Cooling Fans Setup

[Fowlerflap]

I have an Antec case with 2 cooling fans, 1 in front and 1 in back.
What would provide the best cooling, having the rear fan pull air out of the case and the front fan pull air into the case?
Or have both fans pull air out of the case?
Just curious what the normal setup is.

Thanks

[Cricket]

I would set up the fans on the back of the case as exhaust fans to pull heat out of the case. A vacuum will be created inside the case and this will cause cool air to be drawn in naturally from the front of the case. This set up works well and it puts the fans farther from your ears so the system is a little less noisy.

Cricket

[TwoRails]

Yes, both sucking out hot air and drawing air thru the case is more effective. The reason is that you have a laminar (smoother) air flow thru the case, whereas if you put a front fan in, it will create turbulance and has a tendency to "stir" the hot air around in the case.

TwoRails

[ACiDReFLuX666]

great. i find this out just after i buy 4 fans for my comp. guess ill use the xtras on this one. thx for the info

[bigandy]

I agree with the having at least one exhaust fan, and one to pull in cool air. I'm personally using two exhaust fans and one to pull in cooler air.

[silvergrin]

120mm - 120cfm - intake
92mm - 119cfm - intake
80mm - 78cfm - intake
60mm - 70cfm - intake
40mm - about 20cfm - exhaust

lol probably should re-arrange it, but heh.

[Cricket]

silvergrin, you got way too much intake.

Cricket

[mike breck]

The classic fan arrangement is one front/bottom drawing cool air in and one top(or middle)/back exhausting the hot air. This in most instances will work very effectively.

With this setup you could almost say intake and exhaust are more or less balanced.

However, having tried innumerable setups, I found the one that gave me lowest mobo and CPU temps was one front/bottom intake and TWO top/middle/back exhausting.

That will help you achieve a what Hal calls a negative air pressure. In simple terms, this means you have more air being exhausted from the case than being taken in.

However, it must be said, that some have argued that a positive air pressure is best. So it really depends in which camp you fall.

The best way to find out, is to try various fan combinations, measuring the results in a scientific manner, and see what works best for your case.

HTH

[TwoRails]

Actually we need to swap the terms “negative and positive pressures” with “laminar (smooth) and turbulent air flow.” With one or two exhaust fans, you get air drawn smoothly thru the case, which is called laminar air flow. With one or two intake fans pushing air into the case, you get a lot of turbulence in the case.

When air is being drawn out, you get a “change” of air inside the case rated at CFM (cubic feet per minute). This is what cools the system. When air is being pushed into the case, it gets stirred up and the hot air circulates and doesn’t cool effectively.

That why even AMD recommends to *not* use front fans.

The terms of negative and positive pressures are misleading and inaccurate as it really doesn’t exist. If one were to doubt this, it is easy to test. Simply hook up a typical vacuum / pressure gauge to the case and you will *not* see any fluctuations.

If you don’t happen to have a gauge handy, then simply just look at the sides of the case: do the sides bow in when you turn the system on? No. So there is no negative pressure. Do they bow outward? No. So there is no positive pressure.

TwoRails

[mike breck]

Well, TwoRails, I'd like to argue/discuss this with you in a friendly way. As I know you like a good discussion (better known as an argy-bargy in my neck of the woods), I know you won't take offence.



I don't think Positive and Negative air pressure are misleading terms because they are used in describing the flow of air in buildings e.g. attics, rooms. For example, many rooms in Hospitals are described as having a Negative airflow i.e. more air being exhausted than drawn in. The germs in the room are exhausted by the negative airflow. Everything gets swept in the direction of the exhaust fan or ventilator.

Now to acheive this goal, the room doesn't have to be a Wind Tunnel - only a slight difference in the intake and exhaust levels are necessary. So if the isolation room (or PC Case) had polythene walls (like a mobile containment unit) then you might see the walls being sucked in slightly; however, there is no chance of this happening with brick or sheet metal - unless you were using a "huge" amount of pressure.

Now if AMD are recommending that we don't use a front fan and just exhaust fan(s), then, that adheres to the theory of Negative airflow, because more air is being exhausted than drawn in.

As far as turbulent/smooth airflow is concerned, it seems to me that you will get turbulence no matter what you do - unless your remove "all" the components from the case. Any obstructions such as cables, PCI cards, bays, and most importantly, the CPU fan will disrupt the "S" airflow from front/bottom (where the air holes or fan is) to top/ back.

The only thing you can do is set up a strong enough airflow to cool the "particular" components in a "particular" situation. So a hot running CPU, video card, HD, and chipset, will require more air cooling to acheive a certain lower temp than cooler components.

