Heatsink Fan Blowing "Onto" or "Away"???

[JokerJKNY]

Hey all,

i finally got some new ThermalRight SK6+ heatsinks for my dual Athlon machine, but i was wondering which way my Thermaltake 4500rpm fans should go?

right now, i have it blowing away, exhausting the air from the sink.

thusly, i'm running my dual 2100+ Athlon @ 133 mhz bus, riding at a somewhat cool 55 degree C.

should i flip it around? which way works best for you guys? just trying to get a concensus.

 

[markjs]

It's an age old debate, but I go with the manufacturers....when they build the sinks they make em blow down onto the sink.

 

[Peter D]

id say onto the heatsink as well, and just get some exhaust fans in your case

 

[FluxCap]

I did a test with both on my slk800 with YS Tech variable on full speed.

Into: 39c avg idle temp
Away: 41.5c avg idle temp

Enough of a difference to help me make up my mind.

 

[oldfart]

I'll go against the grain. I like away. If you go towards the CPU, all the CPU heat is forced out of the bottom of the HS. That hot air will heat up the components near the CPU like MOSFETS, inductors, the NB and ram. You typically have a rear case fan that will also pull the hot air from the CPU (when blowing away from the CPU) which will get that heat out of the case. Maybe the CPU might be a bit cooler when going towards, but mobo components will be hotter.

 

[FluxCap]

Good point.

 

[amcdonald]

Oldfart, that isn't necessarily true about fan air heating/cooling the mboard components.
When air is pulled through the heatsink, there isn't any air forced on the motherboard.
Especially on an slk-800/900, where there is plenty of open fan area, most of the air entering isn't going to pass over the motherboard.
But if it is blowing into the heatsink, it will force plenty of air on the motherboard.
While this is air is going to be slightly hotter than the other case air, the cooling will be better simply because there is so much more air passing by.
As long as the air is cooler than the component it will help.

In my case, where I have a cpu duct, it ONLY makes sense to blow air into the heatsink.

 

[GooGooCluster]

this would be a good agurment for anandtech to settle. Do some test with sensors next to the mb componets and such. and try it with like 5-6 differnt heatsinks to see if the results will very. And also to prove if there is harm being done.

 

[DAPUNISHER]

It's my understanding that it's dependant upon the heatsink's design and some do perform better sucking instead of blowing or atleast use to. There were a couple Alphas that were recommended by the manufacturer to suck and that their testing showed better cooling that way. oldfart's point is valid depending on the orientation of socket, wether there's active or passive NB cooling, and design of the heatsink too IMO, since I can place my hand on the side of my Aeroflow closest to the NB and it's blowing warm air right at the NB which is also actively cooled causing the air from both to collide albeit the AeroFlow's stronger.

In my case, where I have a cpu duct, it ONLY makes sense to blow air into the heatsink.

I'm not trying to be argumentive here just inquisitive concerning your logic and methodology on this comment but perhaps you could explain why Dell has their CPU ducts sucking air away from the CPU and out of the case then if putting the air on the CPU is the "ONLY" sensible method when using a CPU duct?

 

[FluxCap]

This is true of my slk800 Dapunisher. It has been proven ( I am almost certain) that the fin design on the slk800 prefers air being blown into the heatsink AND performs better with higher RPM fans as well.

 

[DAPUNISHER]

Originally posted by: FluxCap
This is true of my slk800 Dapunisher. It has been proven ( I am almost certain) that the fin design on the slk800 prefers air being blown into the heatsink AND performs better with higher RPM fans as well.

I had the SLK800 too (wish I'd kept it now ) and I can't imagine it working any better ssucking instead of blowing with the way that badboy cooled for me with a 50cfm fan hitting it :beer:

 

[GooGooCluster]

Originally posted by: DAPUNISHER

Originally posted by: FluxCap
This is true of my slk800 Dapunisher. It has been proven ( I am almost certain) that the fin design on the slk800 prefers air being blown into the heatsink AND performs better with higher RPM fans as well.

I had the SLK800 too (wish I'd kept it now ) and I can't imagine it working any better ssucking instead of blowing with the way that badboy cooled for me with a 50cfm fan hitting it :beer:

why did you get rid of it? Was it not performing well?

 

[DAPUNISHER]

I wasn't using AMD in my personal systems for quite sometime there and it was just lying about so I sold it to a friend who wanted it for his overclocked 1600+ AGOIA. Th danged procrastinator hadn't installed it last time ask and was still using a his old SK-6 so I may try to buy it back if he still hasn't used it

 

[Canterwood]

The new hyperthreading P4 machines I build at work, now have a tube which attaches to the inside of the case (via a grill) and locates directly over the cpu fan. This allows the cpu to suck cooler outside air directly onto the heatsink, instead of using warmer air from inside the case. It works pretty well.
The tube comes only with the new case though and not the with cpu itself.

 

[amcdonald]

Dapunisher, Dell doesn't use their ducting system for the same reason I do.
Theirs is an attempt at effeciency & low noise, using only 1 case fan. While its a good solution, Dell's design is targeted at sufficient cooling, not optimal.

