Positive or Negative pressure?

The forum for non-component-related silent pc discussions.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
Skipper93
Posts: 17
Joined: Sun Jan 11, 2009 8:09 am
Location: USA

Positive or Negative pressure?

Post by Skipper93 » Sun Jan 11, 2009 10:47 am

Hello, I am new to the spcr forums.
I have just posted under Cases & Dampening asking for help on find the right case. "The Perfect Chassis.. Does it exist?" If you could help there that'd be great
Anyways, When picking the correct case do you want to create a positive or a negative case pressure by how many fans are either intake or exhaust? I have seen support from both sides and hopefully can get some helpful tips for you guys here at spcr. Thanks,

~skip

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Post by Olle P » Sun Jan 11, 2009 12:01 pm

- Positive pressure is #1 if dust is a major concern. (All filtered intake fans, no dust will enter anywhere else.)

- Negative pressure is better wrt all other aspects.

Cheers
Olle

Skipper93
Posts: 17
Joined: Sun Jan 11, 2009 8:09 am
Location: USA

ok

Post by Skipper93 » Sun Jan 11, 2009 12:49 pm

So does that mean that if a negative pressure is used that dust will increase. I imagine that in a negative pressure system that air will be forced through every little crack that is not an exhaust fan or filtered intake.

Riffer
Posts: 517
Joined: Mon Jan 06, 2003 4:14 pm
Location: Toronto, Canada

Re: ok

Post by Riffer » Sun Jan 11, 2009 1:08 pm

Skipper93 wrote:So does that mean that if a negative pressure is used that dust will increase. I imagine that in a negative pressure system that air will be forced through every little crack that is not an exhaust fan or filtered intake.
Negitive air pressure will pull air through every nook & cranny in the system, positive will push it out.

nick705
Posts: 1162
Joined: Tue Mar 23, 2004 3:26 pm
Location: UK

Post by nick705 » Tue Jan 13, 2009 2:22 am

I like the idea of positive pressure in theory, but I've never been able to make it work very well in practice... whatever the fan setup, temps always seem to be higher than in an equivalent negative pressure system.

You also need a case design which enables you to remove, clean and replace the filter material quickly and easily, or the whole concept falls flat on its face. Personally I'd rather just pop the side panel and blow out the dust once every six months or so...

L2GX
Posts: 51
Joined: Wed Jul 02, 2008 1:36 am
Location: brussels

Post by L2GX » Tue Jan 13, 2009 5:40 am

Higher pressure means a better thermic coupling imho.
Apart from that either way is way below optimum, but both are efficient enough. The biggest problem is the surplus of space in a chassis.
If your pc does not keep cool, ducts are the next best way to go. The best would be a sustained airflow along every component, using countershaped panels.
You only need a millimeter of fastmoving air to cool surfaces, anything more will just cause turbulence resulting in spotty cooling.

Hypernova
Posts: 124
Joined: Sat May 31, 2008 2:33 am
Location: Australia

Post by Hypernova » Thu Jan 15, 2009 12:05 am

nick705 wrote:I like the idea of positive pressure in theory, but I've never been able to make it work very well in practice... whatever the fan setup, temps always seem to be higher than in an equivalent negative pressure system.

You also need a case design which enables you to remove, clean and replace the filter material quickly and easily, or the whole concept falls flat on its face. Personally I'd rather just pop the side panel and blow out the dust once every six months or so...
For +ve pressure filtered set up you need to account for the resistance of the filter (huge). Which means either more fans, fan sandwiches or reduced exhaust. Personally I stuck with the compromise of higher temp at idle and low noise. Using a triple fan sandwich.

alecmg
Posts: 204
Joined: Thu Mar 13, 2008 5:56 am
Location: Estonia

Post by alecmg » Thu Jan 15, 2009 12:26 am

Depends on location of fans and vents.
For my case Arctic Silentium T2 Eco, negative pressure works marvellously. It achieves extraction of hot air, plus all open vents are located near components that need cooling. These receive cold air stream of comparable to slow fan performance.

nick705
Posts: 1162
Joined: Tue Mar 23, 2004 3:26 pm
Location: UK

Post by nick705 » Fri Jan 16, 2009 6:50 am

Hypernova wrote: For +ve pressure filtered set up you need to account for the resistance of the filter (huge).
Exactly - to combine low resistance (necessary for slow fans and low noise) with effective filtering, you need a huge filtered area, which probably requires a custom design using pleated filters.

