Case Airflow: Positive or Negative Pressure?

[Paperdoc]

the absolute easiest way that i have found to show which is better is with a shop vac. take a digital thermometer and hold it in place. have the shop vac blow air onto it. note the temperature. now, do the same with suction. the vacuum test shows that negative pressure is far superior to that of positive pressure.

just thought i would (and being blown over the thermometer yopu propose)add my 2 cents.

You should realize that most vacuums take the air they suck in and, after removing the dust, flow it over the motor as its cooling source, then exhaust it. So any air coming OUT of the vac (and being blown over the thermometer you propose) has been HEATED BY THE MOTOR. Guess why the thermometer says the temperature is higher!! Your test results have absolutely nothing to do with whether suction or blowing is best for case ventilation.

 

[tainted_peak]

(If you don't feel like reading go to the last line.)

I know this is dead but I wanted to point out/make a point. The POINT of this topic was which is "better" positive or negative air flow.

So disregard everything non substantial relating to the topic: dust, cable management, etc

Lets make this purely theoretical, shall we?

Just consider positive or negative pressure nothing else, no what if's or Ilikecablemanagementsoiroxz0rs.

You must remember that blowing a fan on a cpu or mobo is NOT positive pressure its call COOLING the cpu, etc. I know in real life they make a difference but we're eliminating all sources of error and what if's, real this is purely hypothetical. And also ignoring convection currents and the fact heat rises and expands and cold air condenses.

Theoretically gfx card fans, cpu coolers, and PSU must be disregarded since this is not positive or negative to case pressure.
(If you have a problem with this read the line above)

I think everyone would agree ideally a "windtunnel" type situation where equal over qualified fans would work in pairs to blow air across and then suck it out as soon as possible would be most effective (you'd have a totally even and constant airflow of fresh air). But this is not the argument.

Assume you had a case reasonably size (mid tower) and that was perfectly sealed and you did have high enough CFM fans to push and pull air in and out of the case.

Assume 3x120mm that push 120 cfm each and 1x80mm that pulls 50cfm. You'd have positive pressure in the case at this point. This would in turn cause the fans to push less air since they are now pushing against air that has accumulated in the case but of the positive pressure from poor exhaust.

Assume its the other way around 3x120mm pulling and 1x80 pushing. This negative pressure will allow the the fan to perform better since no pressure will be applied to it pushes air.

Something everyone can agree on perfect example
Negative is better why? Go cut off the muffler off your car. Why does it get better gas mileage and HP without the muffler. Lets just say back pressure (positive pressure) is the culprit.

 

[OvrClkr]

IMO as long as my temps don't exceed 60c (overall) I could care less....

I guess this thread would benefit peeps who have cases with VERY bad airflow....

 

[tucansam]

Lots of evac fans with the entire front of the case devoid of fans but full of filtered mesh.

/thread

 

[lotus69]

I think a negative case flow with a wind tunnel setup is the best.

 

[lotus69]

I think a negative case flow with a wind tunnel setup is the best.

 

[DragonBlack]

Cant takes sides Becuz...Actually, with a set up neither favoring positive nor negative, minus the disadvantages of dust being sucked in from small gaps from negative pressure,

both are essentially the Same Thing if one can think slightly out of the box (casing).

Too much exhaust in a room for e.g. with a small hole of cold intake air, it creates a hot vacuum from internal heat due to little cold air coming in. After some time, more vacuum will create a hotter casing like a vacuum flask giving you Hot Coffee.

Too much intake on the other hand, with a small hole for hot air to escape for e.g., will slowly build up more and more pressure. No matter how much cold air coming in, if it cant escape out fast enough, the hot air will still neutralize the cold intake, going back to square one, like a hot Pressure Cooker.

Both extremes are thus problematic. Best is to create a balance pressure like a wind tunnel effect. By standing at any one point in the wind tunnel, both ends pull and push at same pressure without disruption. Then any hot component, if placed in the centre of this tunnel, will be perfectly cooled.

But in a typical pc setup, it is be very difficult to find out all Sweet Spots for every hot component, since all are at different locations, akin to chasing the wind. One will eventually be cooled more or less than the others.

One method to resolve this issue, which is already done by manufacturers for their high end casings is to create 'Heat' zones, separating the casing into 2 or 3 compartments such as power supply at the bottom, hard disks on top and motherboard in the centre, then tackle it from there which is more manageable.

