The new forums will be named Coin Return (based on the most recent vote)! You can check on the status and timeline of the transition to the new forums here.
The Guiding Principles and New Rules document is now in effect.
My new PC's case has a vent at the front set up for a fan so I'm going to get a kick ass one with LEDs in it or something. My question is, should I have this pulling air in to travel through the case and out the back or just have it sucking out air like the other case fans do?
Have it sucking air into the case. That way, you form a cross-draft between air rushing into the case through the front, and air rushing out the back, pushing the air in the same direction both ways. Also, this will let you cool the computer with (presumably) cold air very rapidly, instead of just push hot air out in both directions very slowly.
As has already been said, pulling, specifically because PSU fans almost always exhaust. You don't want to have stagnancy in a case by having all fans in or out. Especially out because that generates lower air pressure in the case, and with air molecules farther apart they hit the heatsinks less often, leaving more heat in the sinks and the chips they're cooling. Consistent air flow gets the heat out of the sinks.
One ceveat: you want more air being sucked out of the case than blown in. This will prevent dead spots from forming. You want to keep intake and exhaust fans roughly equal, but slightly in favour of the exhaust fans.
with air molecules farther apart they hit the heatsinks less often
I'd be amazed if the fewer number of molecules caused an appreciable difference in temperature. What's far more important is pulling the hot air out of the case once it's been heated.
It's basic physics, in order to heat the air there has to be air to heat. Less air means less heat gets from the metal to the air. No imbalance of airflow is useful, because it either prevents hot air from leaving (as you worry about) or it prevents heat from getting into the air due to low pressure equalling low density within the volume. You have to think about air as if it were water, it wouldn't matter how quickly you sucked hot water out of a system if it were only a trickle coming in because that trickle simply can't hold as much heat as a greater mass of water. You always want the most air possible (greatest density/pressure) moving as quickly as possible being equally pushed and pulled (or more specifically, equal mass in, equal mass out per increment of time). Which, uh, ironically makes your point somewhat valid after all for different reasons, namely because the nature of drag and such means you need (slightly) more suction at the exhaust end than at the intake in order to keep the system equal.
I've designed some pretty advanced cooling solutions for servers to be deployed in enterprise class environments, and I consider myself something of a connoiseur of heatsinks and air cooling design. Not that it particularly matters for Joe Gamer's whitebox desktop. So long as there's even stock cooling that'll be enough, but I'm a perfectionist.
If you need absolutely every last bit of performance out of your system, then you want to go with what Philodox and ElectricTurtle have recommended (slightly more exhaust than supply).
However, if you want a low maintenance system, you want a bit more supply than exhaust. The reasoning being that with a negative pressure case, air (and dust) will infiltrate through any part of the case that isn't sealed (like your optical drives). If you run a slight positive pressure, you can put filters on your intake fans and keep the case practically dust free.
If your computer resides in a non-dusty environment, then it isn't much effort to blow it out every couple of months, but if you have carpet, pets, etc, you will need to clean your case weekly to maintain any performance advantage that an unfiltered negative pressure case has over a filtered positive pressure case.
Regarding poor performance due to low air pressure in the case, you won't be able to generate a pressure drop large enough to cause any significant heat transfer inefficiency. Dust build-up will have a much greater effect on the performance of your cooling system (provided you aren't running a literally airtight box).
My old machine runs hot because it's a dual PIII processor machine. I have lots and lots of airflow in that case. I think there are 6 fans in there currently, and the exhaust air is noticeably warmer than room air to the touch. The problem? Fucking thing sounds like a jet. Beware "high speed" fans, go for low-noise.
What kind of heatsinks you got up ins that motha? Most of the heatsinks made for the P3s were pretty shitty... not unlike P3s in general. Oh crap is my fanatical AMD loyalism showing?
(I once built a personal desktop with 13 fans. My last desktop had TWO Vantec tornados on top of ONE heatsink for the CPU moving over a hundred cfm. Believe me, you don't know loud. Never bothered me though, because I think people who prefer quiet to performance are puuuussieeeez. My next system will make all those that came before it look weaksauce in the cooling dept. MUHAHAHAHA I can't wait to soup up some macsteks... mmmmmm. Although I'll probably include some switches just in case people whine about it at LANs.)
DEFINITELY pulling. I used to not have a fan in the front, and my HDDs would overheat and puke. Now I've got an 80mm quiet fan (blue LED for that nifty effect) in front and an 120mm HO fan in the back and my machine has really never been happier. I can run my 2 SATAs and 3 IDEs and not have any heat issues. So definitely intake and definitely do! I don't understand now how computers don't come with fans up front. It's such a big help.
You can make your own fan filter out of those Swiffer pads that they sell for those broom things. You cut it out in the size of your fan and stick it in your case. Saves you from a lot of dust.
Posts
Also, reconsider the kick ass LED if you sleep in the same room as your PC, it's really quite annoying.
I'd be amazed if the fewer number of molecules caused an appreciable difference in temperature. What's far more important is pulling the hot air out of the case once it's been heated.
I've designed some pretty advanced cooling solutions for servers to be deployed in enterprise class environments, and I consider myself something of a connoiseur of heatsinks and air cooling design. Not that it particularly matters for Joe Gamer's whitebox desktop. So long as there's even stock cooling that'll be enough, but I'm a perfectionist.
However, if you want a low maintenance system, you want a bit more supply than exhaust. The reasoning being that with a negative pressure case, air (and dust) will infiltrate through any part of the case that isn't sealed (like your optical drives). If you run a slight positive pressure, you can put filters on your intake fans and keep the case practically dust free.
If your computer resides in a non-dusty environment, then it isn't much effort to blow it out every couple of months, but if you have carpet, pets, etc, you will need to clean your case weekly to maintain any performance advantage that an unfiltered negative pressure case has over a filtered positive pressure case.
Regarding poor performance due to low air pressure in the case, you won't be able to generate a pressure drop large enough to cause any significant heat transfer inefficiency. Dust build-up will have a much greater effect on the performance of your cooling system (provided you aren't running a literally airtight box).
Old 10KRPM SCSI drives don't help.
(I once built a personal desktop with 13 fans. My last desktop had TWO Vantec tornados on top of ONE heatsink for the CPU moving over a hundred cfm. Believe me, you don't know loud. Never bothered me though, because I think people who prefer quiet to performance are puuuussieeeez. My next system will make all those that came before it look weaksauce in the cooling dept. MUHAHAHAHA I can't wait to soup up some macsteks... mmmmmm. Although I'll probably include some switches just in case people whine about it at LANs.)