Case Airflow: Positive or Negative Pressure?

[Meyak2]

There are a few things that I have seen that have not been addressed. The first being fan placement. I understand the want to make a positive or negative system. But I have to agree with "A Bad Day" comment.

An air neutral system is the most effective (CFMin=CFMout) according to the laws of fluid dynamics (the principals that govern air flow) I am not saying that the other options are wring but they do add a lot of stress to components. Thermal failure is not the only issue you need to worry about with fans. Keep in mind that all fans generate magnetic fields.

This is not sufficient to destroy a hard drive or polarize a chipset. BUT it is enough to cause single bit errors in the data bus and we all know that means a slower bus speed, and less reliability. Current chipset manufactures use an adaptive method for data transfers. So over time as chips become more oxygen saturated and begin to fail they slow down. A similar analogy is in solid state drives and how they shrink. Over time the ROM in them goes bad. So the commonly used method developed by intel is that the cells go bad the hdd will no longer use them, so the available space with decrease by that much. So even the way MTBF is calculated is different than standard rotational drives. So the numbers you see are deceiving as apples and oranges. They are not an accurate representation of the same condition.

I digress. So back to my initial point. Location is a very important key, Air Pressure is important because it adds undue stress to components. Positive and negative air flow both have the plus and minus to them!

If you have a serious concern to the dust issue use an ion generator as well as air filters. That would eliminate ALL dust in the system. They themselves create a group of obstacles to overcome. High Voltage (so you would need a secondary power input), large magnetic fields (dangerous to hdd's so shielding would be required) and ultimately space. You can do this but I would not recommend this.

 

[SonicAgamemnon]

Regardless of total intake/exhaust CFM, please note air pressure ratios also change based on selected fan RPM rate. Using my case fan controller, I can easily change fan RPM rates which alters CFM in/out flow rates. Depending on which intake/exhaust fans are adjusted, it is easy to alter overall case pressure into various positive or negative ranges.

My system has nearly 2-to-1 positive case pressure based soley on total CFM in/outflow. Generally speaking, when every fan is running on the lowest RPM level, my case has rather high positive pressure. This initially provided very cool GPU, CPU and overall case temperatures, but after about an hour of normal use both the CPU and overall case temperatures began to slowly rise, although the GPUs remained stable. When I increased the exhaust RPM rates to maximum, leaving the intake RPM fan rates low, the case pressure turns negative. In this configuration, my GPU temperatures were notably higher, but the CPU and overall case temperatures were better duirng normal use.

However, things change once again when I increase system load (hard gaming or video transcoding). Using trial-and-error while monitoring the BIOS and two analog case temperature guages, I have confirmed my system works best during normal use (browsing, nothing major) with equal or slightly negative pressure (intake RPM low, exhaust RPM medium), with better CPU and overall case temperatures but a higher average GPU temperature. When pushing hard, my system works best with positive pressure (intake RPM medium, exhaust RPM medium), with a notably cooler GPU temperature and slightly cooler CPU and case temperatures. Under ultra heavy loads, the best configuration remains positive pressure (intake RPM high, exhaust RPM high).

Another factor to consider is targeting your hottest component and adjusting around that. In my system, the hottest components are two XFX Black Edition R7970s in crossfire. The positive case pressure brings more cool outside air directly onto these components, keeping them cooler and therefore they operate and generate less heat over time. This impacts my system more than any other consideration.

There are many variables.

My system is entirely air cooled. It would be interested to hear more from others who have transitioned to partial or total liquid cooling...

 

[OrcinusAdamantos]

I agree that intake fans only real purpose is to promote airflow in particular areas. If you had only exhaust fans you would still move as much air; it's not like you're creating a vacuum.
Think of a wind tunnel, they only exhaust. Now if computers were shaped like a tunnel you have no need for intake fans.

by your logic, a wind tunnel has magical atmosphere. Air just happens to come out of nowhere and EXHAUST the rear the tunnel.
BUZZ wrong!

A wind tunnel is a neutral air pressure environment. Is air jets blow air at high velocities across surface areas so designers can see the drag effect the components parts.

Something that you can do to test you air flow. (it may not be safe but whatever, neither is half the crap we do anyway) Im a cig smoker, and my PC sits about 6 ft from me. Now when I light up a cig, and exhale. I can watch the smoke get, attracted slowly, then build up velocity as it gets caught ni the air stream that my computer creates. And Since I noticed that, I decided to use my acrylic side to see where the smoke goes. Main intake, is a Huge Super High RPM fan used to cool the CPU. Then I have added several custom mounted exhaust fans. The smoke passes through the CPU sink, then depending on room temps, (if cool) stays low and exits the case past the RAM.