Another cheaper alternative is similar to the automobile's Cold Air Intake concept with heat shielding which essentially creates thermal insulation for air filter, isolate it from the hotter engine components and re-route it to be exposed to the coldest possible incoming air.

One will then create separate mini 'wind tunnels' i.e. air ducts with duct tapes and sealed off the whole CPU cooler, flexible air pipes etc. (a little similar to the water cooling concept, except its shielded.) then exposed it to only one intake and exhaust fans running at same pressure.

There are, of course, other 'secret recipes' method, done by Pros, specially used for overclocking, which is constantly being researched everyday to find the Holy Grail

 

[BOWANDLUKE]

5) A good strong airflow might bypass some areas that need cooling. Has anybody installed an internal fan just to mix up any internal hot spots? Did it help?

i did. it did NOT help much. i took an 80mm fan and put inside on top of my DVD drive and removed a section of front acrylic plastic for max air mass. the case temp drops significantly. (Positive Intake Air)
However no matter what case fan configuration, my cpu temp is the same. I use 2 front intake, 1 side intake and 1 rear exhaust. The PSU is it's own settup it gets rid of it's own heat. It has nothing to do with my case temp or cpu temp. So, i guess i'm a positive person here. lol.

 

[Guest]

I think I may have an answer for the positive vs. negative pressure debate. Most graphics cards blow air from the inside of the case through their internal heatsinks and out through the back of the card. If there is more pressure inside of the case than outside, i.e. positive pressure, this helps blow air through the video cards, helping to keep them cool. If there is negative pressure inside of the computer case, air attempts to be drawn backwards through the video card causing less overall airflow through the card. Using loose threads taped over the video card exhaust to measure air velocity, I tried both negative and positive pressure fan arrangements on my computer case. The air exhausting from the video card moved very much faster when the case had positive pressure. Increasing the airflow to 2:1 ratio (not including the PSU), I was able to see that the airflow through the video card actually reversed when using negative pressure causing an immediate increase in video card temperature.

For everything else in the computer, I can see both sides of the case for positive or negative pressure. From the video card perspective, there is only one answer....positive pressure.

I hope this helps : )

 

[Toni_o]

Hi guys, As hardcore gamer and having my pc in het livingroom connected to my plasma i had to figure a way out to cool good when im gaming and is silent when i wanna see a movie. So what is did.....(with great results)
1. bought a ati 4850 passive cooled pci-e card
2. turned both my 12 cm case fan towards the outside of the case. So that they just suck out all air. And these fans can be manually switched.
3. taped all the air-'holes' in my case except for the second pci slot opening (which is needed by my graphicscard) at the back of my Antac Fusion Case.

What happens is that now these fan are sucking all air out out the case. Due to the narrow inlet directly besides my GPU the airflow there is higher and thus colder. Without the fans on during gaming i measure sometimes up to 70 degrees Celcius. With the fans on i can reduce it to 35 degrees celcius.

The nice this is....when i switch of my fans there is only the cpu cooler running and the rest is nice and silent.

Just a great solution.....try it!

 

[Lazy-B]

Hi all, I'm looking for a case to finally replace my six year old Pentium 4 system and have been doing a bit of research. I want capable, cool, quiet, and clean. For me this means a full tower E-ATX, with more than adequate air flow (the cooling part) and positive pressure (the clean part). There does seem to be a few full tower positive pressure cases out there, but it seems easier to find the negative pressure designs. My thinking is the same as a few others have mentioned. That is that filtered air intakes with higher intake flow than exhaust flow works great for dust free rooms, so why not for computer cases? Silverstone has a nice simple illustrated page that explains it for me...

http://www.silverstonetek.com.tw/tech/wh_positive.php

and HardwareCanucks have a neat little review on Silverstone's venture into positive pressure design.

http://www.*****/forum/hardware-canucks-reviews/12201-silverstone-fortress-ft01-mid-tower-case-review.html

Thanks for the great discussion.

