I can't understand why people stick to blowing CPU fans!!!

Pirata · Post by **Pirata** » Thu May 20, 2004 7:10 am

Why the hell would anybody use a HS and a blowing fan on the CPU to spread really hot air all over the case?

Why doesn't everybody use heatsinks like the alpha, sucking hot air to the outside through a duct? What's the point on using blow heatsinks? I can't understand it!

Besides, does anybody know any heatsink can allows doing away with the CPU fan at all? I mean for a CPU, like mine, a Pentium4 Northwood 2.6GHz/800/512/HT.

sthayashi · Post by **sthayashi** » Thu May 20, 2004 7:30 am

Pirata wrote:Why the hell would anybody use a HS and a blowing fan on the CPU to spread really hot air all over the case?

Why doesn't everybody use heatsinks like the alpha, sucking hot air to the outside through a duct? What's the point on using blow heatsinks? I can't understand it!

Probably because you can use less copper & aluminum if the fan is blowing downwards. When it blows downwards, the air comes in direct contact with both the fins and the baseplate itself. When the fan blows upwards, airflow does not come in as much direct contact with the baseplate. And I think that's less of a problem with heatsinks that use pins instead of fins.

Even then, with the exception of Swiftech, all the upward blowing heatsinks that I know of are really tall, which increases the likelyhood of damage to the motherboard due to weight.

As for a passive heatsink for you, have you looked at the Scythe Heatlane Zen?

fractal · Post by **fractal** » Thu May 20, 2004 10:10 am

Pirata wrote:Why the hell would anybody use a HS and a blowing fan on the CPU to spread really hot air all over the case?

Why doesn't everybody use heatsinks like the alpha, sucking hot air to the outside through a duct? What's the point on using blow heatsinks? I can't understand it!

A picture says a thousand words.

How much air do YOU think is getting through those fins

SpyderCat · Post by **SpyderCat** » Thu May 20, 2004 10:19 am

fractal wrote:How much air do YOU think is getting through those fins

The same layer of dust/cobwebs would have been on top of the fins, shielded from sight by the fan, had the the fan been blowing onto the fins.

I assume you saw the pictures of the dirty Zalman heatsink a few days ago?

This isn't a valid arguement....

Putz · Post by **Putz** » Thu May 20, 2004 10:40 am

In fact, quite the opposite, fractal. It's easy to clean out the dust in that picture by blowing through the heatsink with a can of compressed air, or using a simple vacuum cleaner. On the other hand, to properly clean a layer of dust from between the fan and the top of the heatsink (as would occur if the fan were blowing down), removal of the fan would be necessary.

bomba · Post by **bomba** » Thu May 20, 2004 10:56 am

I suspect that the answer is that high-quality "blow" type HSF's work well enough for a near-silent PC, are priced reasonably and are easy to install.

On the other hand, I've always wondered whether an undervolted P4 2.4C single hard drive system could be entirely cooled with only two 5-7v 80mm L1A's. This system would have two seperate air channels. The main air channel would route intake air over the hard drive, suspended low in the front of the case, perhaps with a Zalman heatpipe hard drive cooler, or homemade heatsinks to aid cooling with minimal airflow. Extraction would be via one of the recent heatpipe "tower" style CPU coolers, configured fanless in suck mode. Suction would be via ducting to a rear case mounted, undervolted 80mm L1A exhaust fan. The PSU would also use an 80mm L1A, but be fed via the 2nd independent air channel (PSU fresh air duct). Graphics would ideally be a low-power, passively cooled solution such as a Radeon 9600 non-pro.

But, bottom line is that very quiet PCs are achievable with careful selection & integration of off the shelf components. Building a custom CPU fan duct is not trivial and at this time, not for me. However, I love studying the expertly crafted ducted CPU solutions created by folks like SpyderCat and Bluefront.

Post by **Rusty075** » Thu May 20, 2004 11:12 am

Ok, to stop this from side-tracking into yet another dust debate:

The reason that 99% of heatsinks blow towards the CPU is because 99% of heatsinks perform better that way. Generally only certain pin heatsinks perform better in "suck" mode. (alpha's usually do, swiftech's usually don't) Nearly every fin heatsink performs better with the fan blowing towards to the CPU.

