large in fact. Couldn't that create a dead spot in the center? If air is flowing through the middle, which direction is it going, since the fan/heatsink model is basically axisymmetric?

You're free to have your own theories, but you seem to have a problem accepting the fact that something you say might not be the truth. And I have a big problem with saying that ANYthing is absolute, especially when there is little to no evidence.

no matter what there is a dead spot with any type of a fan.

the logic to blowing in is that it forces air all the way down. when it sucks air, it brings cool air(ambient case temp) into the heatsink from the outside in. there are many theories of which i will not get into(difficult to understand and explain without pics and examples, i bearly understand them), but basicaly bowing keeps the inner cooler and outer wamer(from warm air coming outwards), and sucking keeps the outer cooler and inner warmer(from cool air warming up as it comes inward) (inner not meaning center, with how close the fan is mounted, there will always be a dead spot int he center of the heatsink, they both average out to be the same temp more or less in the long run. if you want to experiment see what works best for you. it doesn't really matter, if the fan is directly on the heatsink, then it won't really matter, but if it is spaced off, w/ no duct b/t them, it should blow in, sucking will just suck to much air from around and not in/through the heatsink.

 

[Ice Czar]

Originally posted by frag85
no matter what there is a dead spot with any type of a fan.

there is more than one way to skin a cat
that only applies to axials

the correct qualified statement would be, with an axial fan there will always be a deadspot withing X distance from the hub

as you mentioned after you exceed that distance, yet still contain the airflow in a duct
the blind spot will be gone, that is a trick unit that does the job within a short distance due to the radius of the duct, but a straight duct would do it as well given enough distance

a good read Particulary the Intake or Exhaust, laminar or turbulant

the over riding factors that bear on the suck vs blow debate is the ambient temperature of the airflow, which is often in a "case" variable from one config to another would the fan be picking up preheated air or not? is the air cooler near the mobo or above the fan?

General Heat Transfer Guide @ amdmb

heat transfer for conduction would be q=k/L(T1-T2)

We can draw some interesting conclusions from this equation. First, heat transfer is proportional to the temperature difference on the object. If the temperature differential doubles, the heat transferred doubles. Second, the conduction coefficient "k" is proportional to heat transfer. If the conduction coefficient doubles, the heat transfer doubles. Alternatively, for the same differential temperature, twice as much heat may be transferred. The final observation is "L". As thickness increases, heat transfer decreases. Alternatively, to maintain the same heat transfer through a material twice as thick requires twice the temperature differential.

there is a corollary to that for convection
and its found in Newton's Law of Cooling


the secondary factor in the efficency of the suck vs blow debate would be the total airflow (figuring out if there is a pressure change with one vs the other however is beyond me )

and then the design of the heatsink, which could have a design optimized for one versus the other

Ideally, you would want to duct fresh cool air into the case and directly to the fan, and then have that fan far enough away from the heatsink to eliminate any dead spot, most (but not all) heatsinks are designed for blow