Better cooling with air filter??

[Laidbacklux]

My system has one 120mm Panaflo L intake, a 90mm zalman over the cpu and a 90mm zalman out. Put a filter on the 120 in, noticed less airflow but stuck it in, and my cpu temps went down 2 degrees C than usual (usually 20 deg c above case temp, now 18). Took the filter out to clean it and temp went back to 20 above case temp, put back in and temps dropped back to 18 above case. Case temps did not change with or without filter.

Can anyone explain why decreasing air flow might have helped here?

[Massive_gas]

cleaner air? smaller amount of conductive particles? less friction from dust? beats me. I've never heard of this before.

[Clutch]

Do you live in a dusty area? Is your tower under a table or like in a cabinet or something?

[Cool Canuck]

It may have something to do with the air flow pattern at a higher velocity within the case. Interesting.

[DJ-Quack]

Maybe the 120mm was/is doing more harm than good, so decreasing airflow actually helped.

[CCW]

Dust acts as an insulator so the mroe dust going in, the hotter your system will become. Too much airflow can be as bad as too little airflow as you need equal ariflow in as out for best cooling.

Craig

[DexterHolland911]

DJ Quack/CCW hit it on the head. Assuming your 120mm fan is more powerful than your 90mm, you were/are getting stagnant air in your case that is heating up but not being removed. Switch your 120mm fan to exhaust and your 90mm to intake if possible (switch locations and directions, not just directions) and you should get even lower case temps.

[Cool Canuck]

Originally posted by CCW
Too much airflow can be as bad as too little airflow as you need equal ariflow in as out for best cooling.

Craig Care to explain?

[stoopid]

Originally posted by Cool Canuck
Care to explain?

Dexter just did.

[dNA3D]

Better yet, have some negative pressure, meaning more CFM sucking out than CFM blowing in. Sure, you'll have more dust, but then negative pressure also allows air to enter through any holes in your case where fans are not present to equalize the pressure.

Hey, what a revolutionary idea! Way to go dNA.

You can tell I require mental aid.

[Cool Canuck]

/me thinks the laws of physics have been replaced with drugs.

[stoopid]

Originally posted by Cool Canuck
* Cool Canuck thinks the laws of physics have been replaced with drugs.

Think dNA just explained that :p

[izzugotom]

Faster air through the case does not mean cooler at all places. Maybe the Processor fan grasp more air on intake, becase it flow slower over/past it, most probably because of how your other fans were arranged.

[Laidbacklux]

Hmmm, I always heard positive pressure was better?, I am soo replacing the 90mm cpu fan with a 120 (still 90 on exit), i'll post what happens then.

[dNA3D]

Originally posted by Laidbacklux
Hmmm, I always heard positive pressure was better?, I am soo replacing the 90mm cpu fan with a 120 (still 90 on exit), i'll post what happens then.

AFAIK, just about all system and case manufacturers use negative pressure. Coolermaster, Lian-Li, Alienware, Dell, HP, Antec, Chieftec, Chenming... and the list goes on. Most of those listed have 4 fans, 2 intake and 2 uhhhh, outtake?? Then, there's that PSU fan which we must forget. My conclusion.... have marginal negative pressure, not so much that you'll suck your case panels in (not good).

[Geeky1]

positive pressure just causes the fans inside the case (cpu, video card, etc) to be recirculating hot air. negative pressure forces cool air in wherever it can get in- basically any seam, hole, or vent in the case becomes an air intake. neutral pressure would render any vents that don't have fans on them useless. Neutral pressure works very well, as long as you have fans on everything you want to cool. Negative pressure works almost as well, if not a bit better, and it doesn't requre fans on everything if the case is properly designed, so it can be quieter. Positive pressure is essentially useless for most cases on the market because they weren't designed to be cooled using a positive pressure setup. If you were building a case from scratch, any of the 3 could be used effectively.

[Enverex]

Negative pressure is a big problem for drives, i.e. my CD-Rom drives are always full of crap when I open them (very dusty room) meaning they get clogged up easily and I end up having to clean them alot more often.

NS

[DexterHolland911]

Too much negative pressure is not a good thing. It is best to try and keep it as balanced as possible, though marginal negative pressure is what you should be aiming for. One might think that the concept of negative pressure, ie air being sucked into the case, is a good thing, but really it isn't. All the small streams of air cause lots of turbulence and interference, and this is detrimental. Aim for a windtunnel :)

[Enisada]

You could always get a smoke machine and see what is going on in there. You have to admit it also looks pretty cool too. Seeing all that smoke flying around in there. I want to go and do that now....

[Geeky1]

Dexter, that's true, but I doubt that would be a problem in a standard case because I don't think that axial fans can pull anywhere near a large enough vacuum to cause a big problem, altho I may be wrong about that... I haven't done any laboratory testing on this stuff, I'm just going with what I've seen, so...

[izzugotom]

I did it this way:Build PC into a Desk-Removed the drawers.