The main differences are the architecture of the heatsinks used (slk900) and the amount of casefans I use (5 more fans than dell's computer).
Since I have plenty of places to exhaust air, the duct just works best as an intake. I have actually tried switching the entire cooling system around, thinking I might get better case temps, but there was no difference in case temp, and a 7ºC rise in cpu temp. An alpha perhaps could suffer less of a temp increase with a cpu exhaust, but reading from other people's experiences they still work best with an intake duct.

I would like to see some temps reported from a dell user on a high-end p4 that might rival the wattage of my 1.75volt 2100+
Recently I've been fine tuning the airflow and asus probe is now reading 32ºC/25ºC in a 23ºC environment. (under SOB load)
Keep in mind I can barely hear any noise coming from my case, and it is definitely quieter than a dell with a ducting system.

 

[DAPUNISHER]

Thanx for qualifying your remarks It's evident your previous statment just lost a lot by being so brief and emphasising ONLY. I'm certain your system is superior to the Dell in cooling though I owned a Dell 4550 P4 2.4B hot deal until I sold it for a profit recently and it was near dead silent so your statment of being quieter is something I'll just have to take your word on

 

[amcdonald]

I've seen a few in operation, and I'm really impressed with how quiet the newer dell's are..
Mine is quieter, but not by much. Little things like hard drive noise, etc...
The room I keep my computer in has very little ambient sound, so I have gotten obsessive compulsive with the computer's noise.
But I know I'm not too obsessed about noise when I read posts like KenAF's... where he spends $80 on foam.

 

[Jeff7181]

I'd like to quote someone I saw post on this topic a long time ago...

"When you're cooling off your food do you blow air at it, or suck air away from it?"

In addition to that, CPU fans are designed to blow, not suck... and yes, I understand they have to suck in order to blow... but I think it's a safe bet that they can create a larger positive pressure differential than they could create a negative pressure differential.

Didn't anandtech do an experiment with this a long time ago? I'm gonna go digging and see if I can find it... if I remember correctly, they determined that sucking was only good in tight spaces, like 1U server enclosures.

Oldfart... that's a logical thought... however, my CPU fan hangs over the edge of my heatsink and blows onto the northbridge heatsink as well... and I like the fact that it does that... so maybe just like every other thing up for debate on here, it depends on the application =)


*EDIT* That didn't take long... here's the review I found...
linkage

 

[DAPUNISHER]

From the review

The Alpha PAL153U's performance is very good, but depends very much on the fan orientation. In a 1U environment, it is better to let the fan exhaust air from the heatsink; in other configurations where there is more space above the cooler (e.g. 2U+ servers), the fan should blow towards the sink. In any case, it is a good idea to compare temperatures in both configurations after installation, since it depends on many variables (such as ventilation concept of the case, location of air intakes) which configuration is better.

It is not always better for it to blow than suck and the food analogy has no relevance to CPU coolers

 

[Jeff7181]

That review was done using servers with limited space for airflow... even a small Mini-Tower has more airflow than those...

 

[Pulsar]

>>I'd like to quote someone I saw post on this topic a long time ago...

>>"When you're cooling off your food do you blow air at it, or suck air away from it?"

This is a fallacy. You blow air on it because it's too hot to stick your lips on it and suck. Plus, sucking at your food is not exactly considered good table manners (slirping your soup...).

Anyway, I'll say what I've said before.

It's not rocket science. Thermal Transfer is MOST efficient when the temperature differential is greatest. If you blow into the middle, you have cool air hitting the center "hottest" area.

Sucking passes the air across the warm outer sections and has a much lower temperature differential - by the time the air reaches the center it has warmed some already.

Finally, there is the fact that air has momentum. Blown straight down the air will get much further into the fins where air flow is restricted. If you "suck" the majority of the air will have a tendency to travel more across the upper fins closer to the fan.

Yes yes, there are SOME heatsinks that may claim to have a unique design that requires sucking. But if that is the case you will likely discover that another "simple" heatsink like the sk's that simply have straight fin design will perform far better. 99% of all heatsinks are better having the fan blow into them.

 

[DAPUNISHER]

From Alpha's questions&answers section on air pushing pulling Text

Alphas are excellent coolers and the advice about trying it both ways is sound me thinks. That sencond response to the feedback was a guy saying he had 5c better cooling from sucking than blowing on his rig.

 

[Anonemous]

for slk800's i read it was better blowing toward the cpu

FAQ from Thermalright

 

[DAPUNISHER]

Originally posted by: Anonemous
for slk800's i read it was better blowing toward the cpu

FAQ from Thermalright

No doubt about it. There are good coolers though that can benefit from pulling instead of pushing depending on enviroment as the Alpha FAQs state and some Alpha users have experienced so I just wanted it to be clear that anything said to the contrary isn't always the case and users of Alphas inparticular should try both ways as suggested and see for themselves.

 

[Jeff7181]

>>I'd like to quote someone I saw post on this topic a long time ago...

>>"When you're cooling off your food do you blow air at it, or suck air away from it?"

This is a fallacy. You blow air on it because it's too hot to stick your lips on it and suck. Plus, sucking at your food is not exactly considered good table manners (slirping your soup...).

Exactly... too hot to put your lips on... so you blow air on it, you don't suck the hot air off it to cool it down.