200 CFM fan

It really depends on the type of heatsink, whether it was designed for high flow and such.
 
Dang, these things remind me of the 60mm Vantec Tornado lol.

You will see some difference, more or less depending on your heatsink, also what kind of supports and tie down straps you have.
 
200cfm.. holy crap. What kind of vortex does that create inside your box? lol I have a 2000rpm fan that does 77cfm on my heatsink and its throttled so its fairly quiet until its at full speed. I can only imagine 200cfm.
 
I've used 120x38 YS-Tech fans rated around 120cfm iirc, and the noise was horrendous. I had 4 in my system upstairs, and you could hear it downstairs in the lounge - through the floorboards it sounded rather like the gas boiler chugging away. Netted me some great temps though, plus you could place a CD over one of the top blow-holes and it would hover about an inch and a half above it.
 
Lol, im not gonna buy one. I have a zalman cnps10x extreme and it doesnt get that hot on my w3520 @ 4.2

I just got bit by the curious bug and looked up the max rpm on a 120mm fan. Most I had seen until then was 2000rpm.

Although, I put up a wall in my garage this weekend and made a 10x10 computer room. Put down carpet and everything. Just havent added a ac duct yet. Might pick up a few of those to keep the room cool.


BTW If i put my 5870 fan on max... It will blow paper off my printer.
 
equinox654 said:
I saw a few high speed 120mm fans online today. They are around 200cfm of air flow.

Has anyone put one of these on their heatsink and saw any big temp decreases?

Here is one for reference
http://www.directron.com/pfb1212ghef00.html
Click to expand...

I used a large Delta fan (two in push/pull at one time), 150cfm+ each, also loud as all hell, but even at full speed, I might get 5C under my current fan running at 900-1000rpm, from what I have seen after about 70cfm it becomes REALLY hard to cool off the HS any faster and you are starting to be limited by how fast the HS it self can remove heat from the chip to it's fins.
 
Buy grills. I had one of those 80CFM Delta's take a huge chunk out of my finger about 8 years ago. Damn things are dangerous.
 
BlueFireIce said:
I used a large Delta fan (two in push/pull at one time), 150cfm+ each, also loud as all hell, but even at full speed, I might get 5C under my current fan running at 900-1000rpm, from what I have seen after about 70cfm it becomes REALLY hard to cool off the HS any faster and you are starting to be limited by how fast the HS it self can remove heat from the chip to it's fins.
Click to expand...

There's also the fact that rate of heat transfer depends on the temp difference between two mediums (Newton's heat equation). Air at room temp and CPU at 50c means you will hit that barrier pretty quick. On top of that air have smaller heat capacity. At that point you are better served with water or chilled air at least.
 
they usually vibrate too much to run full speed I dont like my cpu and mobo vibrating with that fan and a big old heatsink also.
 
They work rather well. A friend of mine put a couple of those on the back of his Antec 1200. When going full speed, they made the top fan spin backwards. You could also feel them from more than 5 feet away. Unfortunately they are extremely loud and you probably don't want one due to that fact. I honestly wouldn't use them on anything apart from a rackmount server personally.
 
Hypernova said:
There's also the fact that rate of heat transfer depends on the temp difference between two mediums (Newton's heat equation). Air at room temp and CPU at 50c means you will hit that barrier pretty quick. On top of that air have smaller heat capacity. At that point you are better served with water or chilled air at least.
Click to expand...

Yes, air has a smaller heat capacity, that's why you move more air, but it does not matter how much you move if the HS can not keep up. I also could only wish it was loading at 50C. :eek:

load was around 75C-80C (was really pushing it at the time), and did a few tests with room temp being about 50F (quite cool) and results were about the same, other than lower temps of course...Also, water cooling has the same limit as air, as you are still "air cooling" the water, it will never go below a given point (ambient).
 
Hypernova said:
There's also the fact that rate of heat transfer depends on the temp difference between two mediums (Newton's heat equation). Air at room temp and CPU at 50c means you will hit that barrier pretty quick. On top of that air have smaller heat capacity. At that point you are better served with water or chilled air at least.
Click to expand...
It's not really a 'barrier'. Heat transfer is proportional to temperature differences, which - using newton's law of cooling / fourier's law - you could interpret as meaning, "The colder your heatsink fins are, the more heat is transferred", so it really is in your best interest to keep those fins cold by blasting them with air.