A]Cold air enters the desk via Large filters, drawn in mainly by a 220V fan[300mm] right at the rear of the desk. The purpose of this fan is 1-To overcome the resistance the Filters pose[3x VW Beetle air filters] 2- To create a draft through the desk from front to back, thus the air always flow one way for sure.

B]All the stuff PC + Transformers + 12V fans PSU gets from these fresh air. Then I used other 80mm Coolermaster fans to direct these air to the spots I want. 1-2x fans in tandem beneath the CDROM/DVD drives blowing air in from the front of the case onto memory etc,2-1x fan from the lower front bottom sucking air and direct in to the Disk drive.3-1x fan at the bottom sucking air in and blow it to the Grafics card GPU's fan. 4-2x fans at the rear sucking warm air out + some holes at rear. I made use and modify plastic Gutter pieces to accomplish this.

C]The PSU is still in the case, but the sidecovers had holes drilled in on top, so that the PSU sucks it's own cool air in from the front, also via the filters. The PSU has 2x fans in tandem at the rear preventing other fans from lowering the sucking capability of the PSU fan, which happens if you add extractor fans, preventing it from running hot.

D]At the rear the extra PSU [140W of old PC] has its own cooling fan.

E]The PC case in surrounded by foam around the centre, to prevent it from ever sucking in 2nd hand warmer air from the rear.
This is quickly removed, so taking the case out of the desk is no problem.


F]With this experemental method I succeeded to always keep my P4 temp at 4-5c at IDLE or 8-9c at FULL LOAD above room temp. Therefore it is always important to mention room temp when comparing temps, so as to keep perspective.

G]The modem is cooled on top of the desk...Hole in desk to suck air over modem [Holes drilled in casing]. This is after the centre sealing of the case with foam, so pose no dust thread.

Result: Every little component in this PC is cooled. I believe, if you chuck out fans for silencing, and use watercooling, your processor may become the heatsink of your motherboard so to speak. I like watercooling, but will never do it without fans also.

My temps are now:Measured separately with installed sensors and Digital temp gauges in a console. I can see temps when PC is switched off also. If theres action concerning the P4, the temp start rising within 2 sec.
1-Air in=15,6c
2-P4=19,5c
3-GPU=19,0c
4-PSU=18,2c
5-Case air out=17,1c
6-Desk Air out [220V FAN]=17,7c
7-Modem=25,3c [Partially cooled]
8-Monitor=24,8c

[Laidbacklux]

Are vent holes necessary? I am building an aluminum inside, wood outside case that will only have 1 intake, 1 out and psu out. no other vents. Same fan in as out though intake will have a slight filter. PSU is antec true and the dual fans are moving very little air (1500rpm). Either way the case will be slightly negative cause of the psu and the filter (maybe) but no extra vents...will i be cleaning my cd roms often?? Would this type of case work better with positive pressure so that there would be more circulation? The hope, minimal number of large as possible fans, minimal openings for minimal noise escpape, but this means the flow of air is more crucial, so perhaps a positive pressure system would be ideal??? Does my intuition jive with reality?

[izzugotom]

To be honest with you I don't understand the meaning of POS/NEG airflow effecting dust, because if air comes in it will carry dust with it, maybe you just alter the area where the dust is going to land.

[Cool Canuck]

Originally posted by izzugotom
To be honest with you I don't understand the meaning of POS/NEG airflow effecting dust, because if air comes in it will carry dust with it, maybe you just alter the area where the dust is going to land.
Ahhh. Finally someone gets it right. If your fan(s) push air into the case and you filter that air, you get minimal dust. If your fans draw air out of the case, the replacement air will come in from any and all cracks, holes, what have you. You are unable to filter the air entering the case.

Regardless, what air goes in, it has to come out and visa versa.

[izzugotom]

Theres only one way to prevent dust------You must have a Filtered controlled intake, in my case everything is sealed except the Filtered intake and of course the 220V Fan-OUT hole.

[Laidbacklux]

Ok, so now i replaced the 92mm (zalman set to about 45cfm)with a 120mm(papst at 55cfm) fan for my cpu (blows on a zalman copper flower) and the temps went higher...wtf? Thought maybe it was too close to the exhaust so i experiemtned and found the optimal placement, still higher than temps with 92mm fan.

BTW, I checked out the cfm of all the fans, intake 120 panaflo 68cfm, exhaust at zalman 92mm at 45cfm, psu antec true usually running at 25cfm. So some negative pressure here. I took off the filter for the intake and temps went up (thats what started this thread) but then i took off the filter and increased the exhaust by about 10 cfm...this increased temps too.

So i think that since the cpu fan and exhaust are so close and draw from the same source that increasing the exhaust screws with the immediate pressure over the cpu fan and drops the cfm. Been experimenting with the cfm of intake and exhaust but as of right now the optimal balance has been a filtered 120 intake, a 92mm fan over the cpu and a 92 exhaust...any suggestions of a better config seeing as i have this nice papst 120mm fan?

[Geeky1]

calculate the velocity of the air coming out of the fan. I'll bet you that the velocity is lower coming out of the 120mm than it is the 92mm, hence the higher temperatures...