I've looked at Bluefront's designs here with a mixture of admiration and bemusement, but I have neither the practical skills nor the patience to build something similar. Plus I don't have a dusty parrot, so my situation doesn't really warrant the effort. :)

Skipper93
Posts: 17
Joined: Sun Jan 11, 2009 8:09 am
Location: USA

YES

Post by Skipper93 » Fri Jan 16, 2009 9:59 am

Thanks for all the comments and tips, both setups sound like they are good but negative sounds lower maintenance and most cases I've looked at have more exhaust fans than intake so the choice is obvious to me :P
Thanks again,

~skip

xan_user
*Lifetime Patron*
Posts: 2269
Joined: Sun May 21, 2006 9:09 am
Location: Northern California.

Post by xan_user » Fri Jan 16, 2009 10:28 am

As usual ill be odd man out.
I run positive with no filter. It reduces dust getting in to optical and floppy drives, but doesn't eliminate it in the rest of the case.

Dust is not a big issue here, but I hate getting it in my optidrives.

NeilBlanchard
Moderator
Posts: 7681
Joined: Mon Dec 09, 2002 7:11 pm
Location: Maynard, MA, Eaarth
Contact:

Post by NeilBlanchard » Fri Jan 16, 2009 7:41 pm

Hello,

As you might imagine, this topic has come up more than a few times in the past. Here's what I wrote the last time:
Neil wrote:I'm of the opinion (based on my observations) that pulling air is much easier/more efficient than pushing it. Because nature "abhors a vacuum", and this principle causes the low pressure area to be "filled in" easily, and the air is put into motion farther away from the fan.

Pushing air into a case and increasing the pressure inside the case causes the air to "try" to go back out through the fan itself, since it is closer than the exhaust. This is basic fluid dynamics.

What it really comes down to is: your choice about how you want to deal with the dust, that is inevitably in the air. in order to cool the computer, you must move enough air through it -- and that air will contain dust, no matter how you configure the air flow.

With negative pressure, you can either:

* Clean it regularly, including any filter.

* Or limit/control the intake as much as possible, to force air to pass through a filter -- by covering as many openings in the case as possible, except through the filter. If you want this to be quiet, you must provide enough filter area, and the filter must be low enough restriction to not force the fan to run faster, but still be effective at stopping dust.

With positive pressure you can either:

* Clean it regularly, including any filter.

* Construct a larger filter that can pass enough air, even after some dust accumulates. If you want this to be quiet, you must provide enough filter area, and the filter must be low enough restriction to not force the fan to run faster, but still be effective at stopping dust.

Negative pressure can move air through multiple openings with one fan: some in the front (over the HD's) and some through the side/back onto the video card. It also often concentrates the air flow past the CPU heatsink.

Positive pressure is less able to cause air to move through multiple openings (see above), because of the way a higher pressure zone will dissipate outward in all directions -- and it will take the shortest path.

Another way to think about it: a negative pressure case is moving air into the whole room, where it easily dissipates; and therefore the fan only has to work against negligible pressure. A positive pressure case is moving air into the case, which is not going to dissipate as easily, and therefore makes the fan work against a little greater pressure, all else being equal.

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

Post by whiic » Mon Jan 19, 2009 11:42 am

"Pushing air into a case and increasing the pressure inside the case causes the air to "try" to go back out through the fan itself, since it is closer than the exhaust."

Backpressure. But that's also present when trying to pull air out of the case... when the case has a vacuum, it tries to pull air back in through the exhaust even though there's a fan in there. Whether pulling or pushing, air resistance will fight against fan's capability to provide airflow.

