Discussion on Positive Case pressure

[pHaestus]

Several people have contacted me in one form or another regarding my comments on the benefits (or lack thereof) of lots of intake fans.

My intention was to point out that care and thought regarding airflow was important and that the whole positive vs. negative was not that big an issue. Plenty of cases and OEMs have been designed for negative pressure and to just place fans for the purpose of having positive pressure may harm cooling rather than improve it. Perhaps this wasn't stated clearly, and if so I figured that it was only fair to give others a chance to correct it.

So here is an area where everyone on any side of this issue can feel free to discuss it.

[beav]

Well, i don't have any first hand experience, but when i was visiting my brother, i stopped by a procooling regular zoson's place. He showed me that when he reduced the speeds of his front intakes, his temps dropped a good 2-3 degrees.

Quite interesting if you ask me. I've always believed that more exhaust was better, but people kept telling me to use intakes for positive pressure. Quite a shacky topic if you ask me.

[mkosem]

on my old antec SX1240 I had very low case pressure. this was done by making 2 120mm blowholes with 5 volted H1a's and puting pcpower and cooling silencer fans in the power supply and both rear exhausts. the only fans I had blowing in were a 120mm M1a on my heatercore and a 80mm L1a in front of my HD. When I had that case I arranged the fans in that fassion after days of trials with different suck/blow combinations of fans. I am a believer that low case pressure is best for cooling.

--Matt

[WebMasta33]

Well...

Let me toss in my 2 cents.

Having positive case pressure, keeps more air in the case. So of course, it's going to cool less, because the air lingers in the case longer, and can transfer the heat to the incoming air before it leaves the case.

Negative case preassure, keeps the flow in check. You always have flow, because the airpreassure is forcing air in the case, and the fans are always pulling it out.

I think my case, is perfect... because I can tell you right now, it has no positive or negative preasure. I have 2 80mm exhaust in the top back, the psu, and then a 120mm 5volt Pan H1A exaust pulling through BIX, and a 120mm H1A 5volt BIX blow in the front. But I also, have a grate on my case, that's about 1sq foot in area. This pretty much eliviates any preasure problems that may or may not occur.

I feel that the more holes you can just put in the case, and just have fans blowing on the things that absoluty need it, and have every other fan removing air from the case is a good thing.

[bigben2k]

I think that this whole case pressure issue is a moot point: the pressure inside the case, wether its positive or negative really can't be that far off from atmospheric pressure, including its own variances which should make this whole issue a waste of time.

If anything, I would argue that humidity has a larger impact.

I've seen the articles about proper case ventilation, how the air should come from the bottom, at the front, and exit at the top, in the back, but I have yet to see this explained with the side holes/fans/ducts, which can direct cold air directly above a CPU/GPU, which seems to have a much more beneficial effect than anything else.

For what it's worth, I'll back the negative air pressure recomendations, but I'll also support, to a greater extent, the use of side fans, and proper climate control inside the room.

[Joe]

I was trying to cool down a server I had, and found something very interesting.

with a Ballanced 1 80mm in the front and 1 80 mm in the back I was able to keep a Dual P3 450server with 5 HD's, 1 CDROM, 1 DLT cooled. if I added ADDITIONAL intake, it almost cooked teh DLT! cause now the heat that the DLT generates was being blown back into itself and through the drive. Almost completely toasted the drive.

the case was a AOpen SV320.

[Heavy_Equipment]

If I have to use air cooling...

I have an old ATX case here, that is tiny and cramped. In this, I ran a P3B-F with a Celeron 400 (6 x 66MHz fsb) on a slotket adapter.

I couldn't get any more than 500 MHz (6 x 83MHz fsb) out of this chip, with an 80mm intake in the bottom front of the case. (No temp data...adapter didn't have a probe)
In an era of 75%+ OC's, I knew there was more, even though the 400's weren't as good as the 300a's.

I cut a blow hole into the case, directly above the slot, and moved the 80mm case fan to the top as exhaust, with no intake.

600 MHz on the first boot, and rock solid for 2+ years @ 100 MHz.

Negative pressure has my vote...exhaust the heat, physics will take care of the rest.

Ok, low wattage example...but I haven't hardly used a case since then, let alone air.

[zoson]

I'm going to break this down into very very simple terms, using only one equation. This may not be true for all setups, but it is an actual law of physics/thermodynamics. PV=nRT where P is pressure, V is volume, n is the number of moles of substance, R is a constant, and T is Temperature.
From this equation we can see that temperature is directly related, rather than inversely related to pressure. Thus, as you lower pressure, you will have a temperature decrease as well. Keep in mind that this IS only to a certain point, as if you have no air in your case your stuff will blow up in your face. :eek:
As beav stated above, I have demonstrated to him this effect. BTW beav, isn't seeing it yourself first hand experience???:confused:
Here is my procedure:
2x 80mm 50cfm ADDA High fans intake
2x 80mm 50cfm ADDA High fans exhaust
1x 120mm 108cfm ADDA High fan blowhole(exhaust)
The fans are controlled by a rheobus that can vary the fans from 0v to full 12v.
I am using a DD5 to measure ambient temperature at various locations within my case (immediately behind the intake fans, at the bottom rear of the case, and at the socket).
I am using the 8k3a internal diode reading of my CPU.
The computer had been running for several hours before hand with all the fans on full. I would turn the two intake fans all the way off with the rheobus, and immediately the temperature of the air behind the fans would drop steadily, leveling out at about 3-4C below what it had previously been at. The effect would then spread slowly through the rest of my case, with an end result of all locations where the temp probes were being about 2-3C lower temps.
I would like to note that the probes were all in large open spaces, and not between my pci cards etc. I found that if I left my front case fans off, my sound card would overheat and produce crackling, and eventually no sound at all. This was remedied by turning the fans back on.

So the moral of this story? negative case pressure is good - to an extent. You have to have air in the case for your cooling to work, but having low pressure is good. :)
-Zoson

[WebMasta33]

The reason why when you lower the preassure you lower the temperature is because you lowering the preassure is displacing the molecules farther apart, so you reduce the interaction, which reduces the number of molecule collisions, which reduces kinetic energy, which reduces heat...

[pHaestus]

I think you guys are misapplying the ideal gas law (and a few other ones as well). The concept of "case pressure" is a definition for discussing the relative amount of intake or exhaust IMO and so consequences of gas law aren't really relevant. I dont think delta P is very large regardless of the ratio of intake to exhaust and I don't think you are generating enough pressure to compress gases (a requirement of increasing n). Before that happened, the fan would just slow down due to the increased resistance.

[Heavy_Equipment]

Heh heh, we're not creating or reducing pressure here, we're controlling the draft.
That would be cool, but case fans aren't that good.:D

Ok, some pressure change is present, otherwise a fan wouldn't work, but it can be measured in "ths" I'm sure. .000...

We are just controlling whether the air is forced in, to find it's own path of escape, or the opposite.

IMO, the reason I saw success, (temps would have been nice to quote) was the fact that with the exhaust directly over the cpu, I reduced the "hot ceiling" effect in my case. Considering that the cpu sat in the top 3" of that case, and the psu was mounted vertically in front of the cpu, evacuating this area could only help.

I don't think you can argue push/pull is better...efficiency is key, and the method that produces the path of least resistance is the setup that will work the best.

IMO, "pull" just happens to be the simple way to lower that resistance.

[zoson]

Does some1 then want to explain the results that beav and I saw when we were testing w/ my rheobus?
-Zoson

[Heavy_Equipment]

I wasn't touching that one...

notice the liberal use of IMO in my reply. ;)