 

[Motopsychojdn]

This is a well read and answered thread, but I'm new to building and these forums so this is a simpletons view , my story is..
I had one 80mm exh built in to side panel opp. cpu and the cpu fan itself when it arrived, since buying I've added 120mm exh at rear, 2x60mm ram coolers, one pci slot exhaust under gfx card which also has onboard 50mm, I had planned on another 120mm exh on top at some point and 2x80mm at front bottom of the case as intakes but I run at 20-30'C as it is so it would be overkill tbh, Now given that most if not all cases arent airtight, any extra air in the case WILL find a way out,especially if its hot and in a state of agitated expansion, whereas a negative pressure would try to balance itself out by finding more air from somewhere, ie it would suck air in through the same gaps, in a general day to day situation, physics will balance this out whatever way you choose, if your case is in fact airtight, things would then need looking at in more detail.
Hth,
Moto

 

[corniger]

I have a Big Tower case with 2 side panel intakes and 2 large exhausts in the rear (one behind the CPU, one above the power supply to blow all the hot air out that passes by the CPU), so it's about balanced.
But the difference I have here: super silent HD coolers that "produce" a great deal of cool air right inside the system. So I can't really tell how the pressure inside is behaving. Would be interesting to determine, maybe with coloured smoke....

 

[Motopsychojdn]

I think negative for the reasons stated above, My case is a Tsunami Dream with 1 120mm intake, 1 120mm exhaust as well as a 90mm side exhaust and the power supply as an exhaust.

I was also thinking about adding a PCI slot cooler, but wanted a couple of opinions on how effective they are...

I run a slot fan directly under my graphics cards fan, I dont O/c to any mad degree, but the case is always nice and cool, I would say get one mate, a little extra exhaust cant hurt you any
Moto

 

[SchoolDaGeek]

I once had so much negative air flow out of the case that it was sucking air through my keyboard!

Seriously though, dust accumulation is a direct result of static electricity on the components. Moving air causes static electricity, just as any moving part does, such as fan motors and blades. That is why the dust sticks to them the most. Materials composition has alot to do with it as well. Metal parts will not collect as much dust as plastic parts will. Isolated metal parts will collect more dust than grounded metal parts.

Most of our typical tower cases have the inlet on the front bottom and the outlet at the power supply at the rear top. Just as overall PSI is not going to change from 14.7PSI at sea level, etc. the theory at least is a basic premise, cold air sinks, hot air rises. The other thing to consider is the slight increase the power supply has to put out to power so many fans, some usually being very close to the power draw of a hard drive. The more electricity that flows the more static charge will be built up throughout the case, as well as the higher the power supply has to work to power several additional fans.

I have found throughout my 10 years building computers that an ounce of prevention goes a long way.

A computer that only has the power supply as the only fan in the case is always going to suffer. The Dell plastic shroud over the CPU heatsink does take the hottest component and vent it right into the PSU inlet, but I have seen Dells that over time have so much debris caked around that shroud and the fins of the CPU.

Here is what I do for a normal computer NOT overclocked:

I get some anti-static spray from Radio Shack (or otherwise) and treat all the plastic pieces to a nice dose of it. This includes any fan blades front and back. Let dry. I also spray the fins of the heatsink of any component that has one, north bridge, south bridge, GPU, CPU. Let dry. If I am worried about the service environment or the cleanliness habits of the house the computer will live in (such as one with pets) I will sometimes also drill a hole in one of the CPU fins and GROUND the metal heatsink to the chassis or motherboard screws with a small solid copper wire. This helps with static buildup.

Lastly, I use simple scotch tape and tape the four corner edges of the CDROM and floppy. Both these items again have exposed moving parts and will attract dust by themselves even if there was zero airflow. If the design of the CDROM is good, you can tape off alot of its back panel as well, such as over top of the jumpers or audio connection and around the interface plug. Then spray these items with the anti-static spray. Spray down all the jackets of your IDE/SATA connectors and power supply wires. Now assemble your computer. Use mini zip ties and take up any loose slack in your wires.

The FINAL component after all this is done is to worry about balanced airflow. If you have a power supply that has a 120mm fan, you should have a case that has has a spot for a 120mm fan in the bottom front. If it is 1 or 2 80mm fans on the power supply, you should have the same amount of 80mm fans blowing INTO the case, again, usually at the front bottom or one of them at the side pointed as has been mentioned directly at the CPU or GPU.