Negative pressure:
+ one fan, multiple intakes => less fans, big diameter and low rpm => low noise
+ fan typically in rear => noise is where you aren't listening
+ easy to design airflow: each hole is an intake
- impossible to filter all air into the case
- impossible to prevent air entering floppy and optical drives
- air exiting exhaust opening is highly turbulent because of proximity of the fan, creating more noise and turbulence doesn't contribute to cooling when it has already exited.
- in theory, it's lower pressure, thus less dense and has less molecules to transfer heat. But in reality, pressure difference with low rpm exhaust fan and low-restriction intake is miniscule.

Positive pressure:
+ possibility to filter all air entering the case
- difficult to do so without creating noticeable resistance to airflow
+ possibility to prevent dust from entering floppy and optical drives
- this also requires filtering
+ turbulence is used to cool whatever is in the case. Turbulence occurs always after the fan and in this case, "after" is inside. In negative pressure, "after" is outside.
- if the intake fan is in lower front as it typically is in standard cases, turbulent airflow is used to cool HDDs which are far from hottest components in the case. Any cooling performance benefit is most likely wasted. If CPU was there, it would be a benefit to CPU cooling... but then the whole case air temperature would rise because of it.
+ in theory, air pressure is higher and heat transfer is increased. (In practice, proper filtering will reduce airflow much more than increased performance of higher pressure would gain.)
- all this summa summarum: very difficult to make it advantageous for silent computing, but very good for low-maintenance computers for dusty environments. Difficulty to design proper airflow can be solved by adding ridiculous CFM or multiple small fans that are in direct contact with ambient air. Point cooling at it's extreme.

Zero-pressure:
+ preventing air from entering or exiting the case outside of wanter intake and exhaust. No dust in floppy and optical drives even if air is not filtered (only needs to balance exhaust and intake fans to match the airflow)
- more than needed fans, some of which in front => more noise
- as theres no pressure difference, object outside the main path of airflow have zero airflow (where as in negative and positive pressure there's always air leaking in or out of the case from every imaginable crack and hole). Typically intake lower front and exhaust rear top: no airflow over PCI area and CD/DVD bay. Good cooling for HDD bay, southbridge, northbridge and CPU. RAM is not necessarily in airflow path thus poorly cooled.

I like negative pressure the most. There's some ways to combine it with positive pressure cooling. For example: use a tower heatsink for CPU and put the exhaust fan to push air into CPU heatsink, ductworking CPU exhausts outside the case. It's more efficient to do it this way than to install an exhaust fan and ductwork it to suck air through the CPU heatsink. Why? Higher pressure + turbulance after then fan. Thus you need less rpm to cool the CPU (but the airflow needed to cool the rest of the system remains the same) and you also place the fan deep inside the case, which should attenuate the noise as well.

QuietOC
Posts: 1407
Joined: Tue Dec 13, 2005 1:08 pm
Location: Michigan
Contact:

Post by QuietOC » Mon Jan 19, 2009 12:20 pm

whiic wrote:For example: use a tower heatsink for CPU and put the exhaust fan to push air into CPU heatsink, ductworking CPU exhausts outside the case. It's more efficient to do it this way than to install an exhaust fan and ductwork it to suck air through the CPU heatsink. Why? Higher pressure + turbulance after then fan.
I found a ducted fan sucking air through a heatsink (Ninja) worked better than the same fan pushing air into the heatsink. I think the problem with pushing fans is that most of the air short circuits out the sides. I didn't try a pushing fan with a duct on the Ninja. I am sure in both cases there is a lot of turbulence. I doubt axial fans produce laminar flow at this scale with any reliability.

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

Post by whiic » Tue Jan 20, 2009 2:54 am

to QuietOC:
That makes sense. Pulling though pipework may be better than pushing without pipework. "Short-circuiting" airflow is a bad thing. Best examples of this are stock coolers and other top-blowing coolers which tend to re-circulate the same air over and over again ad infinitum. (Though, they do benefit the most from ducting.)

I did some tests with OCZ Vindicator (that shameless clone of Ninja) with single fan push and single fan pull configuration in a practically passive cooling environment (motherboard nailed to a wall, PSU sitting on top of speaker case) and push configuration totally owned pull configuration. No ductwork. Forced airflow direction was upwards in both configurations. Pull configuration with 12V -> 3.6V undervolt had noticeable higher CPU temps than same voltage pushing air in.