The reason that side ducts aren't common is simple: because mobo makers don't always put the CPU socket in the same place. Get a new mobo, and you're likely to need a new duct as well.

Here's an experiment to try Pirata: reverse your CPU fan and have it suck through the duct, rather than blow outwards. The CPU is the most temp sensitive component you have, why have it cooled by air that's been warmed flowing through the rest of the case. Instead, use the duct to feed it the coolest air possible: the air outside the case.

And Pirata, whatever you did to triple post this thread....please don't do it again.

1398342003 · Post by **1398342003** » Thu May 20, 2004 11:15 am

With the fan blowing down you get the "countercurrent" effect that is shown on the SPCR front page. The real problem is the fan's dead zone.

If used properly blowing down or up can be quiet and cool. I use the down method, and I get 25°-30° above ambient with about 10-20 CFM.

vortex222 · Post by **vortex222** » Thu May 20, 2004 12:41 pm

my SLK800 proformed equil both ways i figure, it might be 1 degree worse sucking, but i found it localised the heatouput to an area that allowed the rear casefan to exaust it before the psu did.

i then took it a step further and put in a duct and it works great. but i also have a reasonably cool running computer compaired to some other powerboxes. with a hotter cpu in my computer i would probibly have to do somthing different.

mheffel · Post by **mheffel** » Thu May 20, 2004 12:42 pm

Why do heatsinks work better with fans blowing air on them? Maybe thinking of it this way will help.

When I'm hot, I get a fan and position it so the air blows towards me ... seems to work much better than positioning it where the air blows away from me. But why is this?

I think when the fan is blowing air right at me that most of the high velocity air hits me dead on. The high velocity air is concentrated as a narrow stream aimed right at me. Now, when the fan is blowing away from me, the air moving around me is at a much lower velocity, because the fan is pulling air from a broad area. Most of the air going into the fan doesn't even pass over me, and the further away the fan is the less air movement I get.

Same thing with a heatsink. Aiming a fan to blow air towards the heatsink means most of the high velocity air is hitting the body of the heatsink. Now, if the fan is pulling air from the heatsink, a significant portion of the fan's intake never passes over the heart of the heatsink, only the tips of it's fins ... of course, I think ducting to force the air to be pulled through the heatsink would help significantly, but most heatsinks aren't shaped to allow this.

vortex222 · Post by **vortex222** » Thu May 20, 2004 12:54 pm

yeah thats what i think too. the duct idea has the added bonus of removing the cpu's heat from the case directly though. dont forget, us silent fanatics dont care about a couple of degrees difference, we care about a couple db. and that the cpu is "cool enough" to run stable.

like i said if i had a hotter cpu, i would either have to turn up my duct fan, or have a fan on the hs blowing down to keep it cool. however that is noisyer, and im happy with the speed of my computer.

SpyderCat · Post by **SpyderCat** » Thu May 20, 2004 1:32 pm

The problem is: We, the customers, buy MODULES, and a bunch of modules thrown on a heap make up our PC's.
Most manufacturers make modules. They concentrate on making / marketing one kind of module really good so we buy lots of them. We, the customers are, and want to be, responsible for picking the right modules.

When selecting the modules we want to use most of make the mistake not to "design a system" first. How can you expect to build a good system when you didn't design one first? Once it is designed, you can pick the best modules for your design, and with each specific module you add to you "heap" you have to evaluate the design again, and change the design when needed.

This process is where system-builders like Dell and HP excel. They sell carefully balanced systems. Whole systems.
This is why people like Bluefront and Ralph Hutter succeed over and over again in building good systems. They design first, and all their modules complement each other.

Stop thinking: "what module needs to be swapped?" , start thinking "where can I improve my system as a whole?".

Trip · Post by **Trip** » Thu May 20, 2004 1:33 pm

One reason blowing down is better that I've read in here is cool air hits the top of the fins first and then warms as it approachs the base of the fins so that more of a heat gradient develops on the heatsink. The gradient promotes heat from the warmer base to transfer to the cooler fins and also the cool air hitting the warm fins picks up more heat than if it were already warmed up and hitting warm fins. By the time it reaches the base, it is warmer but the base is much warmer and a lot of heat is still transferred.