Then if its hot, it will split and raise upward also and get caught into the GPU's intake vortex. Then again at the PSU. So if you guys really want to solve your own theories, smoke out your computer (not overkill like, but enough to be seen) and post back here what you find.

 

[BoltHead513]

i have tried both positive and negative pressures in my HAF X and the temp difference is monumental. i really perfer positive, because of temps(open the case and get a blast of cold air) and that my 7990 OC to 1175 core and 1650 mem lives below 80C under sustained strenuous load. i couldn't do that with negative

-it may be possible that the haf x is made for + pressure, so my point may become invalid

 

[ScrewySqrl]

whoa. thread Resurrection from the dead. I had posts here nears 3 years ago

 

[ToastyMozart]

Positive to keep the dust out

But if you are pushing more air in than out therefore positive than you are pushing in more dust as well...

Except you can put dust filters over fans, it's much harder to filter the cracks and gaps in your case.
Positive airflow means you can control the path of the incoming air, where contaminants matter.

 

[reust]

Follow the ideal gas constant, p1 v1 = p2 v2, if you could ideally make 1 mole of Oxygen go from 22.4L at 1 atmosphere of pressure, to 2 atmospheres of pressure, you still have 1 mole of Oxygen just 11.2L in size now. At this point you would have raised the kinetic energy of the oxygen molecules causing them to bounce around on surfaces more, spreading and transferring heat more. why wouldn't you want to have twice the volume of oxygen molecules forced through the fans installed onto your graphics cards, cpu, ram and what ever other internal fans inside the case. Keep in mind simple things like hot air rises, you can't have all fans taking air in, but keeping it positive will do good. PS Does anyone else think the new computer cases coming out with this "mesh" external design just a way to melt your... everything faster? I sealed my computer to be air tight, besides the intake/ exhaust fans, and i dropped my GPU 15 degrees celcius. -- runs at 45 degrees now 550TI Asus( know this from hours of gaming on several accounts). positive pressure ftw. 210CFM intake, 150CFM exhaust.

 

[Fiftyfifth]

I don't really get all the debate in this thread, I think it's over complicating things. Your case is designed to accommodate a certain number of fans so in my mind you should try to work to that as much as your means will allow. Buy good fans. And in my opinion a side fan is a must, the difference in temperature to the motherboard is significant.

I think all that's required here is to install as many fans as your system (and wallet) will allow and then apply logic as to what the best layout will be. As for increased longevity of hardware, to me this point is made redundant by the speed at which technology is evolving. Your computer will last years without you even trying to preserve it but I bet you won't wait that long before you upgrade it - or at least parts of it.

I have 6 Noctua fans in my case (3x140 3x120), 1 of each fitted to my heatsink. At idle, my mobo runs 20-25'C, my CPU (i5-2500K OC@ 4.3GHz) runs 22-30'C and GPU (Asus 560ti top oc) runs 30'C. On load I rarely go over 50'C. So my conclusion is get good fans, install them using your brain and if you're unsatisfied with the result just make a few thought out adjustments. This will make a quick task of what's now been a 6 year debate

Oh and if you don't want to pay high-end fan prices don't scared to buy second hand fans. One of my Noctua's is second hand and it runs just as well as the rest, the technology invested in them is so advanced now that it's highly unlikely you're going to get some squeaky or rattling thing delivered to your door.

 

[bosa]

I just have my Power supply fan positive the rest negative. Love all the post for this question..lolol

 

[Sage Connel]

Let's use science and reasoning to solve this. Okay, if you have a tube, submerged in water ( a liquid) that behaves nearly identical to air ( a fluid) and you place a fan in the middle if there is no resistance what will happen? Water will flow at equal rates both in front and behind the fan or blower. Move it to the front it back of the tube, nothing changes. Rate in = rate out. Now add resistance and everything changes. Your going to have increased flow by the fan, and decreased the further you are from it, which will cause positive or negative pressure. Until you reach the threshold then performance will decrease and become equalized at said pressure. Introduce an equal blower at opposite ends and what does this do? Increase potential work against resistance creating flow and equalizing pressure in a steadier stream. Equal pressure is best. It creates the best wind tunnel and follows the fluids natural path. And this is why my cfi case with 8 fans front bottom intake/back too exhaust with120mm seidon on 4.4ghz fx8350 keeps it 35-39'C. Air is constantly flowing through parts, fresh blown ambient air equally through the tunnel. It flows fast enough to carry heat from the gpu to my rad and still cool effectively. My pc's exhaust is never hot, no pockets just a cool running PC

 

[abo2]

Stuff I've read suggests negative pressure is better for cooling since it's less likely to have stagnant areas in your case, but positive pressure is better for controlling/filtering your intake and avoiding dust. So I propose the following solution; positive pressure (for low dust) at low system loads, and negative pressure (for high cooling) at high system loads.