Push configuration benefits from close proximity to heatsink as turbulence contributes to cooling in the same way as airflow speed itself. Or course with very low airflow speeds, the air exiting heatsink is roughly the same as heatsink temperature, thus improving heat transfer from heatsink to air (by means of adding turbulence) will not yield any benefit, only more wooshing noise.

In pull configuration, adding ductwork is beneficial. Always. Even when there's no risk for looping hot exhaust air, adding a spacer between fan and heatsink is a good thing. Fans aren't designed to suck air while it's blades are almost in contact with a grille or heatsink. You have probably noticed but fans blades are close the side where air is pulled than where it is pushed. The rods supporting the fan hub and blades are on the push-side of the frame, typically. Like pushing air to HS without ductwork creates turbulence, so does pulling without ductwork. The only difference is, that turbulence caused by pulling occurs outside the heatsink and contribute nothing positive.

judge56988
Friend of SPCR
Posts: 455
Joined: Sun Jul 27, 2008 10:39 am
Location: England

Post by judge56988 » Tue Jan 20, 2009 5:31 am

Doing a bit of reading on the internet about ventilation in the workplace seems to indicate an overwhelming preference for push-pull systems, with the push being less than half the amount of air than the pull; resulting in net negative pressure.
However to avoid short circuiting, pull outlet(s) must be very close to, or ducted to, contaminant, (heat) and obviously not close to the intake.
Arrange inlets to avoid dead spots. Not all inlets need to be mechanically pushed, as the case is negative pressure, air will enter from carefully positioned vents.
Turbulence causes friction therefore more pressure is required to move air - in a fully ducted instance (say a cpu tower type cooler) laminar flow is preferable. In an open case, turbulent flow would reach more ares.

~El~Jefe~
Friend of SPCR
Posts: 2887
Joined: Mon Feb 28, 2005 4:21 pm
Location: New York City zzzz
Contact:

Post by ~El~Jefe~ » Tue Jan 20, 2009 4:58 pm

they arent working with a system that is 3 feet tall, and they arent using a fanned PSU.

Let me explain what I tell people about this:

If you ahve negative pressure, which most of my cases did before I knew better, you will suck air from the outside into the psu, and into every crack in the unit. Dust is always a problem, unless you live alone and dont keep the psu on 24/7 and are a budding serial killer who puts plastic over everything is his home and washes his body with cholorine bleach because he is a very bad bad boy.

So, If you use a fanless psu, it doesnt really matter as much, but still dust will become an issue.

When I had fanned psu's, bottom 120's, or pushed by 80's from the side, the suction of the air out of the back or side of the case would cause the psu to heat up. I did not know this intuitively. One day I got myself a very good fan controller and some papst 120 mm fans. I put the fan controller up when i felt thepsu get hot. It kept getting hot, so I kept it on at higher voltages. One day, I lowered the fans blowing out to 5 volts each. The psu cooled off. The psu went into its lowest operational mode. I decided to game, I kept the fans on lowest setting, my system temps were the slightly higher. Was it starved for air? ok, maybe it was. I cut my case up, this helped everything, but the psu would still get hot as its 1 inch from the 120 mm papst fan. I put a 120 nexus on 10-11 volts in front and a 120 nexus on 5 volts at the rear. The psu became silent and cool/cold to the touch after gaming. positive pressure. I did this with all cases except my current setup. I have water cooling now and only one fan on rear and a fanless Phantom 350.

Fanned psu's work against negative pressure in a large way. Psu's generate more noise than any case fan.

negative pressure sux in many ways.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Post by Vicotnik » Tue Jan 20, 2009 8:51 pm

In my last couple of builds I've used negative pressure, with only a couple or a single exhaust fan(s). Dust is only a problem given enough time. If you clean the computer once in a while it's not really an issue and since I usually do inexpensive upgrades more than once a year the stuff seldom have time to get very dusty during the time I use them, cleaning or not.

ntavlas
Posts: 811
Joined: Mon Jul 16, 2007 2:35 pm
Location: Greece
Contact:

Post by ntavlas » Wed Jan 21, 2009 1:38 am

I`ve also experienced the issue of the case exhaust competing against the psu fan though with a psu that sucks air through the bottom it should be less of an issue.