The fact that more air comes in contact with the base plate seems to be the most significant reason why blow works better than suck though.

One heatsink that had me really excited for awhile (till I decided it wasn't worth the extra money) is the Aerocool HT-101. The DP-102 is another interesting one but the HT-101 may be more ductable.

Putz · Post by **Putz** » Thu May 20, 2004 1:39 pm

mheffel wrote:Why do heatsinks work better with fans blowing air on them? Maybe thinking of it this way will help.

When I'm hot, I get a fan and position it so the air blows towards me ... seems to work much better than positioning it where the air blows away from me. But why is this?

I think when the fan is blowing air right at me that most of the high velocity air hits me dead on. The high velocity air is concentrated as a narrow stream aimed right at me. Now, when the fan is blowing away from me, the air moving around me is at a much lower velocity, because the fan is pulling air from a broad area. Most of the air going into the fan doesn't even pass over me, and the further away the fan is the less air movement I get.

Same thing with a heatsink. Aiming a fan to blow air towards the heatsink means most of the high velocity air is hitting the body of the heatsink. Now, if the fan is pulling air from the heatsink, a significant portion of the fan's intake never passes over the heart of the heatsink, only the tips of it's fins ... of course, I think ducting to force the air to be pulled through the heatsink would help significantly, but most heatsinks aren't shaped to allow this.

Very well said, mheffel. I agree 100%. (I just happen to be the type to try to make it work in "suck" mode, by ducting, etc.)

And Welcome to SPCR!

Bluefront · Post by **Bluefront** » Thu May 20, 2004 3:06 pm

Humm...I really don't think it is valid to compare how you feel with a fan blowing on your face (rather than sucking air off your face), with how a CPU heatsink responds temp-wise to the same fan conditions. It just doesn't work like that.

I suspect that blowing or sucking would yield similar results, but only if the heatsink were specifically designed for each type of airflow. As it stands right now.....most heatsinks are designed for the "blow" method. So you just cannot reverse the fan and come to a valid conclusion about which method is best.

Pirata · Post by **Pirata** » Thu May 20, 2004 3:13 pm

@Rusty075

My triple posting was a mistake. I got angry with my DSL connection because it was not moving. I started hitting the damned submit button... dumb me...

I cannot turn my CPU fan. It is a 60mm fan boxed with the Pentium. I don't know how to detach it from the HS, and even if I did, the plastic structure that wraps the fan doesn't seem to be able to be reattached the other way round. I also haven't tried to invert polarity to invert spin sense. I did try with another fan I have around and it didn't even started to spin, so I guess polarity inversion does not inverse spin sense, after all... Anybody know something I could do about this?

Duxt must be kept away by keeping case pressure to 0, and using filters in intakes. Sealing cases is too much of a chore to me. Balanced case cooling seems much more elegant.

One question: how much does a duct increase the noise of a CPU fan as perceived from the outside @1m?

Have you ever heared about the ducting mod? Does it remove the fan's dead zone problem? Does it reduce the CPU temp?

I guess a conclusion to all this coulsbe: blowing air needs lower rmps, but the case heats up more. Sucking needs more rpms, the case keeps cooler.
Dust is independent.
Is this right?

Trip · Post by **Trip** » Thu May 20, 2004 3:29 pm

Bluefront (or anyone else), what's the best sucking heatsink you've heard of? the Scythe Kamakaze?

Bluefront · Post by **Bluefront** » Thu May 20, 2004 3:30 pm

Really you cannot come to a simple conclusion like that....there are too many variables.

Heh...I was responding to pirata.

IMHO......ALPHA is the best for "sucking", maybe with it's copy-cat in second place.

You would think this Swiftech heatsink would respond well to "sucking" with the addition of a DIY shroud like the Alpha has. I might try it sometime.

Trip · Post by **Trip** » Thu May 20, 2004 3:41 pm

I meant the copper Synthe Samurai EDIT: actually, that's a completely different design from the kamikaze...

synthe is a copy of alpha?

Bluefront · Post by **Bluefront** » Thu May 20, 2004 3:50 pm

Yeah....Alpha came first with that design (I think anyway). I've had my two Alphas almost three years. The Sythe Kamakaze is a recent copy of an Alpha.

