Case fan Algorithm?

[daos]

is there, or does anyone know of a specific guideline to follow for case cooling when using fans? i was under the impression from an article i read about a year ago that said there was a specific equation to it. maybe im making this too complicated but i would like to know these things in as much detail as possible.

are you supposed to have a slightly negative pressure in your case to keep air flowing? or do you want a positive pressure? should i use 2 intake fans and 1 exaust, or 1 intake and 2 exaust fans? these are the kind of things im talking about.

whats the best air cooled setup for a basic case with overclocking in mind?

 

[greasepain37]

i recomend, a positive pressure. i currently have 3 intake fans only. i'm sure other will post on how this is wrong, but if you ask them i'm sure they'll say the world is still flat. my case is also dust free becouse of the positive pressure, you would also need good hepa filters for each fan.

 

[beatle]

Seems like no pressure would be ideal (equal intake and exhaust.) I'm running more exhaust than intake right now for the simple reason that I don't have enough places to put intake fans (not cutting into this one yet!)

 

[mechBgon]

I think you're supposed to do this:

• add up the thermal load (in watts) of the internal components
• figure out the total air volume of the case. An approximation will work here, it doesn't have to be exact.
• determine your general elevation. Air gets thinner as your elevation increases, so you can't just assume that everyone else's numbers are right.
• measure the vent slots/holes as precisely as practical, and calculate the percentage of the airpath that's obstructed by slats/grating.
• check the airflow versus backpressure on the fans you've considered. For instance, here are some for the Panaflo hydrodynamic-bearing 80mm models, very popular: link The graphs are on page 2.
• look at all that nonsense, say "the heck with it" and buy a well-vented case like an Antec SX835II, and spend $11 to buy a set of five NMB low-noise dual-ball-bearing 80mm fans for it (link) plus some 3-pin-to-4-pin adapters.

Done! With the five fans (three intake, two exhaust) you will have a nice crossflow with minimal noise. You can also flip the power supply over to pull stagnant air out of the top of the case: instructions

edit: the Antec PerformancePlus and PerformanceII cases have removable air filters, for what that's worth. Cuts the airflow down a little, naturally.

 

[daos]

shouldnt your PS need cool air as well to keep it cool? wouldnt that be giving the PS hot air to try and "cool" it? sounds like an interesting idea though.

that Antec case your pointed out is the exact case i have. should i put 2 intakes and 2 exausts? i hear that if the pressure is positive then staggnent air starts to build up waiting to be moved out the case and mixes hot with cold air causing friction. in turn that can obviously damage PC components. on the other hand if you have a negative pressure then it can form a vaccum and thats not good either.

im thinking that 1 intake and 2 exausts is the way to go. this way i do have air coming in, but it doesnt have enough time to sit around because the exaust is twice the intake, so the air moves quickly through the case.

anyways, any more advice is well respected from anyone who has any input. thanks.

P.S thanks for the information mechBgon, i really appreciate it buddy. im gonna definatly try the PS mod.

 

[mechBgon]

I think you're over-engineering things but for what it's worth, I have that setup and it does stay very cool in real life: Turbo lives in that case with those fans. Remember that there's a third exhaust fan as well, the one in the rear of the power supply, so this is pretty close to balanced. It's not the flashiest case around but it works for me.

edit: the PerformancePlus 880 in my rig profile is the same physical chassis as the SX835II but with an alternate power supply.

 

[greasepain37]

Originally posted by: daos
shouldnt your PS need cool air as well to keep it cool? wouldnt that be giving the PS hot air to try and "cool" it? sounds like an interesting idea though.

i hear that if the pressure is positive then staggnent air starts to build up waiting to be moved out the case and mixes hot with cold air causing friction. in turn that can obviously damage PC components. on the other hand if you have a negative pressure then it can form a vaccum and thats not good either.

more about positive pressure : description from a server site : "All of AvidHosting's custom web servers are equipped with a positive pressure filtered-air system. Four large fans pull filtered air into each server's protective case and the components within are cooled by fans which circulate this purified air. This constant introduction of clean air into the case creates a positive pressure environment ensuring dust and particles remain outside of the server." friction and staggnent air are old wives tales

 

[Drewpy]

Originally posted by: greasepain37
i hear that if the pressure is positive then staggnent air starts to build up waiting to be moved out the case and mixes hot with cold air causing friction. in turn that can obviously damage PC components. on the other hand if you have a negative pressure then it can form a vaccum and thats not good either.

more about positive pressure : description from a server site : "All of AvidHosting's custom web servers are equipped with a positive pressure filtered-air system. Four large fans pull filtered air into each server's protective case and the components within are cooled by fans which circulate this purified air. This constant introduction of clean air into the case creates a positive pressure environment ensuring dust and particles remain outside of the server." friction and staggnent air are old wives tales

You need incredibly fast airflow to create enough friction to actually get a charge in the air. Think up and down drafts in a storm cell creating lightning. There is virtually no way that the airflow through a computer case would be capable of generating the required air velocities.