I`d like to add another advantage of positive pressure: the reduction of dead spots. While my cpu barely benefited from the addition of two intake fans, many other hard to cool components did.

~El~Jefe~
Friend of SPCR
Posts: 2887
Joined: Mon Feb 28, 2005 4:21 pm
Location: New York City zzzz
Contact:

Post by ~El~Jefe~ » Fri Jan 23, 2009 11:33 pm

that is a great point. the air being blown around really makes for a much cooler motherboard, parts that dont have temp sensors on them, ram, and hd's.

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

Post by whiic » Sat Jan 24, 2009 4:03 am

PSU fan competing with exhaust fan for airflow is a bad thing as PSU fan ramp-up typically causes much additional noise compared to same air moved by exhaust fan. (PSU fans aren't the quiet type, they aren't softmounted and there's turbulence inside the PSU.)

But luckily this competition for cool air isn't anything alarming if the case has low air restriction. Do cablegami and use a case with low-resistance intake, and I'd say you'd be ok. The exhaust fan will cause low-pressure within the case and thus resist PSU fan from cooling the PSU, but it also ensure case temperatures drop and lower amount of airflow within PSU is sufficient to cool it.

While negative pressure cooling may be quieter than inadequately cooled (as PSU won't ramp up), positive pressure cooling may be even more quiet as intake air to PSU would be low (as with negative pressure cooling) AND the positive pressure would reduce need for PSU ramp-up (even possibly allowing removal of PSU fan... if you like gambling). There's some tweaks you need to do, though.

Negative pressure: each hole in case is an intake. Blocking holes that aren't needed will increase airflow from the holes that direct airflow to HDD stack, PCI area, etc. Cooling holes on left side panel are typically useless as air will mostly flow directly to exhaust and help cool nothing (will bypass even GPU). 5.25" bay might be another unless it's used for suspending HDDs. But blocking them will increase low-pressure build-up and cause risk of PSU fan ramp-up, so block any holes only if you experience high temperature on HDD tray.

With cases that mount PSU at the bottom or on top in an isolated chamber, high negative pressure in case wouldn't cause PSU fan ramp-up. High negative pressures still aren't something you hope to achieve, though.

Positive pressure: worst leak sources would probably be left cover openings and exhaust openeng. PCI cooling is better done by leaving some of the covers plates off. Exhaust opening may or may not be good to leave open. Blocking it would potentially allow removal of PSU fan, leaving it open would allow for lower positive pressure and more airflow for the same rpm on front intake. 5.25" bay leaks are less significant as airflow would need to make a U-turn. Also, leaks with positive pressure aren't a real problem: with negative pressure the leaks reduce HDD cooling, with positive pressure, the only case fan is blowing at the HDDs, and where the air escapes has no effect for HDDs. (It would for the PSU, if you have fan removed from it.)

Combination: positive pressure builds typically still have at least one exhaust... PSU fan. They are still positive pressure builds. Negative pressure builds can also use a very low rpm intake fan with less airflow than than exhaust fan + PSU fan combined. That way you don't need to block all unnecessary intakes to increase air through HDD stack. Small negative pressure in combination with a removed PCI slot cover or two would create enough airflow to prevent black spots. Completely zero-pressure-differential is IMO a bad thing. It doesn't cause optical and floppy drive dust build-up but it would create a heck of a lot of black spots, including almost still air in PCI area and 5.25" with temperature induced natural convection and maybe some assist by turbulence of intake fan. But if there's significant turbulence, it's probably not spinning at low rpm anymore...

My two most recent builds both use a combination. My quadcore build uses net negative pressure with low airflow resistance Centurion 530 case. SpeedFan controlled front intake (ultra low rpm) and hardwired exhaust fan (@3.6V, a high rpm fan spinning at low rpm). Corsair HX520W exhausting as well.

Prescott build is most likely net positive pressure. Undervolted Nexus Real Silent Case Fan(TM) as intake, PSU fan as only exhaust. PSU fan is both swapped from high-rpm to low-rpm type AND undervolted. Fan only acts as some extra lifeguard for PSU as intake fan alone could either cool the PSU or the CPU sufficiently but probably not both (they aren't on the same path of airflow so it'd need quite a bit of balancing to get it to work).