This can be done by careful use of mixture of fixed RPM and PWM or TC fans. Put the PWM or TC fans on your case exhausts, and use fixed RPM fans on your intakes. Note that most PSU's have temp/load variable fans in them and are normally part of the exhaust flow for your case. At low system loads, the exhaust fans will be running at low speed, so the fixed RPM intakes will give you positive pressure. When your system gets loaded, the exhaust fans will spin up and give you negative pressure.

 

[Motopsychojdn]

Let's use science and reasoning to solve this. Okay, if you have a tube, submerged in water ( a liquid) that behaves nearly identical to air ( a fluid) and you place a fan in the middle if there is no resistance what will happen? Water will flow at equal rates both in front and behind the fan or blower. Move it to the front it back of the tube, nothing changes. Rate in = rate out. Now add resistance and everything changes. Your going to have increased flow by the fan, and decreased the further you are from it, which will cause positive or negative pressure. Until you reach the threshold then performance will decrease and become equalized at said pressure. Introduce an equal blower at opposite ends and what does this do? Increase potential work against resistance creating flow and equalizing pressure in a steadier stream. Equal pressure is best. It creates the best wind tunnel and follows the fluids natural path. And this is why my cfi case with 8 fans front bottom intake/back too exhaust with120mm seidon on 4.4ghz fx8350 keeps it 35-39'C. Air is constantly flowing through parts, fresh blown ambient air equally through the tunnel. It flows fast enough to carry heat from the gpu to my rad and still cool effectively. My pc's exhaust is never hot, no pockets just a cool running PC

This is fine logic assuming a clear undisturbed path for the airflow, however your case I'll wager is not 100% leak free and there are impediments to the airflow that create eddies, equal pressure in a sealed tube is obviously optimum but variables can introduce many a tricky situation which is why a lot of us advocate experimentation, only you can find the optimum setup for your case, in your room.
@Abo2, most decent cases/Psu setups take from outside the case and vent to the rear of the Psu, which means there is no effect on the internal airflow from the Psu, other than possibly a small amount of heat emanating from the Psu into the cabinet space

 

[ScrewySqrl]

wow. this thread got ressurected from the dead. Multiple times. The original post is from 2007!

 

[Motopsychojdn]

Yup I got a notification mail so thought I'd put a little more into it
Moto

 

[slang103]

For the rigs I put together I've always had to be cognoscente of the dust issue. I myself live in a place that for some unknown reason is prone to dust no matter how well we clean or how recent the furnace filter has been replaced. That is a mystery befitting its own topic. Also for the people I have put rigs together for, they have little desire to frequently be taking their computers out and deep cleaning them. Nerd moan all you like about how it is good to do, the fact is, most people wont do it.

My solution: More fans (requiring less cfm each), easy to clean filtration for air and if I have to choose... just a hair on the positive pressure side.

My current rig has:
4x 120mm intake fans (2 front, 1 side & 1 psu).
2x 120mm rear exhaust fans (liquid cooled cpu radiator attached)
1x 200mm top exhaust fan

Based on the placement of fans & set rpms, there are both no places in the case where air is still (hot pockets) and pressure is just a hair on the positive side.

People can judge all they want about needing to "just clean your computer" but I was having to do it weekly before changing my methods and adding filtration. With 2 tower computers at home and a bunch at work that I take care of, that was beyond stupid. Now, they run cooler because the radiators and heat syncs aren't covered in dust AND cleaning consists of turning the computer off & vacuuming the filters off.

Fans don't cost much relative to your performance computer and nice ones @ slow rpms produce very little noise. Get a case that allows for a good distribution of quite a few fans and you won't have issues of hot pockets. My rig after 1 week of running dropped 10c on both cpu & gpu compared to what it would have been as soon as I made my changes (filters & side mount 120mm fan). Cold air in bottom & lower front, out top & upper back.

 

[Brian Blair]

Negative. Heat can't be a problem if it can stay. Not to mention the latter brings dust hat creates major heating issues. But my case is a Antec 300, And set up with 2 front intakes. So I have kinda a front to back flow going on. And it works.

 

[jaimelmiel]

I have two gaming machines so it is all about negative air flow. Especially with my Fx 9590 Machine
I use an Msi R9 290x wideo card. This is heat city. It is about moving the air. 140 cfm coming in the lower and mid front
and 270 cfm going out the back and top of a Coolermaster Haf 912 case. The sides have 2 120mm fan openings
that allows cool air to be drawn in due to the difference between The intake and exhaust. I truly believe if your a gamer
a properly designed negative flow setup is the only way to go. Yes there will be dust Every so often I take the system apart
an clean it. But if your system doesn't have a dedicated video card and you do not have a high GHZ cpu then go positive airflow.