If you only have one fan, use it as an exhaust at the rear of your case. Otherwise I recommend trying to balance the airflow within your case
CFM in = CFM out
or with slightly positive airflow if you can't bablance it.

 

[Noid]

I'm amazed at some of the replys. I didnt use any science, just my stupid logic (lol).

fan 1.) Suck air OUT from CPU fan from the rear case hole with a 90mm.
fan 2.) Push air IN from the side blow hole I made with a 90mm. (placed directly above the CPU / northbridge area)
fan 3.) Push air IN from the lower front of the case with a 120mm (pushed over the cards with 2 rear case slots open)
fan 4 & 5.) PS fans suck air OUT from top.

Both IN fans have mesh screens to keep the dust, dirt, and hair out. (after removing the crappy case mesh holes)
Tie all wires and cables to sides and corners.

I have more CFM in than going out, reguarding the fans. BUT, the open slots in the rear equalize any pressure.

I figure sucking the hot air out immediately, and forcing as much cool air in, is best.

I dont burn CD's or have heavy HD access. So, this works good for me.
If I did,,, I'd put a fan in FRONT of the HD's and probably put a fan ABOVE the CD burner.

 

[BalAtWork]

Let me give you a little advice from a ChemE.

Heat transfer does not exist in a void. Although you cannot create a "void" you can decrese the pressure in the case, therfore decreasing the effective heat transfer that can take place. Remember, a decrease in pressure basically means less matter in a given area as compared to the standard. Hence less matter to absorb the heat.

BUT, positive pressure is also bad. Then you have a decrease in "new" air into the case. Hence the equilibrium that is formed causes the case temperature to be higher because you saturate the air with heat before it can escape.

The perfect scenerio would be perfectly balanced system. But you will probably never achieve that unless you take into account every nook and cranny leaking air in/out. SOOOO aim for balance, but lean to positive to keep dust out. Case temp might be a degree or 2 higher, but dust could cause the same temperature rise once in your case and on your components by decreasing their ability to radiate heat effectively.

 

[mechBgon]

Originally posted by: BalAtWork
Let me give you a little advice from a ChemE.

Heat transfer does not exist in a void. Although you cannot create a "void" you can decrese the pressure in the case, therfore decreasing the effective heat transfer that can take place. Remember, a decrease in pressure basically means less matter in a given area as compared to the standard. Hence less matter to absorb the heat.

BUT, positive pressure is also bad. Then you have a decrease in "new" air into the case. Hence the equilibrium that is formed causes the case temperature to be higher because you saturate the air with heat before it can escape.

The perfect scenerio would be perfectly balanced system. But you will probably never achieve that unless you take into account every nook and cranny leaking air in/out. SOOOO aim for balance, but lean to positive to keep dust out. Case temp might be a degree or 2 higher, but dust could cause the same temperature rise once in your case and on your components by decreasing their ability to radiate heat effectively.

Advice from a failed ChemE student here : remember that typical mid-output 80mm case fans hit full airflow stall at just 3.5mm H2O backpressure, so it's not like they are going to be able to create a significant pressurization or vacuum. For comparison, one atmosphere is over 10000mm H20 at typical temperatures.

(edit for typo )

 

[BalAtWork]

LoL

I din't mean for anyone to actually take what I was saying at actual macrocosim levels.

I was overexagerating on purpose...could you imagine actually creating a void in your case? Be one hell of a strong case.


To put it as simple as possible...heat exchangers require moving matter (air in this case) in order to have the maximum temperature differential to remove as much heat as possible. Any scenerio besides a balanced load will eventually lead to less air flow around the heat sink and therefor higher temperatures.

Better?

 

[mechBgon]

Hehe, I know I do see people get fixated on certain things, such as case material and flow balance, beyond a reasonable level. I worry more about the flow patterns and making sure the fans are working together instead of against eachother, figuring these cases usually have enough loopholes and extra holes to balance themselves.

For instance, if I had strong fans in front of my case, and weak ones in the rear, the strong front fans can simply loop extra air up to the top of the case and down the inside of the bezel, and thus move the air they want to move without getting it from outside. And there are enough cracks and crevices in most cases to allow a decent purge stream if needed, or the opposite. Most of the optical drives I see these days have gaskets around their drawer, so it's probably not a super-big dust risk.

 

[spanner]

Computers are not really that fussy about cooling. i.e the best combination is found by trial and error. In general, just put a case fan where there is a space for one and you will be fine. exhaust fans at the back and intake at the front and sides. As for negative and positive pressure, it doesn't really matter, you just want to have your fans near hot components so that there is maximum (turbulent) flow in that area.