TR Forums

I have all the components (to be listed shortly) that I want to put in for the time being. My question is how to get a decent, expandable cooling setup that will last. Ideally, that will last even past the MB, which seems like it should be possible since I liquid cool the CPU. In the past, I've kinda half-assed it and that has kind of bitten me in the ass, although I've been getting slowly through the years.

A probably irrelevant quick note: I started putting this together last summer, and then sundry health issues intervened and I haven't had a chance to get back to it since that time; so if the parts choices seem odd for a 2014, well now you know. Where I felt it would matter (or defray embarrassment) I've tossed in a couple of notes.

Internals

Motherboard: ASUS Z-87
CPU: Intel i7-4400K 3.5 GHz ###I have never overclocked a processor, but it might be nice to have the option
CPU cooler: Corsair H105
RAM: Corsair Vengeance 16GB kit.
Graphics: EVGA GeForce GTX690
Audio: Xonar Xense
Case: Corsair 900D
PSU: Corsair AX 860i

Storage

OS: 512 GB Agility
Steam:Crucial M500 960GB
[b]Mass: 4x4TB WD Black, RAID 5
Optical: 2x LG BD rewriter
I have a great many odds and ends around from consolidating eight years worth of random single purpose boxes into a general purpose Ubuntu WS, one Fileserver/backup/highly automated home webserver, and a Raspberry Pi XMBC box. The suggestion that I am pointing all this out so that people don't accuse me of having a porn habit of frighting proportions is a foul accuracy.

Okay. To restate the question, how to I cool this monster. The 900D was serious overkill; I liked my 650D because I could get my hands in it, and if some is good, more must be better, right? but it isn't the crux of the issue. The issue is I don't really understand what I'm doing. Where do I want static pressure? Where do I want airflow? how much? (for that matter, why do I want one or the other)? For that matter, is there a good guide on here for the clueless? I was in laptops from 2001 - 2008 and I really don't feel I've ever caught up.

Positive pressure is good because you control where the air enters the case and you can filter the dust out before it comes in. If you run your case with negative pressure, the unfiltered dirty air can enter through any crack that it finds.

To expand on JAE's post, positive pressure means having more intake airflow than exhaust.

The 900D is already cooling overkill for your rig, esp with the H105 cooler. I'd just run it with the stock fans and your current setup (3x120mm intake, 140mm exhaust, PSU fan and H105).

The general suggestion is to bring cool air in the from the bottom and exhaust it out of the top. For positive pressure, make sure that there is more intake than exhaust.

This generally means mounting intake fans on the front of the case at the bottom and on the bottom of the case at the front. Exhaust fans would be mounted at the top of the case and on the back of the case at the top. In the case of liquid cooling, a radiator would be mounted on the top.

Static pressure fans are meant to be used with heatsinks.

Because everything is running at stock and you have a very large case, cooling shouldn't be an issue and you should be fine with the fans that came with the case.

i have the 900D and i can tell you its a great case - best i have ever seen/owned

as for my setup i have the front 3 fans as intake and a 360mm radiator on the top with 3X 38mm fans also as an intake
the rear fan i have as exhaust

i have 7 hard drives and 2 ssd's all in the bottom area (instead of at the front)

just about any fan setup will do as long as you have the majority of the fans as intake and imo just having the rear fan as exhaust will do, i used to have more fans in it along the bottom section but they wasnt needed and just added more noise so i removed them (none of the temps changed noticeably)

there are many options with this case as far as fans go and just about anything will do especially if you are running stock, i am running a 6 core 3930k at 5.0ghz in mine and everything is running nice and cool

i think as long as you have more intake compared to exhaust you will be fine with just about any setup

Awesome, thanks. It is really nice to know that I don't have to fuss -- I was afraid I was going to need to have days to spend fussing.

[EDIT: Something I'm not sure I understand]

Which direction do I want to move air across the radiator. I assume I want it drawing in so as to increase cooling on the CPU as well as net internal pressure.


I had kinda figured I should be good with the stock setup, but when I poked around I couldn't find a single example of anyone doing that. Granted, as I said the case is massive overkill - I kind of regret it; on the other hand at the time my immune system was well launched into its plan to destroy its greatest enemy, my joints and any other bodily systems that stood with them. Or near them. So a case that I didn't have to squeeze around tight spaces to get into had a huge appeal. Also, in the event that my home is invaded by a bevy of commandos, I can take shelter behind it, secure in the knowledge that it wills stop anything up to and possibly including a Browning .50.

