Question Top Case Fans Realization

ToolPunkXS

Commendable
May 10, 2021
13
1
1,515
So I know this could sound backwards at this point now that I am realizing some things... It wasn't until a couple builds ago that I had PC cases that had top fan mounts. So personally my thoughts were always air in and "out the back".... So i mounted the fans on these new cases at the top, to pull air "in" along with the front to exhaust out the back.

Just now though, as I was planning another PC build, I was questioning this. Heat rises, yes? With x2 front and x2 top fans pushing in and down, am I trapping heat towards the bottom and slowly pulling it up the back side with the x1 rear fan exhausting?

Then I wondered.... If heat rises and I have x2 top fans, should I completely exhaust out the top and have my rear fan pulling in also?

What is the ideal "in and out" for x2 front, x2 top and x1 back?
 

Phaaze88

Titan
Ambassador
Heat rises, yes?
PASSIVELY. Passive cooling arrangements.
It doesn't take much force, like a low rpm case fan, to completely overpower that natural motion. There are other reasons top case intake doesn't work well. The ones where it does work are hardware setups that aren't following the current trends.


What is the ideal "in and out" for x2 front, x2 top and x1 back?
You have to figure that one out yourself by testing multiple configs and comparing the differences.
 

Paperdoc

Polypheme
Ambassador
Ignore the old rule of hot air rises. While true, it really applies ONLY when there are NO other forces moving air. In a computer case with decent fans (such as yours) the fan action has vastly more impact on air movement than air temperature. Concentrate instead on maximizing air flow.

Your should have a near-balance of fans that force air flow in and ones for out. Ideally (in my opinion) there should be s slight excess of intake capacity so that the net result is that air pressure inside your case is slightly higher than outside. That means that, at any cracks where air can leak, it will leak OUT and NOT allow unfiltered room air to leak in and carry dust. As a FIRST approximation you can use the number of fans and the rated air flow maximum capacity of each to estimate this balance of POTENTIAL air flow. But that is never the final answer. For one thing, any INTAKE fans (like, at the front) should have dust filters in front of them, and these reduce the actual air intake capacity somewhat even when they are kept clean. IF you are using a AIO liquid-cooled system where the rad and its fans are mounted in the front to act as air intakes, the rad also reduces those fans' air flow. (You do NOT indicate you are doing that.)

In your case of planned 2 front, 2 top and one rear, achieving a balance as I suggest is difficult. As others have said, the best arrangement would be two front as intake, and two top plus one rear as exhaust. IF you could add one more front intake fan that would make things easier, but for now I'll assume you will not do this. I assume also that all five fans are identical, although you might be able to use larger fans (140 mm instead of 120 mm) in the front intake locations for more intake capacity.

This arrangement would give you good overall air flow, but maybe not the pressure balance I would prefer. You may feel that is sufficient. IF you want to try for my ideal of slightly positive interior pressure, there is a way. First, you need to split the fans into two groups for separate control. Connect both front intake fans to one mobo SYS_FAN header, and the two top plus one rear fans to another header. Then you go into BIOS Setup for those fan headers. MOST have options to use the default standard Fan Profile, which is how the header decides what speed to run the its fans according to the temperature on the mobo. Use that for the intake fans. But another option is to set your own custom "fan curve" of speed versus temperature, and you can use that feature for the exhaust fans. Then you set those exhaust units to run a bit slower than the intakes for all temperatures.

How much slower? That takes experimentation - you can NOT calculate and predict this. After your system is set up and working, get a small source of smoke like a cigarette or an incense stick. With the system running a low workload, go around the outside of the case near any small slits or cracks where air can flow through. Put the smoke source near the crack and observe its motion. What you are aiming for is slow flow AWAY from the case. If it is flowing into the case, then you need to slow down the exhaust fans more. If it flows VERY fast either way, the fan speed needs adjusting. After you have that adjustment made, do it over again with a high workload on the system. Over several tries you can adjust the exhaust fan "curve" to create modest outflow of air at case cracks for all workloads.

Just a note of caution that may NOT apply to you. In many cases with lots of fan mounting spots there are positions at the top of the front panel, and at the front of the top panel. These two combined can create a little pocket right adjacent to each other. If you mount fans in both of those, the air flowing in from the front top gets pulled out immediately by the exhaust fan in the top panel at the front. That air then has no benefit for case cooling. In that particular arrangement it is best NOT to mount an exhaust fan at the front of the top panel.