Fantastic results with fan upside down

rx2 · Post by **rx2** » Thu May 12, 2005 3:29 am

Hi,

Thought I would just share some interesting results. I have just built a system :

Antec SLK3000B
P4C800-E
P4 3.2 Norwood
Seasonic S12 430
2G Corsair RAM
2*SATA 160 G
1*IDE 80 G
XP-90 Heatsink
92mm Panaflo for heatsink

I have the stock 120mm tri-fan that came with box set on low. I dont have any front fans.

After putting it all together I ran Prime95, and monitored the temperature with SpeedFan. I have also set up the BIOS to use Q-FAN @ a ratio of 11/16. The result was a CPU temp of about 50-51. Not too bad I thought.

I then decided to flip the fan on the heatsink, and see what happens. Well doing the same tests the CPU has not gone above 40. One thing I have noticed though on SpeedFan is the fan on the seasonic has speeded up slightly, but not to make any noticable noise difference.

All in all I am pretty pleased, nice quite box!

Cheers

Robert

BrianE · Post by **BrianE** » Thu May 12, 2005 8:39 am

So just to clarify, you're saying that you changed the fan direction over your XP-90 from blowing air onto it, to sucking air through it?

10 is a pretty good difference. I should do more research on this later on, as I was kind of thinking about the suck vs blow debate as well. Probably depends on the design of the heatsink though....

benx · Post by **benx** » Thu May 12, 2005 8:51 am

Same here with SKL900 with 92mm fan my temp dropped 4/5C

sundevil_1997 · Post by **sundevil_1997** » Thu May 12, 2005 9:25 am

Interesting results. Is it possible somehow that drawing the air through the HS from the bottom (near the processor) that it's getting cooler air? Maybe from some grillwork nearby? Usually blowing air into the HS, the fan is always drawing air from the middle of the case, which is probably heated up by most all components. Just a thought.

EDIT: Another thought...blowing air straight down onto a flat surface, it's going to spread out and probably be easily drawn back up to the low pressure area caused by the fan intake. So that would probably tend to result in a lot of air recycling, and thus general heating up. But I'm guessing there's probably not a surface close to the top of the HSF, so air exhausted that way probably isn't quickly or easily drawn back down to the motherboard, and thus not recycled as easily or often.

I wish I could try this on my media pc. It's an ASUS P4T motherboard and it has the HSF positioned almost completely under the PSU. Unfortuately, the PSU I have has a 120mm fan that's pulling air up out of the case...and about 2-3 inches below that fan is the HSF trying to pull air down and into the HS...so somewhere between that 2-3 inch clearance, there's probably a small vacuum going on.

If I could reverse that HSF, then it'd be pulling air from the base of the motherboard, through the HS, and directly into the PSU fan and out the case.

Unfortunately, this is the stock HSF for the 423 socket, which is one big piece of plastic....no simple screws to flip the fan. I might tinker with it and see if I can pull the plastic off, maybe the fan actually is a seperate component inside there.

Otherwise, does anyone know....could I just flip the connector for the fan and reverse the polarity? Does that work to run it "backwards" as opposed to flipping the fan itself. Seems like most fans are meant to only run in one direction.

mongobilly · Post by **mongobilly** » Thu May 12, 2005 9:55 am

You could reverse the direction that way, it probably won't like it very much tho. Anectodal evidence: I recently mounted a Yate Loon the wrong way, started to make clicking noise right away, fixing the polarity made the noise go away.

BrianE · Post by **BrianE** » Thu May 12, 2005 1:15 pm

I think most fan blades that you see are only meant to spin one direction... so just reversing the motor might not work as efficiently as actually flipping the fan over.

sundevil_1997 · Post by **sundevil_1997** » Thu May 12, 2005 1:35 pm

BrianE wrote:I think most fan blades that you see are only meant to spin one direction... so just reversing the motor might not work as efficiently as actually flipping the fan over.

Heh..nothing is very easy. I'll try removing the whole plastic "enclosure" to see if maybe the fan itself really is removeable. I sense already that I'll probably end posting in the CPU forums shortly about finding a good HSF replacement for a 423 socket.....

Sooty · Post by **Sooty** » Thu May 12, 2005 2:13 pm

Rx2,

With your heatsink fan blowing downwards, it’s competing for the very same air as the 120mm case exhaust fan. With the XP-90 fan reversed, there’s no conflict between the two fans.

cpemma · Post by **cpemma** » Thu May 12, 2005 2:46 pm

sundevil_1997 wrote:Otherwise, does anyone know....could I just flip the connector for the fan and reverse the polarity? Does that work to run it "backwards" as opposed to flipping the fan itself. Seems like most fans are meant to only run in one direction.

Brushless fans will only run one direction - reversing the polarity will at best result in no motion at all, at worst blow the semiconductors switching the fan motor.

GUNNER · Post by **GUNNER** » Sat May 14, 2005 5:10 am

This thread reminded me to try the same thing.

From my setup (see http://forums.silentpcreview.com/viewtopic.php?t=21871 ), I also have 3 120mm fans all in the upper rear of the case. A 120mm Nexus blowing out the back, the 120mm in a Seasonic S12-380 pulling up-and-out, and a 120mm Nexus blowing down on a XP-120 heatsink.

I reversed the fan on the heatsink to blowing UP and also saw a temperature drop. (Not 10C, though.) The digital sensor I have next to the heatsink base went from 42C to 39C. The motherboard reported temp went from 52C to 47C.

Other temps in the system stayed the same.

Fantastic results with fan upside down

Fantastic results with fan upside down

Re: Fantastic results with fan upside down

Re: Fantastic results with fan upside down