[Laidbacklux]

The cfm for the 120 is higher, it's supposed to be and i also can feel a noticable difference...as to the speed of the air its a good point, any easy way to measure it. For 55cfm to move through a 120mm fan vs 45cfm to move through a 92mm fan, I'll brush up on some physics...but care to explain why...

[Geeky1]

A 55cfm 120mm fan yields a LFM (linear feet/minute) rate of ~354.84, or about 4mph. A 45cfm 92mm fan yields a LFM rate of ~493.94, or about 5.6mph. Assuming that the heat carrying capacity of the air is the limiting factor (which it almost certainly is) the faster the air is flowing, the better. The other issue is the amount of the air reaching the heatsink- a larger fan is going to mean that less of it's output is going to reach the heatsink, which, unless compensated for with higher total airflow, could produce the increase in temperature that you are experiencing.

Both of the above factors contribute to the performance of the Delta "focused flow" fans- they have very high airflow for their size, and it's focused into a relatively small area, so it's moving at a much higher velocity than normal.

[Laidbacklux]

ok ok, more info, case open, cpu temps drops to about the same as with ol' 92mm, case temp may go down but not much, not even a degree C. So I am gonna try to put a old cpu hood on the 120 opening toward intake in an effort to shield any pressure effects of the exhaust...i'll keep anyone interested posted...would appreciate any advice.

[Laidbacklux]

k, i didn't feel like taping and or melthing plastic so i just sat the hood behind the intake so it directed air at about a 45 degree angle upward, case and cpu temp dropped (1 C case, 3 C cpu)...the only if is the video card temp that i cannot check on...but the negative pressure is sucking the air in through the vents near the slots so I am hopeful...not a final config but what ever is.

[DexterHolland911]

Laidbacklux - Are you having problems with temps or do you overclock? This much fine-tuning isn't necessary.

Geeky1 - Velocity of air is not so important, the cfm of the fan is (assuming all the air is being blown onto the heatsink). But you are right, the 120mm fan is 'wasting air', the heatsink is not large enough to use all of its air of the 120mm fan. Especially when it comes to case-cooling, cfm is what matters.

[Enverex]

80>60mm converters are a nightmare, they blow more air back on themselves than they blow into the heatsink itself and consequently this makes lots more noise, same would be true for 120mm>80mm converters, so regardless of what you do, try and stay away from converters.

NS

[DexterHolland911]

My old review/huge heatsink round up is lost, but I did a lot of testing on converters. NS is totally right, stay away from them.

[Laidbacklux]

Not overclocking, just trying to have a quiet system. My system:

Athlon XP 1600+
Asus A7V266
512 pc2100 generic
ATI AIW Radeon
SBLive 5.1 Plat
WD 80gig Special Edition

I have a Zalman flower heatsink that is actually bigger than a 92mm fan, the 120 covers it. My temps right now: ambient 25
Case 27 CPU 47

So I would ideally like to lower noise while not raising temps, so i figured a larger fan (papst 120) instead of the 92mm zalman fan over the cpu would do that. Noise is slightly less, temps went up though.
The hood i placed on the bottom of the case does not narrow the air flow, it just redirects and it seems to workfairly well.

As to the sound level is my pc, the loudest thing in there now is the WD HD which i am thinkng of sandwiching (via silentpcreview). I read 7volts negative review but it seemed that he did it quite differently and its not that costly compared to a smartdrive silencer( WD is too hot for silentdrive).

I think the velocity vs cfm is a good point when dealing with 2 fans near each other competing for the same air. Cause regardless of whether the 120 is wasting a little air, temps should not go up when using a fan with larger cfm, velocity makes intuitive senes to me but that may not mean much in the long run.

[DexterHolland911]

*sigh* Gotta love intuition... but that has nothing to do with the real world. The other part of it could be that the Zalman flower heatsink is a skivved heatsink... sort of. Therefore it needs a high-pressure, high-speed airflow through it to be effective. Laidbacklux, I was a supermod with 1500+ posts at the old icrontic probably long before you ever came here. Now I know I have 22 posts at the new icrontic, and even in the last 6 months at the old icrontic I didn't really contribute, I had other things to focus on. But I have been around computers for a long time, I'm not just pulling stuff out of my ass (which I have seen a little too much lately...)

My advice, you want to get rid of noise? Get watercooling :)

*edit* When I said wasting air, what I meant was that the area of the 120mm fan is larger than that of the heatsink. Therefore the 120mm fan will be blowing air not only on the heatsink, but on the surrounding motherboard where there is no heatsink. This air is being 'wasted'. A smaller fan will be blowing only on the heatsink, its air is not being 'wasted'.

[Laidbacklux]

DexterHolland911
Not only do i appreciate your comments but i take them into serious consideration, as I do with most intelligent comments posted regardless of number of posts. I am not quite to the point of watercooling, maybe eventually. For now I tweak with the wind, and as you point out, it may not really be worth it. If not, then perhaps others who are thinking about replacing their 92 with a 120 for their cpu will learn from this less than successful effort.