I was going to put this as "finally" but I realized it is probably pretty important: how can I ensure that I have positive vs. negative pressure without doing repeated practical experiments? Is there some good way to sum airflow volumes? And while I'm thinking about it, wouldn't (in the abstract; I'm happily reassured WRT my actual build) a positive pressure setup have lower total cooling than a system with neutral pressure? (After thinking I decided that negative and positive pressures would cancel out, although that is pure intuition).

Incidentally, assuming that I do decide to overclock -- I'm not disinterested, it has just always seemed to require knowledge that I haven't wanted to accept the opportunity cost to acquire -- would should I expect to make substantial modification? Or, to approach it from the other side, what would I have to do to need to get aggressive with cooling?

And if no one minds me expanding a bit and filling some holes in my knowledge:
Why does case size make a difference? is the increase in air volume that big a deal? I mean, the specific heat of air might as well be zero, but my take away from that is that there is something that I'm missing.
What does static air pressure actually mean, vs. airflow? For that matter, what property of the fans makes them different?

Again, if this is getting out of scope, I apologize.

jangove wrote:

Is there some good way to sum airflow volumes? And while I'm thinking about it, wouldn't (in the abstract; I'm happily reassured WRT my actual build) a positive pressure setup have lower total cooling than a system with neutral pressure? (After thinking I decided that negative and positive pressures would cancel out, although that is pure intuition).

The easiest way is to look up the CFM (cubic feet per minute) of the fan. A fan's CFM is the amount of air that moves so just have more CFM directed into your case than out of it.

jangove wrote:

Incidentally, assuming that I do decide to overclock -- I'm not disinterested, it has just always seemed to require knowledge that I haven't wanted to accept the opportunity cost to acquire -- would should I expect to make substantial modification? Or, to approach it from the other side, what would I have to do to need to get aggressive with cooling?

If you're already water cooling you don't need to do anything else unless you're going for an insane overclock.

If you decide to add more fans just add intake at the bottom and on the front. Some people use a "push-pull" configuration on radiators or other heatsinks to dissipate more heat. Push-pull just means having two fans (or four) attached to the radiator/heatsink, half of which are pushing air onto the radiator/heatsink and half of which are pulling it away.

kumori wrote:

The general suggestion is to bring cool air in the from the bottom and exhaust it out of the top. For positive pressure, make sure that there is more intake than exhaust.

This is the golden rule, rather than a general suggestion;
Positive pressure increases air density, so in addition to preventing dust ingress, it actually raises the efficiency of all heatsinks slightly.

After that, you just have to decide what components you want to prioritise:

• To get the coolest graphics cards and internal temps, put the H105i radiator at an exhaust point.
• To get the best CPU overclock, have the radiator at the bottom as an intake.

Intel Core i7 4400K? Is there such an SKU?

jangove wrote:

CPU: Intel i7-4400K 3.5 GHz ###I have never overclocked a processor, but it might be nice to have the option

I'm guessing you mean 4770K.

Seriously, if you go out and get a $350 case (getting a case that is more expensive than you processor is another story) you better build a real monster of a system and overclock it to the max with a crazy custom water cooling solution... otherwise, why not just get a $150 case?

jangove wrote:

Why does case size make a difference? is the increase in air volume that big a deal?

In general, it provides a greater volume of air to blend. Unless you have a blower-style GPU, the GPU and CPU exhaust their hot air inside the case. If you've got a smaller case with poor ventilation*, the ratio of hot air (CPU+GPU) to cold air (air sitting inside the case from the intake fans) increases. An example being: adding 5oz of boiling water to 1 gallon of cold tap water (big case) or 10oz of tap water (small case). You have exhaust fans to get that air out, but the case temp will always be higher than room temp.

*Ventilation plays an important role in all this. Having a huge case doesn't give you a free pass to not install any case fans. Also, other factors such as airflow restrictions and distance from the intake fans can affect the end result. Despite it's diminutive size, the Silverstone TJ08-E achieves great cooling performance even with high power systems because it's basically a short wind tunnel. Distance from intake fans is also the main reason why installing a side panel fan is the most effective way to lower GPU temps...the fan drawing in cold air is basically touching the GPU. Additionally, hard drive bays can have a profound effect on case cooling. Some drive bays/cages are very restrictive and act as a wall behind the front intake fans.

jangove wrote:

What does static air pressure actually mean, vs. airflow? For that matter, what property of the fans makes them different?

You could write a whole series of articles on fans and airflow.

Static pressure is a measure of the fan's output; specifically, its ability to overcome resistance. Heatsinks with densely packed fins or with great depth will have more resistance to air flowing through the fin structure - these types of heatsinks need a fan with higher static pressure ratings to push the air through the fins. Hence, why static pressure ratings for fans are generally only useful for heatsink fans are rarely case fans.