As a case is not a sealed unit, I wouldn't imagine that there would be a perceptable great difference in the pressure inside the case relative to the room. However, according to the standard definition of negative air pressure, if more air is exhausted than taken in, then it would not be incorrect to use the term in a case cooling situation.

It makes sense when you think about it. The longer hot air has time to lurk inside the case, the hotter the case will be. So the quicker you can get rid of that heated, the cooler the case will be.

However, as I've already said, I don't think this method is set in stone, and freely accept, what was the best arrangement for my case and components, is not necessarily the best for "all" setups.

I would still encourage people to try different arrangements and see what works best for their situation.

Your Honour, I rest my (PC) case.

[minsonngo]

I agree with mike breck, just try different arrangements and use the one that best suits that particular system.

[TwoRails]

Originally posted by mike breck Well, TwoRails, I'd like to argue/discuss this with you in a friendly way. As I know you like a good discussion (better known as an argy-bargy in my neck of the woods), I know you won't take offence.

a: I’d take it no other way! (But it does feel a little funny because we seem to have similar thoughts most of the time.) b: Yes I do c: Right again you are!

I don't think Positive and Negative air pressure are misleading terms because they are used in describing the flow of air in buildings e.g. attics, rooms. For example, many rooms in Hospitals are described as having a Negative airflow i.e. more air being exhausted than drawn in. The germs in the room are exhausted by the negative airflow. Everything gets swept in the direction of the exhaust fan or ventilator.

The last time this subject came up I redid some research on the subject. I spoke to 4 or 5 engineers on the subject and did some reading. In regards to rooms in your house or the like, if there is any negative or positive pressures present then they all stated that something is wrong with the system; there is either a leak somewhere or something is clogged to account for it.

In regards to hospitals and the like, air is drawn out of certain rooms and it works as the air flows smoothing by being drawn out. If the air was being pushed into the room, then you will have turbulence and things get stirred up and don’t exit the room as desired.

Now to acheive this goal, the room doesn't have to be a Wind Tunnel - only a slight difference in the intake and exhaust levels are necessary. So if the isolation room (or PC Case) had polythene walls (like a mobile containment unit) then you might see the walls being sucked in slightly; however, there is no chance of this happening with brick or sheet metal - unless you were using a "huge" amount of pressure.

Some of my work is done in Negative pressure buildings and rooms, and yes, to actually keep in what needs to be kept in does require sufficient air flow as to be noticeable. When you enter, you can actually feel the air go past you. On exit, it is difficult to open the door, like it was an extremely heavy door, until it is opened somewhat.

To get Negative pressure you need several things, some of which are (and can vary depending on the actual conditions / requirements): restricted air inlet and a relatively sealed off area, both of which don’t really exist in PC cases. And, of course, a strong enough system to draw even a partial “vacuum.” More on this later.

Now if AMD are recommending that we don't use a front fan and just exhaust fan(s), then, that adheres to the theory of Negative airflow, because more air is being exhausted than drawn in.

Sorry, but their reason is the same as mine: you have too much turbulence when air is pushed into a PC case, room, or even a pipe!

Let’s take the pipe one more step. If you put a fan on a pipe so it will draw air thru the pipe, then you have smooth, laminar air flow thru the pipe. If you turn the fan around and push air thru the pipe, there is a lot of turbulence thru the pipe. The laminar flow is not due to Negative pressure, and the turbulence is not due to Positive pressure.

As far as turbulent/smooth airflow is concerned, it seems to me that you will get turbulence no matter what you do - unless your remove "all" the components from the case. Any obstructions such as cables, PCI cards, bays, and most importantly, the CPU fan will disrupt the "S" airflow from front/bottom (where the air holes or fan is) to top/ back.

Yes and no. With air being drawn out, the general flow is laminar, but no, it’s not perfect. Just a lot better than with air being pushed in. Drawing it out insures that there is a Constant Change of air thru the case for the CPU fan to draw from. It is this constant change of air that cools everything from your car radiator to your CPU. When air is being pushed in, yes there is some exchange of fresh air, but a lot of it just gets stirred around in the case before it exits.

Think of a bucket with a big hole in the bottom of it, but not big enough for the bucket to not hold water. Now, with a hose filling the bucket from the top, and the water running out the bottom, even though there is water in the bucket, you can easily imagine that the water gets changed out at a steady rate. Now think of the same bucket, but with no hole in the bottom. With the water still being put in, it now overflows the top of the bucket. How long will it take for the water that is Not near the top to be changed out? There might be a lot of water circulating in the lower part of the bucket but not getting changed out with fresh water.