The fan at the front bottom of the case should always have some kind of filter in front of the fan. I like to use the thin spongies that you normally get when you purchase your PCI cards and the like. I either remove the front cover of the tower and place them over the fan inlet, then use a piece of heavier tape or something to secure it in place, then replace the front cover and allow the cover to hold it permanently in place through pressure. Or if the face plate is too difficult to remove, I choose the largest hole usually at the very bottom, and stuff a folded piece of spongie in there and let it unfold inside where it filters the largest opening.

By doing all this you basically achieve the closest you can to a filtered windtunnel effect of 'neutral' airflow in CFM (you can match up the fans exactly if your PSU advertises its CFM) and the air will be forced in cool, still be mixed around due to the HDD placement, and tied up wires, but will be repelled at the static electricty level for a very long time until the spray wears off, and then vented across the CPU and into the PSU to exit (the hottest two components) and out.

In rare cases where there is either a large fan opening at the rear OR a very hot running GPU, I will either close off the fan opening with solid tape or another spongie, or at the very least leave a PCI bracket open below the GPU exposing it to outside air. If the case and power supply are mismatched, meaning you have a 120mm fan on your power supply and only a single 80mm fan opening on the case, then I will buy a PCI slot cooler and place it below the GPU passive heatsink.

Good luck everyone!!!

 

[astrodudepsu]

I feel educated. Thank you to all previous posters.

 

[Wizardsblade]

I vote positive.

Truth of the matter is that sinces cases will be almost neutral, as long as you have a good air flow path it is more important to have higher total fan CFM then if it is pos or neg. I would rather have 3+ and 1- or 3- and 1+ (dependant on where case vent are if any) with a good air flow path then have 2 and 2 as the 3 and 1 combination will give 50% more fan CFM.

 

[Floatsup]

My old rig blew, and I had to replace it. I am all in favour for home systems to be positive for dust.
Especially if you have pets as I do.
I used an existing case and monitor, and HD. I also had a 550W watt PSU. But for about $450 I think I did fairly well.

Heres the Rig
http://secure.newegg.com/WishList/PublicWishDetail.aspx?WishListNumber=13826486

One more note.
Spend the extra bucks on good heatsinks, and Thermal compunds. I also modified my case, and have 4 fans. The air flows from front to back.

This is the back side.
I took the PSU that blew and used a fan, alwas canibalize what u can.
If it means modification, then thats what your puter does. also I'm not a big fan of that phrase
heres the back side, witha little electrical tape can do

This is with everything in it.

Be aware of the size of your case. If you notice that video card, its huge. Make sure you have a descent size case to start off with.

First system test

I put a non matallic screen over the case opening, and this is the finished product. Eventually I may replace the fan with an led one. Dont really care about the glitter effect.

The bottom line, under full load, Im running 32c on MOBO, get this chit, 30c on the CPU, and GPU an astounding 46c.

I havent over clocked this rig yet. But I also don't have a reason to right now.
That my friends is COOL RUNNINGS.

On a side note I have an elcheapo PSU. I bought it for 15 bucks. Its a 550Watt with a single +12volt rail. 28 amps. (light in my book)

Knock on wood, my last one was an elcheapo, and lasted 3 years. It did exactly what I asked of it. This one is rated at 550, but most 400-450 will undermine this. So basically I am pleased with the internals of this model of VIOTEK, Very steady outputs, but again only (x1) 12volt rail. Again if this lasts 3 years, and takes my system out, well it did its job. Then I may blow another $400 for a system to get me through 3-4 years. I am not a rich man, so I try to plan on these events.

If all works out well though, I may wind up replacing it at the end of the year.
I dont like the MTBF report.

I must admit though. for my needs, Im under-rated on the 550, and the components looks very well manufactured and placed together.
I have about 25 years in systems, so I take my experience on these matters from first hand knowledge.
My first system was an 8086 back in the day.


Also this is a great thread, and I'm glad someone revived it. It should be used as a guide for any newcomer to the world building their rig. First the case should be highly considered the starting point. I have never done any major overclocking, and I would think that if thats your thing, then your cooling with pumps ( liquid ).

The only game I play is Aces High 2. I do a little video editing but non-commercial.
I also build some web sites with some graphix programs, again nothing commercial.
Heres a link to our website if you want to check it

http://www.precisionsquad.org

Stop by the guestbook if you dig it.

Cheers


OBTW.

