Case fans, Positive or negative pressure?


Jan 1, 2017
So I wonder if u should have positive or negative pressure in the front of the case?

Top and back also ofc! xD
From cooling perspective, positive or negative pressure doesn't matter... neither is somewhat of a problem in that it does not allow you take advantage of two of your case's best cooling features:

a) The rear grille
b) The vented slot covers

1. From a practical standpoint, 99% of the time a well designed case will have

a) Dust filters @ front openings
b) Dust filters @ top openings
c) Dust filters @ bottom openings

So it make sense to use them.

2. Also, from a practical standpoint, you want the heat blowing away from you (out the rear and maybe top) rather than at you and "in your face."

3. On a well designed case, all dust filters are removable; the top one however is usually somewhat different as the grille and frame remains while the filter itself can be taken out ... and left out.

a) In an air cooling build, if the numbers work, (and often they do not) you want these top fans blowing out. You don't want air filters there if they are exhaust fans so remove it and leave it out.
b) In an a water cooling build, you want these fans blowing in (no exceptions). You want air filters there as they are intake fans.

4. You must always remember that air filters cause a restriction which can reduce air flow by as much as a third. While most would assume that 3 fans in and 2 out (all identical) would result in positive air pressure, the truth is you are closer to neither positive or negative ... 3 intakes with a 1/3 restriction of air flow means you only have the effective flow of 2 unrestricted fans which matches your 2 exhaust unrestricted fans.

5. Dust is the most commonly given reason for maintaining positive case pressure. And while that is certainly an issue, to my mind the more important issue is what else is coming in the the dust ? Let's take a look at the very common scenario below

(2) front fans blowing in
(2) top fans on a CLC blowing out
(1) rear fan blowing out

Let's "do the math" where EF is the fan equivalent of one fan with no restriction

Intakes: 2 front fans - 1/3 restriction for filter = 1.33 EF
Exhaust: 2 top fans + 1 rear fan = 3.0 EF

You have 2.25 times more air being sucked out than pushed in.... meaning you have to suck in 1.67 EFs worth of air thru the rear case grilles. The logic applied to the CLC blowing out concept is that the user wants the hot air outside the case .... but the reality is that this is not happening as common sense might dictate because various factors are not being considered. The CLC will remove the 75 - 125 watts of CPU heat, but cooling of the CPU is nerfed by using pre-heated case air to cool the CPU as opposed to cool ambient air from outside the case.

In addition, you have the heat exhaust from that 650 watt PSU and that 250 watt GFX card that was exhausted out the back being sucked right back into the case thru the rear case and slot grilles along with all that dust since it gotta get in somewhere since you are exhaust 2.25 times as much air than you are getting from your intake fans.

In short:

Front = In
Bottom = In
Sides = In
Rear = Out

Top = Out on air cooled builds **if** you have positive air pressure. With 3 fans out (2 x 120mm on top + rear fan), you'll need 5 fans blowing in (5 x 2/3 = 3.33) to prevent hot PSU and GFX card exhaust from being sucked into the case thru the grilles. So say 2 front + 2 sides + 1 bottom intake and 2 top and 1 rear as exhaust

Top = In on water cooled builds. With just rear blowing out, and a 2 x 120mm CLC blowing in plus 2 front fans blowing in from front, you have 4 x 2/3 = 2.66 blowing in and 1 blowing out, leaves you will 1.33 blowing out thru rear case grilles ... simply "blowing away" that hot PSU and GFX card exhaust and keeping that heat and dust out of your case.


Jan 1, 2017

So if I got this right the fans in the front blow in,
The one in the back blow out and the one at the top blow out?

And I should also remove my dust filter at the top? because it is blowing air out?
Most of that is wrong without knowing all the parts used. You cant just say 1/3 restriction from filters. Then comes the size and CFM of the fans and not a clue if the case in question even has filters.
1. If you read the whole post, you will see that I took pains to state:

" On a well designed case..."

The words "well designed case" and "case w/o filters" are mutually exclusive. So no, by the stated definition, it's not wrong.

