Author Options:

How do I measure pressure inside computer case Answered

. I've setup a home file server and would like to be able to maintain positive pressure in the case to keep dirt out (two dogs in the house). I've cut some filters to fit the fans blowing into the case (front, rear, and side) and installed a "PCI slot" exhaust fan (it and the PSU are the only exhausts). All these are connected to a four-channel controller.
. Using the handy-dandy strip-of-paper-in-front-of-a-case-opening method, case pressure is real close to being equal to ambient with all intake fans on high and the exhaust fan on low (700W PSU has two fans that seem to work very well). Even under full-load, BIOS/SpeedFan says the 1.8 GHz Celeron temp stays below 140 degF under full load and idles in the mid-to-high 90s. HDD temps (2x400GB RAID 1 + spare and 40GB boot) stay below 105, except when formatting when they might rise to 110.
. I have some clear tubing, so I could rig up a manometer, but don't feel comfortable hooking up a water hose to my computer. Any ideas on a cheap-and-easy way to measure DeltaP (I'm guessing 0-5 inWC would work) without liquids? Could a automotive Manifold Absolute Pressure (MAP) sensor (I think they have a range close to 0-30 inWC) be adapted?
. Someone point me in the right direction and I'll post an Instructable when I figure it out.


Update for anybody that cares . Rigged up a 2L plastic bottle, latex glove, and tubing. Works very well. . Even after removing the PCI-slot fan and making all the case fans blow in, I still had neg press. PSU fans are too strong. Blocked part of the PSU inlet with tape and stuck a t/c inside. . So far, so good. Still seem to have more than enough air flow. Waiting to see what happens when the filters start plugging up.

. How would the exhaust fan suck muck in (LOL is that a motel in Hanoi?)? . The PCI-slot exh fan doesn't seem to make much difference regarding case/CPU/HDD temps - it's just to make sure there's some circulation around my EIDE/RAID card (not that it really needs it). I have no control over the PSU fans (yet) - I'm assuming they are full-on all the time, but may be temp controlled (it's a 700W unit and may not be getting warm enough to kick the fans to high). . The gaps won't change what I'm trying to read (unless I put the end of the sensing tube next to a gap). What kind of pressure did you measure? . If I'm running positive pressure in the case, I won't have to worry about all the gaps (I won't have to "Mind the Gap!" heehee) - as Trebuchet03 pointed out.

(My rig) I mean there are great big holes either side of the drive, so I'm not going to get any positive pressue. (Yours) Anything blowing out will suck air in, unless your feeding plenty of filtered air in to compensate. And you probably are. My case tends to accumulate bloody cat hair on the filters... (better there than inside I suppose) L

Anything blowing out will suck air in, unless your feeding plenty of filtered air in to compensate. And you probably are.

. OK, I think I see what you mean.
. That's just it, I'm _not_ getting enough clean air to compensate right now.
. I may put a controller on the PSU fans, but that's a last resort.
. What I want to do is balance the air in vs. the air out so that (in priority order): temps are kept within reason (this doesn't appear to be a problem unless I decide to make all the fans exhaust and maybe not then, since cases are so leaky), positive pressure is maintained, and noise is kept to a minimum. If not for the noise, I could just plug all the intake fans to 12V and remove the PCI-slot fan.
. If I have some means to measure the case/ambient DP, I can fine-tune fan speeds to achieve minimum noise. I don't need much DP - anything > 0 will work.
. For now, I think I'll use a manometer (I have ~100' of tubing, but it still makes me nervous - I suppose I could rig up a "catch-cup" in the case) to get me in the ballpark. Then use a balloon as a tattle-tale.
. I'd still like to have an electronic DP monitor. After establishing baselines, it could be used to determine when the filters need to be cleaned (or other malfunction, eg, fan failure) and trip an alarm.
. I know what you mean about the animal hair. ;)
. PS: Sorry to hear that your pet is bleeding. LOL

I'd use your tube, just be careful how you do it. Have a long U shape, say at least 12" and start with only a couple of inches of water. L

. You, sir, are much braver than I. :) . I'm sure that, at the pressures involved, the chances of sucking water into the case are slim, but I've spent way too much time on this project to assume that risk.

If you've filtered your intakes, the air entering the case will be as good as your filters. Your only chance of a problem would be the exhaust fan sucking muck in. So how is the temperature without the exhaust fan? I did just have a go at checking pressure (tube) on my machine, but there are big gaps around the 5 1/4" drive, which I couldn't be bothered to seal-up. So I don't know whether this would work. L

When you say really close to being at ambient pressure (that sounds really funny) - do you mean in the negative or positive direction? If it's positive (and I think that's what you said)... That's it -- you're done ;) Dust particles can't go upstream -- even if that stream is flowing very very slowly ;) Just remember that no filter is perfect -- you're still going to get coated with dust on the interior. This is why, for the home/office user the designers weren't too anal about it. Big dedicated server rooms are another story. If you really want to keep dust out... you could go for a submerged setup using oil ;)

. It's negative (see other msg). . I don't expect the filters to be perfect, nor even close. I doubt the fans could handle the pressure drop of a "real" filter - the "HVAC outlet vent" filters I've cut-to-size are bad enough and they are advertised as low-DP. . I have a few aquariums (aquaria?) laying around and have thought about trying a submerged setup. Not that I need it - it just sounds cool. I'll probably never get around to it.

. I tried sticking a lighter in front of the side vent and now I have a hole burnt in the filter! That tells me I still have a slight vacuum inside.
. Judging by the internal temps, I have plenty of air flow. Which kind of surprises me, what with all the filters I've installed.
. I've never heard of a "balloon barometer system," but it leads me to believe I should be able to lightly stretch an elastomer over, say, the cut-off end of a 1L bottle and run some tubing from the case to the bottle. That should indicate pos/neg pressure (all I really need). Heck, I don't even need the bottle - just run tubing from the case to an intact balloon.
. What I would like to find is a cheap DP transducer with a mV output. I could hook that into a bay-mounted temp controller I have and alarm on high pressure.

I don't think positive pressure will help much, as you NEED airflow though the case (for a normal box anyway) and dust and such will follow the air, though the inlet. Filters should be fine. But It you REALLY want to measure pressure, I'm thinking some kind of balloon barometer system. If you make it REALLY wide (6 inches, maybe more), then use a fairly long needle, It should be sensitive enough.