Made a change 2 months later
I also Just cutdown 5 degrees C off my North Bridge by adding an addtional fan.
I happen to have one around, and the unit plugged into a 3 pin connector on the MoBo.
Figured what the hell. It does not do anything other then cool.

Took 30 seconds to wire tie it to the cpu cooler.

Well I got my new PSU today. Installed it, and very pleased. Nov-01-10

Done

I also Installed a new 160gig SATA drive for Ubuntu 10.10. I like that operating system. Now I cleaned out windows just for my gaming.
ubuntu is now a browser/video veiwer/and web development platform.

 

[techguygary]

As long as the fans are cycling the air in the cabinet at least 1 time a minute
it shouldnt matter if it blows in or blows out !!

 

[joedillon]

Hello,
The dust that accummulates in house on blinds etc is this due to gravity - air flow and or positive negative ? Any smarts out there ? THANKS

 

[Guest]

Hi, I just spent the last hour and a half reading this whole thread. Someone mentioned thier PS and Video card in the Negitive pressure theroy. With neg pressure these components are getting less flow. As most powersupplies are the least ventalated, have the lowest flow through them you could even risk contamination of your case with the hot air within. The same goes for the GPU. The air may have the "Capacity" to get hotter when Dence, (Trapped Hot Chambers) but with out flow of air, well, things will warm up. Its all about flow, faster flow=cooler temps. I believe having a over all neutral pressure equal to your atmosphere pressure (1atm) would work better than neg press. Thinking how components are tested when designed, This seems resonable.

 

[Guest]

Because remember, While Colder air is more dence, It also has the capacity to cool more, Theres simply more there! Once it gets hot it becomes less dence, Make sense?

 

[jmak49]

I have 5 (!) HDDs in my tower.

One 12´´ fan and two 8´´ fans blowing in
( incl. one fan in side panel blowing directly over CPU ) and
two 8´´ fans + PSU fan blowing out. With this setup I get an average low 30 Celsus. It´s a setup with by far lowest temperature among my configurations.

I have tried different setups also with the very same fans but with much higher tempeatures as a result.
I get then about 45 C as an average of all readings (HD, case core etc) .
Even sporadic lockups.
It was enough to make a lower rear fan to blow "in" instead of "out" - everything else was the same. And even other configurations.

Another, more theoretical thought - I´m not so sure if fans in the lower front of the case you positivly think suck air from outside and make your case a "high pressure case" are really very effective in doing so. At least you must really take care to seal it egainst vent. Add then you have air resistance in the vent.

Josef

 

[waffleboi9]

Hey All, found this thread through Google and have read through it all.

Good stuff all around, some very intelligent discussion here.

I have but one tiny question that I'm not sure about. Here's my situation

2x120mm intake fans on the bottom front, sucking cold air in over the hard drives
2x120mm exhaust fans on the top and back, blowing hot air out

So far, we have an approximately even flow of air, right?

Well, I have one more side panel 120mm fan, positioned above the CPU. Should I make it an intake or exhaust?

Intake, would be pulling cold air directly onto the CPU.
Exhaust, would be blowing out the heat generated by the CPU.

I'm stumped. Advice? I think I'm going to try both, run load testing, and measure the results.

 

[Floatsup]

Hey All, found this thread through Google and have read through it all.

Good stuff all around, some very intelligent discussion here.

I have but one tiny question that I'm not sure about. Here's my situation

2x120mm intake fans on the bottom front, sucking cold air in over the hard drives
2x120mm exhaust fans on the top and back, blowing hot air out

So far, we have an approximately even flow of air, right?

Well, I have one more side panel 120mm fan, positioned above the CPU. Should I make it an intake or exhaust?

Intake, would be pulling cold air directly onto the CPU.
Exhaust, would be blowing out the heat generated by the CPU.

I'm stumped. Advice? I think I'm going to try both, run load testing, and measure the results.

Im a positive pressure person myself. I would make it an intake with a filter. Here some nice filters too you can use.
http://www.performance-pcs.com/catalog/index.php?main_page=product_info&cPath=36_463_500_466&products_id=25562&zenid=dadddb0c534d299e4c590506ac7f8de3

Also if can get a tunnel for your fan to blow onto your cpu, that channels air directly in thats even better.

G/L Cheers