The only thing you can do is set up a strong enough airflow to cool the "particular" components in a "particular" situation. So a hot running CPU, video card, HD, and chipset, will require more air cooling to acheive a certain lower temp than cooler components.

If you by “more air cooling” you mean sufficient air flow, a.k.a. sufficient CFM, thru the case, then we agree. Just “more power” doesn’t cut it, however, and is the main reason that no matter how many fans people put in their system, they still have a cooling problem. As you say in your closing, playing with arrangements are often necessary. There to many situations, nonetheless, where a PC is cooling is vastly improved by eliminating the overkill on fans and reducing their number to just one or two.

As a case is not a sealed unit, I wouldn't imagine that there would be a perceptable great difference in the pressure inside the case relative to the room. However, according to the standard definition of negative air pressure, if more air is exhausted than taken in, then it would not be incorrect to use the term in a case cooling situation.

Actually, you’re reinforcing my side here! Everything from physics to smoke testing PC cooling systems / cases prove the large benefit of laminar versus turbulent air flow. If something is so negligible that it requires expensive, highly sensitive scientific equipment to even (possibly) measure, then how can this negligible factor be doing the work that the smooth air flow has been tested to do?

It makes sense when you think about it. The longer hot air has time to lurk inside the case, the hotter the case will be. So the quicker you can get rid of that heated, the cooler the case will be.

Exactly !!! That’s why you want the constant change of air that smooth air flow provides, and turbulence doesn’t. Again this is proven by smoke tests.

However, as I've already said, I don't think this method is set in stone, and freely accept, what was the best arrangement for my case and components, is not necessarily the best for "all" setups. I would still encourage people to try different arrangements and see what works best for their situation. Your Honour, I rest my (PC) case.

Mostly agreed. I also don’t believe in “one-size-fits-all” and adjustments are often necessary. My place to start is one or two rear fans (depending on the case, of course). I don’t add fans if temps are reasonable. My experience shows that I get no added benefit from installing more fans and don’t have any front fans in any of my personal systems.

In closing: Negative pressure is pretty easy to achieve if you think about how little effort is required to cave in the sides of an empty, plastic 2 liter soda bottle by drawing air out with you mouth. Now, hopefully to remove all doubt about what actually does the work in cooling (as we are discussing it), try this little test (I have) :: Take a 80mm case fan. Blow up a paper bag. Put the open end of the bag around the frame of the fan, with the fan pointed to draw air out of the bag. Seal the bag around the fan (a rubber band works nicely!) Turn the fan on. Gee.... guess what? The fan does not even have the power to suck the bag in and collapse it! The Negative pressure (if any) it develops isn’t even enough to “punch it’s way into a paper bag” ! LOL


Humbly and Respectfully submitted for Your Consideration, Dear Sir



Sincerely
TwoRails

[mike breck]

Well, TwoRails, you have put an really, excellent arguement; but you have me at a disadvantage; because I can't find any info on the advantages and disadvantages of Laminar v Turbulent airflows in a PC case.

However, in the meantime, lets' just look at the question of Negative and Positive airflows and their effects on cooling.

In regards to rooms in your house or the like, if there is any negative or positive pressures present then they all stated that something is wrong with the system; there is either a leak somewhere or something is clogged to account for it.

Not necessarily. See this link for a description of house ventilation using both Negative and Positive pressure:

http://216.239.59.104/search?q=cache...n&ie=UTF-8</a>

Some of my work is done in Negative pressure buildings and rooms, and yes, to actually keep in what needs to be kept in does require sufficient air flow as to be noticeable.

I suppose this will depend on how much Negative pressure is required in a given situation. Negative air pressure doesn't have to be so preceptable as you suggest.

"Establishing an Isolation Room with Negative Pressure

Pressure differential

The minimum pressure difference necessary to achieve and maintain negative pressure that will result in airflow into the room is very small (0.001 inch of water). Higher pressures ( greater than or equal to 0.001 inch of water) are satisfactory; however, these higher pressures may be difficult to achieve. The actual level of negative pressure achieved will depend on the difference in the ventilation exhaust and supply flows and the physical configuration of the room, including the airflow path and flow openings. If the room is well sealed, negative pressures greater than the minimum of 0.001 inch of water may be readily achieved. However, if rooms are not well sealed, as may be the case in many facilities (especially older facilities), achieving higher negative pressures may require exhaust/supply flow differentials beyond the capability of the ventilation system."