In broad terms, the main characteristics which are important for fans to enthusiasts are static pressure, airflow, RPM, starting voltage and noise. Finding the right balance for a given user and fan can be tricky, and to the majority of users, isn't particularly important. Usually, it's the silence enthusiasts and OC enthusiasts who will be most interested in these characteristics, to varying degrees for different aspects.

For most users, the stock fan, motherboard and case configurations should suffice.

jangove wrote:

Why does case size make a difference? is the increase in air volume that big a deal?

A huge case is a benefit in the non steady-state condition when you first start your game. If you run your benchmark during the first couple of minutes that the computer is turned on from a cold state, a large case will still be warming up, so you'll get a better e-peen number.

At steady state, after you've been gaming for a few minutes, the size of the case becomes irrelevant. All that matters at that point is the air that is being moved into and out of the case. If you're bringing in a large mass flow rate of very cool air and exhausting that same large mass flow rate as hot air, that's what's carrying the heat away.
q^· = m^· ΔT c_p

.
q is the heat flow rate (in watts).
.
m is the mass flow rate of air (in kilograms per second).
ΔT is the change in temperature between the cold air entering and the hot air leaving (in degrees Celsius).
c_p is the heat capacity of air (about 1005 joules per kilogram per degree Celsius).

So, you bring in the air as cool as you can (usually taking it from down low at the front of the case because it's likely to be fresh cool air rather than being mixed with some hot air that was previously exhausted from the case). This gives you the best possible ΔT.

You see that the size of the case doesn't figure in... except as it can affect the mass flow rate of air. A large case may have very few obstructions to air flow inside. It may have room for more large fans and grilles for intake and exhaust. A small case may be so jammed full with components inside that the flow of air is restricted. A small case may also have fewer and/or smaller fans and grilles for intake and exhaust of air.

If your high-end overclocked gaming PC is turning 400 watts of electricity into heat, the air coming in is at 25°C and the air leaving is at 40°C, you'll need 0.0265 kg/s of air flow = 1.41 m³/min = 50 cfm. If you let the outlet temperature rise to a warm 50°C, you'll need only 31 cfm. If your outlet is a cool 30°C, you'll need 146 cfm.


That's what we see when we treat the entire PC as a system. When you look at individual components, you have to consider the conditions at their heat transfer surfaces.
q^· = U A ΔT_LM
.
q is the heat flow rate (in watts).
U is the overall heat transfer coefficient (in W/m²°C)
A is the surface area (in square meters)
ΔT_LM is the log-mean temperature difference between the hot component and the air (in degrees Celsius).

So, if we want to keep an individual component cool...
We give it a large surface area (big cooler with many fins).
We ensure that the air flow over the fins is turbulent rather than laminar so that the heat transfer coefficient is high.
...and we expose it to the coolest possible air.

If you put your CPU's water-cooling radiator on the case inlet, the CPU cooling will be great because it will see the coolest air as it enters the case. Everything else in the case will have worse cooling because it will see the air that has already been warmed by the CPU's radiator. If you put your radiator on the case outlet, the CPU cooling won't be as good, since the air to the radiator will have been warmed by the other components in the case, but all of those components will run cooler because they are supplied with cooler air. I installed my CPU cooler's radiator on the outlet.

JAE you just gave me a flashback to a Materials Engineering class I took about 10 years ago. I didn't retain too much from it. If a larger case has say 2 fans on the front and 1 fan at the top rear for exhaust will the case still have positive pressure if the entire top of the case perforated (even though there are no fans mounted on top)? My first thought is that some of the warmer air inside the case would naturally rise out of the top of the case with no fans while the rest would be exhausted out of the rear fan.

The chimney effect inside your case is much smaller than the air movement generated by the fans.

iamjsmith83 wrote:

JAE you just gave me a flashback to a Materials Engineering class I took about 10 years ago. I didn't retain too much from it. If a larger case has say 2 fans on the front and 1 fan at the top rear for exhaust will the case still have positive pressure if the entire top of the case perforated (even though there are no fans mounted on top)? My first thought is that some of the warmer air inside the case would naturally rise out of the top of the case with no fans while the rest would be exhausted out of the rear fan.

The warmer air inside the case will also be less dense than the (hopefully) cold air coming into the case, so there would still probably be a net positive pressure (mdot between intake and exhaust fans will be positive into the case, but of course in equilibrium the net mass flow is equal, which is the whole point of positive pressure - forcing air out of vents instead of drawing it into the case).

I've moved this thread to the cooling forum.

--JustAnEngineer, SBA forum moderator