Plekto
Posts: 398
Joined: Tue Feb 19, 2008 2:08 pm
Location: Los Angeles

Post by Plekto » Tue Jan 27, 2009 6:35 pm

Note - an undervolted a bit slot fan(7V usually) is a godsend in any negative pressure build.

gud4u
Posts: 30
Joined: Sat Feb 18, 2006 6:57 pm

Post by gud4u » Sat Feb 14, 2009 4:19 pm

I was about to say I favor positive-pressure, but I'm actually running identical 120mm Panaflos at 7.5V in push-pull front-intake and rear exhaust.

I guess you'd call this simply forced front-to-back air flow. No other intake or exhaust ports. Even flow over all components. HD cages removed, HD is elastic-suspended in air-stream.

My case air intake configuration is a bit different, since I use 3M spun-glass filter/AC filter media surface-mounted over the front intake fan. This stuff is corrugated, about 1" thick overall, cut-to-fit - inconvenient to mount any other way than surface-mount.

Good air flow, quiet and hospital-clean case interior.

Hope this helps!

psiu
Posts: 1201
Joined: Tue Aug 23, 2005 1:53 pm
Location: SE MI

Post by psiu » Sat Feb 14, 2009 6:43 pm

Depends on the case a lot. My HTPC works good with 2 120mm intakes, and the only exhaust is the fan in power supply--and it gets air before any hot components. Lots of other openings for air to escape though.

No dust filters unfortunately. That is one thing that I do love about the P180--dust filters that are easily accessible. Conventional cases with conventional components though, probably negative or equal airflows.

pcy
-- Vendor --
Posts: 137
Joined: Sat Aug 13, 2005 2:46 am
Location: Central London, England

Post by pcy » Mon Feb 16, 2009 9:13 am

Hi,


It makes no difference.


All that matters is pressure difference.

The pressure differences we are talking about are tiny in relationship to Atmospheric Pressure, so there is no significant impact on air density to consider.

In consequence the only thing the air is sensitve to is the pressure difference between any two points, and the absolute value of the presure has no significance.

Air does not easily flow thoiugh narrow cracks unless driven by a substantial pressure difference. Whether or not you PC stays clean inside depends only on the filter on the main air intake.




Peter

~El~Jefe~
Friend of SPCR
Posts: 2887
Joined: Mon Feb 28, 2005 4:21 pm
Location: New York City zzzz
Contact:

Post by ~El~Jefe~ » Tue Feb 24, 2009 2:27 pm

I do not know what you are referencing and the scope of what you are referencing. Cooling a computer is actually not simple. There are many strange sources of concentrated heat. The total heat amount isnt great. References to cooling buildings or facilities or single objects or dwellings has little to do with what makes a computer happy. Take the VRM's. they really are very low watt units but they are tiny and without heatsinks most of the time. VGA card is upside down and blocking upwards air flow. The removal of total heat from the case is only one concern. Noise is generated mostly by the psu. Most people dont want to mod a psu fan and many enthusiasts game at some point in their computer carreers, so really, who wants to tinker with a 300 watt draw?

You need turbulence in a case if you want to go low on the fan speeds. positive pressure from 1-2 front 120mm fans create a whirlwind of air that cools the most expensive and sensative parts. The total heat probably only is like 70 watts normal usage and 180 watts gaming. that's really tiny, but the concentration of the heat is insanely tight.

I cant think of anything where the force of direction has no effect on the system. rear wheel drive cars are more efficient at insane acceleration, pushing vs pulling. A fan in a case at the bottom pushing air with heat naturally rising on the top, cant be the same thing. Air being sucked into a case causes a steady stream of just pressure with low turbulence, blown in it has more. theres more force at the source of the blade. the blades push the air faster, this isnt some closed tube system. It is about removing heat from items quickly as prime importance. Only thing that will be the same or close to it is total volume of air moved. the way it does and where it comes from its so completely different in a push vs pull case setup.

Post Reply