"To establish negative pressure in a room that has a normally functioning ventilation system, the room supply and exhaust airflows are first balanced to achieve an exhaust flow of either 10% or 50 cubic feet per minute (cfm) greater than the supply (whichever is the greater)."

http://www.berriman-usa.com/negative.htm

To obtain a greater Negative pressure than 0.001 inch of water, then you've got to seal the room (or PC case) and/or increase the CFM of the ventilation fans. I would imagine the Negative air pressure requirements for a electronics "clean" room might be more than the attic of a house, a PC, or even an Isolation room or tent in a Hospital. A difference of 10% between the intake airflow and exhaust airflow isn't that much to achieve Negative pressure.

As for the paper bag analogy, the 80mm fan may cause Negative air pressure inside the paper bag without drawing it's sides in. After all, it doesn't have to be strong enough to displace paper - just air.

Having said that, there's certainly enough of a change in air pressure for a strip of paper to cling to the 80mm 45.2 CFM intake fan on my machines.

Let’s take the pipe one more step. If you put a fan on a pipe so it will draw air thru the pipe, then you have smooth, laminar air flow thru the pipe. If you turn the fan around and push air thru the pipe, there is a lot of turbulence thru the pipe. The laminar flow is not due to Negative pressure, and the turbulence is not due to Positive pressure.

With the fan sucking air, you are creating Negative air pressure inside the pipe. When the fan is blowing into the pipe, you are creating Postive air pressure. This is exactly the same as sucking thro a straw. When you suck in, the air pressure inside the straw is Negative.

However, if we look at what has been said so far in this discussion, it seems to me that we agree on one thing - you are better with more air being exhausted than drawn in. This remains true whether you have a case fan at the front or not.

What we disagree on is:

a) Whether you should call this state a Negative air flow.

b) Whether you should have an intake fan bottom/front because of the turbulance it causes.

The most common definition of Negative airflow is:

"Negative air pressure means that air is drawn into the building to replace the air removed by exhaust systems or other building conditions."

However, you may be correct in that a PC case with one intake fan and two exhuast fans is not an exact model of Negative air flow. Can it be still be called Negative air flow when air is being pushed in, instead of drawn in?

To be quite honest, after the amount of contradictaory info I've read over the last 24 hours, I can't be certain.

It may infact be closer to what one Ventilation system manufacturer called Push/Pull Ventilation. Here are some definitions of three types of Ventilation:

"With a Negative Pressure system, exhaust fans are used to try and suck heat and fumes out of the building. Negative pressure neutralizes in all directions -- top, bottom, left or right. So, if the emissions or heat are in the middle of a plant, negative pressure is very inefficent. The challenge with a stand-alone exhaust system is that most do not have the power to gather a large portion of air, and will only remove the heat and fumes directly in front of the exhaust fan. A simple illustration of Negative Pressure: try extinguishing a candle by inhaling rather than exhaling."

Their definition of Negative air pressure is closer to your position of having no intake fans.

"With a Balanced Air, or Push/Pull, System there are fans at one end of a building pushing air toward an exhaust system either on the ceiling or opposite wall. The challenges with a traditional Balanced Air System is that it requires expensive duct work, and there will quite often be "dead spots" as air flow can only travel limited distances depending on fan size or duct locations."

This infact, could be another explanation of what happens when you have an intake fan at the front, as well as exhaust fans. The air is drawn in by the exhuast fans and "pushed" in by the intake fan.

The air from the intake fan is circulated around the case, taking heat with it, and then forced to the top of the case by a mixture of natural convection, possibly Negative air pressure, and the new air comming in. When it reaches the top, it is "pulled" out by the exhaust fans.


"In a Positive Pressure System, shrouded fans are used and air is introduced into the space, increasing air pressure equally at all points inside the structure. When an exhaust opening is created, all of the interior air moves in one mass to the exhaust point. The key is to seal the opening with incoming air so the air within the building can only escape through the exhaust point. You must have more air coming in than going out."

Well some people do it this way, as well.

I must admit, your theory of Laminar air sounds really elegant, but having tried innumerable experiments with case fan arrangements with my T/Bird 1200 in a full tower (including just using one, two , and three exhaust fans and no intake fan), I have to say that two exhaust and one intake worked best for my situation.

This gave the the best temps of CPU 35c and mobo 22c.
At 22c, the mobo/case was just one degree above ambient room temp. That ratio was maintained as the ambient temp went up.

So Turbulent works for me and Laminar works for you and the existence of Negative or Positive air pressure inside a PC case is still open to debate.

We're back where we started - different strokes for different folks - and PC cases.

[TwoRails]

Hi mike breck,

I've very pressed for time "rat" now, and do need to re-read the above, so I'll have to get back to you and everyone later (hopefully tonight..). Before I do, however, I'd like to read the stuff in the link you provided but I could not get it to work.