Trip · Post by **Trip** » Thu May 20, 2004 3:52 pm

the samurai looks similar to a thermalright sink, except for the "Wave Stacked Fins". Even the Heatlane Zen looks like it may have originated from TS Heatronics.

vortex222 · Post by **vortex222** » Thu May 20, 2004 3:57 pm

the slks with there wedge design work really well. because it has a larger area to take are in on the bottom, whereass most square hsf wont work so well

Post by **Rusty075** » Thu May 20, 2004 4:03 pm

Bluefront wrote:I suspect that blowing or sucking would yield similar results, but only if the heatsink were specifically designed for each type of airflow. As it stands right now.....most heatsinks are designed for the "blow" method. So you just cannot reverse the fan and come to a valid conclusion about which method is best.

That is the dead-on crux of the whole issue.

In reality it makes little difference: you can achieve the same results either way, depending upon your hardware

The Swifttech "V" heatsink performs surprisingly bad in suck mode. I've played around with attaching Alpha-esque ducts to it, and the results are always worse than just blowing air at it. Not exactly sure why.

It does appear, at least on the surface, at producing a good performing "sucking" heatsink is somehow more difficult. Clearly Alpha has it figured out, but virtually no one else does.

The Scythe Samurai has very little in common with any Thermalright. (unless you consider using copper copying

) And it does perform pretty well in suck mode.

Trip · Post by **Trip** » Thu May 20, 2004 4:10 pm

The Scythe Samurai has very little in common with any Thermalright. (unless you consider using copper copying )

haha, well it does at first (i've only seen pictures). Those Wave Stacked Fins look interesting. Have you compared it to the P4 Alpha?

Post by **Rusty075** » Thu May 20, 2004 4:19 pm

If I had one to compare it to I would.

From my brief playing with it (it's in line for review testing) I'd say that it's probably not quite as good as the new Alpha is, at least according to the few reviews of the Alpha that I've seen. But the Samurai does seem to enjoy low airflow.

Pirata · Post by **Pirata** » Thu May 20, 2004 4:56 pm

@bluefront: why can't I draw such conclusions? It seems logic: for same case temp, with blowing you get countercurrent cooling and the benefit of air molecules actually hitting the HS, whereas with sucking air just runs a little across the HS pins before living for the outside. Positive pressure within the HS is better for cooling, since air really crashes against the heat. With ducting, the difference increases: with sucking, you suck hot air from within the case. With blowing, you take cool air from the outside. You get the CPU cooler, though the case still takes the hot air in.

The best solution, a combination of both worlds, is the countercurrent mod linked in the silentpcreview homepage. It's cheap, yet difficult to build.

Bluefront · Post by **Bluefront** » Thu May 20, 2004 5:58 pm

Ok then.....Explain the Alpha. By your logic it should cool better in the blow mode (no ducting). However the opposite seems to be the case. It is difficult to explain the reason for this, but it's true.

I suspect there's no one answer here....if you come up with one, you can design the perfect heatsink, the perfect air-cooling setup. Everyone seems to be trying.

Trip · Post by **Trip** » Thu May 20, 2004 9:10 pm

I was thinking it had something to do with only having so much of an opening at the bottom that air rushes in and is in such a tight squeeze that heat is transferred well. The greater space in the center relieves pressure thus promoting air flow to there. And it doesn't work well on a thermalright b/c the fins don't allow air to flow as easily around the heatsink. I think some tape on the top half of a thermalright would help out.

One reason why air doesn't flow down as easily is it is drawn to the center? or b/c more opening is ideal for blowing down than sucking. Who knows, i gave it a shot

Pirata · Post by **Pirata** » Fri May 21, 2004 8:59 am

@ Bluefront: you're right.

@ Rusty075: you talk about a new alpha, newer than the 8942?

And once again, anybody already played arount with Thermaltake's Ducting Mod?

Trip · Post by **Trip** » Fri May 21, 2004 2:25 pm

thermaltake ducting mod? i have one. it doesn't seem to help with slk 800A

I can't understand why people stick to blowing CPU fans!!!

I can't understand why people stick to blowing CPU fans!!!

Re: I can't understand why people stick to blowing CPU fans!

Re: I can't understand why people stick to blowing CPU fans!

Re: I can't understand why people stick to blowing CPU fans!