2. After having personally conducted days of testing, please explain why I can't say that ?

a) Again, if you read what I actually wrote ...

"air filters cause a restriction which can reduce air flow by as much as a third ". Is it your position that we would normally see more than a third, because everything from 0 to 33% in fact, fits that definition ? Now with a clean filter it's certainly less, probably on the order of 10% but that has no significance here. When we design a building, we design for the largest load and that's what I have done here. You can object all ya want saying "it's all wrong because the snow and wind loads aren't there all the time".... but the fact remains, when it is there, if I don't design for the worse case condition, then the occupants of the building die when the roof collapses.

Since doing those tests three years back we have always tested every completed build. Don't always break out the fog machine, cigarette will do but it is a simple test to undertake. Our last build that came in, the user brought in a box of parts, a Corsair Spec 01 and new red LRD fans for every fan mount

Front: (x2) 120mm
Top: (x2) 120mm
Rear: (x1) 120mm

He asked for all 5 fans to be installed and we did it with the top fans as exhausts as requested. That gave him 3 out and 2 in. So I gave him the above speech and it just didn't seem to make sense in his mind and we did the siggie test.... that smoke went right into the case. Reversed one of the top fans and did it again.... granted wouldn't actually do this for a working build because of obvious short circuiting issues but it's valid for the experiment)

Now we had 3 in and 2 out. Lit the ciggie again and now air was clearly coming out the rear grilles and a decent clip.... no doubt as filters clogged up a bit, this situation would deteriorate somewhat. Just for "schitz and giggles" we did it with 2 / 2 and again, smoke went into the case. He went home with 4 in and 1 out simply because after testing all potential possibilities, 3 in 2 out resulted in negative pressure.

b) I can say because we have tested this on our test bench and have physically measured the impact using various fan combinations using 6 thermal probes, a infrared thermometer, a fog machine and various fan combinations at varying speeds. We took out the fan filters, laid them out on newspaper, misted them with water and then took some dust from a shop vac on them (was recently used to clean whole house filter and heat exhangers for house forced air system) and let them dry. Then we used the vacuum to suck from the "clean side" at a reasonable distance and finally gave them a shake to get off the loose stuff and pulling off any dust bunnies by hand till we got a "look" similar to what we normally see in a typically user's build when brought over to get something fixed.... certainly much less than i typically see in my youngest sons build and those get cleaned every 3 months.

These were then installed in a box that has 10 fan mounts and we tested by having the fog machine discharge at the rear grille. Results w/ 140mm @ 1200 rpm:

(1) in (1) out - negative pressure fog sucked into PC
(2) in (1) out - positive pressure, no fog sucked into PC, stream pushed away
(2) in (2) out - significant negative pressure fog sucked into PC
(3) in (1) out - significant positive pressure
(3) in (2) out - neutral, barely perceptible movement in fog stream being pushed away from PC
(3) in (3) out - strong negative pressure fog sucked into PC
(4) in (1) out - very strong positive pressure, fog really getting a kick
(4) in (2) out - significant positive pressure
(4) in (3) out - slight negative pressure, some fog sucked into PC
(4) in (4) out - strong negative pressure fog sucked into PC

skipped the 5s and went to 6

(6) in (3) out - strong positive pressure
(6) in (4) out - neutral, barely perceptible movement in fog stream being pushed away from PC
(6) in (5) out - medium negative pressure significant movement in fog stream
(6) in (6) out - strong negative pressure fog sucked into PC

Went thru the same schpeil at 900 and 600 rpm.... with decreased fan speed the effect of air resistance was somewhat less. At 600 rpm for example, with (6) in (5) out, the effect was relatively neutral. At higher rpms which we didn't test, I imagine the effect of air restriction would be somewhat greater. And we did it again with 120mm fans

What fans are used is immaterial when all are the same, at the same rpm and size. As a corollary, in industrial building and kitchen design, we really don't use intake fans at all except where energy code requirements dictate using pre-heated intake air. If you exhaust air out if at space, all you need is an open window or intake louver to allow air in. Think of any window fan, kitchen hood exhaust or attic fan in your house.... you don't have a corresponding intake fan

Similar threads