Later -

TwoRails

[FLG]

OK i have and idea im not quite shure if it will work or anything like that but here it goes..you know those fog machines well why not get one and make the fog go into you pc and see exactly where the air is going?the only thing is that you need a case with a window. now the fog that comes from these machines are made from a water mixed with some other stuff to make the fog so water residue could build up but i looked around and it said that the residure only forms directly in front of the machine so if you put it away from your pc but enough so the fog goes in then it should be ok, and the other thing is these machines are about 60$ for the cheapest one.

[TwoRails]

Hi FLG,

That's what I was refering to with the "smoke" testing and I think it's an excellent idea

TwoRails

[mike breck]

Hi TwoRails,

That first link on House Ventilation is not right. This should take you there.

http://www.certainteed.com/solutions...out_howwhy.pdf

This is link pertaining to Push/Pull airflow

http://www.fitventilation.com/howitworks.htm

FLG

Yes, smoke testing is an excellent way of analysing airflow; however, from a PC enthusiast's point of view, the bottom line is the CPU, mobo, HD, and Video card temps.

The airflow that gives the lowest temps in a given situation will be the optimal one. However, obtaining the optimal airflow may come back to the old tradeoff between temps and noise i.e. how to obtain the lowest temps with minimal case fan and airflow noise.

[TwoRails]

Hi mike breck and Everybody,

I am at a disadvantage as I have severe time limitations during the work week to give my response the proper time it deserves. So, being pressed for time, I will have to make this short.

In one sense I will concede that the terms Negative and Positive exist, but only in the strict “technical” and “scientific” sense as this is the way I have come to believe it is being referred to in this context.

I do so *only* in the definition that: as I type this, or anybody that makes a post or any other movement with their body, my fingers are making little pockets of negative and positive pressures. Any movement creates pockets of pressure differentials.

But that is not was does the cooling. Cooling is done, and everybody will agree with this, by air (or liquid / fluid, etc.) circulation. All heat exchangers function under this principle. Air must move thru or around the heat exchanger to work.

My whole point is that laminar / smooth air flow is much more efficient in cooling than is turbulent air flow. This is supported not only in the science of Physics but is backed up by individual testing such as done by AMD and other manufactures. Even case reviews by places like www.tomshardware.com often discuss a cases poor design in as such as it creates turbulent air flow.

Just look up why a golf ball has dimples and whichever link you follow will tell you they are there because they give the golf ball more laminar flow which reduces the Negative pressure pocket behind the ball, thus giving it longer flight. Laminar flow is also why such things as air planes and submarines have slender tails (i.e.: tapered at the end )

The two links above don’t really apply, in my humble opinion, for several reasons. One is that they a sales pitches to their own products, and of course, their products are the best and will do exactly as advertised, as all products do. The first, about attic ventilation doesn’t apply as it’s for special circumstances, such as ice dams, water expansion at freezing, and shingle damage.

The second link reinforces the difference of turbulence and laminar air flow. Just look at their own photo of a “conventional” fan: it shows tons of turbulence, as it should. “Their” fan, as they claim, produces less turbulence. They are describing how laminar flow is better than turbulent flow but only using different wording. The whole purpose is to get enough air changed out to cool the system without recirculating the existing air: Read: efficiency. Read: laminar

The key to cooling is keeping the air moving. You can do it with a dozen fans, as long as enough cool air crosses the components to be cooled. For efficient cooling, which also means “more quite” in our situation, you want air flowing *thru* the case, not recirculating around inside the case.

That is my usage of laminar and turbulent.

We also agree that some experimenting is often required due to case design and other variables. One of the hottest running case I every built with (and it was only one due to the poor design) had an extreme number of holes punched into it for decoration. The poor design came from all the holes where in the 1/3 rear of the case. Even with multi rear fans, all the air was being drawn in thru these holes in the rear and not thru the case. That case did require a front fan to keep cool.

So, yes, it may take one, two, even three fans to cool a box down. In my own experiments, I have found in cases without weird holes all over the place, that one or two rear fans is more than plenty. But I have learned to start with just the rear fan that is typically in most modern cases.

Even my wife’s computer, with a 1.333 Ghz T-Bird, runs plenty cool with just one 80mm rear fan and no front fan. But then the only intake is in the front of the case.

In closing I would like to really (for real) apologize for not having more time to spend on this tonight.

In the mean time, Happy Computing and keep cool!!

TwoRails

[mike breck]

Thanks your reply TwoRails.

As you're really busy at the moment, we shall end the discussion there. I dare say the discussion will be revived at some point in the future

It's been interesting and given me food for thought.

Sorry for waylaying your original Post Fowlerflap, but I couldn't resist discussing this subject with TwoRails.