 

[ScrewySqrl]

the scary thing is this thread is only 3 pages long, and its 7 years old!

Personally, I tend toward a neutral or slightly positive flow.

 

[booyaah]

Here is Silverstone's take on it (see the nice animated diagram):

http://www.silverstonetek.com/techtalk_cont.php?tid=wh_positive&area=en

p.s. It has a really high google page rank...google's algorithm must like old stuff.

 

[RESONANT]

I think negative or positive pressure isn't really the answer to heat management in a computer case. If your fan system is creating either positive or negative pressure then you can assume that it isn't designed with the greatest effectiveness.

Your only concern should be replacing the hot air inside the case with fresh air as quickly as possible. you can measure the systems effectiveness by how many times per minute the air is replaced inside the case. If you have positive or negative pressure, your airflow will be lower as it's bottle-necking somewhere in the design. equal pressure into and out of the case would be best.

Remembering that there are quite a few vents on a good case, you should consider those negative pressure zones, and then add more input fans accordingly. This means that even though you have more fans blowing in, your overall effect is an equal pressure airflow design. (closing a few vents off for better flow is sometimes actually a good thing)

If your really serious, then temperature sensors throughout the case and testing is the only way to be sure. Over-engineer what your not sure about and you'll be fine. If you don't have enough fans, then allow the vents to be your output, and focus the fans onto the key components directly.

My personal design for crazy cooling involves low viscosity mineral oil submersion, and an electric thermocouple (peltier) connected to a water cooler and a quad radiator (either directly on the cpu, or cooling the oil, not sure yet, but I think the oil would just freeze and the properties would not work throughout the case), the oil submersion allows for sub zero cooling and no condensation. (think liquid nitrogen)
"It's mad scientist kinda crazy!"

 

[jaimelmiel]

I think negative or positive pressure isn't really the answer to heat management in a computer case. If your fan system is creating either positive or negative pressure then you can assume that it isn't designed with the greatest effectiveness.

Your only concern should be replacing the hot air inside the case with fresh air as quickly as possible. you can measure the systems effectiveness by how many times per minute the air is replaced inside the case. If you have positive or negative pressure, your airflow will be lower as it's bottle-necking somewhere in the design. equal pressure into and out of the case would be best.

Remembering that there are quite a few vents on a good case, you should consider those negative pressure zones, and then add more input fans accordingly. This means that even though you have more fans blowing in, your overall effect is an equal pressure airflow design. (closing a few vents off for better flow is sometimes actually a good thing)

If your really serious, then temperature sensors throughout the case and testing is the only way to be sure. Over-engineer what your not sure about and you'll be fine. If you don't have enough fans, then allow the vents to be your output, and focus the fans onto the key components directly.

My personal design for crazy cooling involves low viscosity mineral oil submersion, and an electric thermocouple (peltier) connected to a water cooler and a quad radiator (either directly on the cpu, or cooling the oil, not sure yet, but I think the oil would just freeze and the properties would not work throughout the case), the oil submersion allows for sub zero cooling and no condensation. (think liquid nitrogen)
"It's mad scientist kinda crazy!"

 

[jaimelmiel]

The point is depending. on how you set your fans along with the design of the case you will either have positive or
negative pressure. It is up to you to figure that out. You need to choose. Choose wisely.

 

[Outlander_04]

Positive/negative discussion is bogus to the core

There is no way to get any useful pressure rise using the kind of axial fans computers use when the case is also essentially open .
What is important is air flow over the hotest components

 

[B_S_N]

I have read almost all the answers....

I will go for negative pressure because.... Its more efficient..... There is no chance of forming a Air pocket inside(in negative pressure).... But you must put an air filter at the inlet....... Because Negative pressure tends to pull the air.... which pulls the dust particles also.... the dust particles not only restrict heat transfer it will damage your case too......

So my suggestion is Negative pressure with air filter and periodic cleaning.....

 

[vriend]

I prefer Negative - there is more air exiting the case via exhaust fans than fresh air entering the case via intake fans and vents. The CFM for exhaust fans exceeds the CFM from intake fans.

The main reason is that most of the internal components in a conventional PC are individually cooled; think about the GPU, most cards have twin fans and conventional CPU cooling has of course always a minimum of 1 fan, and some people also have RAM cooling etc.

So I see no need of importing more air into the case, except for cooling down the radiators of the already present cooling parts. But I think the effect of that is minimal, especially if you have good fans installed to the radiators.

I believe that the focus should be on relieving the case from all the hot air that is exhausted by the cooling fans of the individual parts.

Also you have far less dust in on your internal PC parts. This is a positive contribution to lifetime of your equipment.

But of course it all starts with good premium case that allows and maintains a good negative air pressure. This is imperative.

Cheaper and budget cases will not work as good as a highly developed PC case.