Although in reality here, we're dealing with a 'fixed' heat load. Remember that the generalized heat equation is Q = UA(Th-Tc). Since Tc (cold/ambient temp) and Q (heat flow) are fixed by our operating conditions the only way to reduce Th (hot temp) is to increase A (surface area) or increase U (generalized heat transfer coefficient). U represents a combination of things, including the conductive resistance within the chip itself, the IHS, and the heatsink, and also the convective resistance between the heatsink fins and the cooling fluid. This convective resistance (typically represented as h) is usually based on Nusselt Number, which is a dimensionless quantity itself based on a variety of factors including fin geometry and fluid Reynolds Number, which represents the turbulence of the fluid flow. For forced convection over a flat plate, h is nearly linearly proportional to reynolds number, which means that more airflow increases h, which increases U, which lowers your chip temperature. See section 5 of http://www.cheresources.com/convection.pdf for a relationship between Nusselt number and Reynolds number for flat plates.

This doesn't always work out as expected because heatsink designers try to add many fins to increase the surface area. When the fins get closer together, it shrinks the turbulent flow boundary layer where a lot of convective transfer occurs. More airflow can help preserve the turbulence, but only to a point. If the fins are really close together then all the airflow between them becomes laminar and increased flow does very little. For larger fin-spacing, it takes less airflow to get to this point where the laminar boundary layer is already minimized, so high CFM helps even less here. It's for this reason that high CFM delta's are only marginally better than conventional fans.

Please correct me if I've gotten my facts mixed up.
 
Last edited:
Kaldskryke said:
It's not really a 'barrier'. Heat transfer is proportional to temperature differences, which - using newton's law of cooling / fourier's law - you could interpret as meaning, "The colder your heatsink fins are, the more heat is transferred", so it really is in your best interest to keep those fins cold by blasting them with air.

Although in reality here, we're dealing with a 'fixed' heat load. Remember that the generalized heat equation is Q = UA(Th-Tc). Since Tc (cold/ambient temp) and Q (heat flow) are fixed by our operating conditions the only way to reduce Th (hot temp) is to increase A (surface area) or increase U (generalized heat transfer coefficient). U represents a combination of things, including the conductive resistance within the chip itself, the IHS, and the heatsink, and also the convective resistance between the heatsink fins and the cooling fluid. This convective resistance (typically represented as h) is usually based on Nusselt Number, which is a dimensionless quantity itself based on a variety of factors including fin geometry and fluid Reynolds Number, which represents the turbulence of the fluid flow. For forced convection over a flat plate, h is nearly linearly proportional to reynolds number, which means that more airflow increases h, which increases U, which lowers your chip temperature. See section 5 of http://www.cheresources.com/convection.pdf for a relationship between Nusselt number and Reynolds number for flat plates.

This doesn't always work out as expected because heatsink designers try to add many fins to increase the surface area. When the fins get closer together, it shrinks the turbulent flow boundary layer where a lot of convective transfer occurs. More airflow can help preserve the turbulence, but only to a point. If the fins are really close together then all the airflow between them becomes laminar and increased flow does very little. For larger fin-spacing, it takes less airflow to get to this point where the laminar boundary layer is already minimized, so high CFM helps even less here. It's for this reason that high CFM delta's are only marginally better than conventional fans.

Please correct me if I've gotten my facts mixed up.
Click to expand...

Yes i guess barrier is the wrong choice of word. Although expanding from your explanations it is easy to see how water cooling helps.

Given limited space on the CPU Q = UA(Tc-Tw). Given that water has higher heat capacity U goes up. And then you will get the best result with the lowest Tw possible. On the rad side given effectively no space constraints you can jack up A in Q = UA(Tw-Ta) as much as possible. This enables lower U with slower fans.
 
Hypernova said:
Ha, little girls gagaing over puny 200cfm fans.

How about 252.85 cfm and PWM controllable from your M/B.

Or if you just want something that is 12cm:

S-Force 11000 rpm with 5inH2O pressure

AFAIK this is the most powerful 12cm fan in the world, but your normal computer PSU won't drive it.
Click to expand...

What the hell is it made out of, stainless steel?

k1pp3r said:
OP: These are faster and less expensive

I would suggest ear protection
Click to expand...

Meh, when people write server grade fan as description for this they weren't kidding. High cfm, high static air pressure and high noise...and no way to undervolt these directional delta fans without enjoying the annoying high pitched motor whine.

I think a lot of people overlook Deltas AFB line in favor of the louder and higher cfm PFB/TFB/etc lines. I didn't even know about it until two weeks ago when I saw a delta AFB1212SHE fan in a radiator fan review. Bought me two of these and really satisfied so far. Not uberdelta loud in anyway and two of them are more then enough for Megahalems in push pull config, easy to undervolt with Lamptron FC-2 however for some reason if I only connect one of them per channel the fan doesnt start up. I need to connect two of them for the fans to work or a combo of any other fan(s).
 
Last edited:
k1pp3r said:
Dang, these things remind me of the 60mm Vantec Tornado lol.

.
Click to expand...

Oh man don't remind me. Delta had a small black label fan back in the day that was loud as hell. Cooled good with the copper Alpha heatsinks at the time though :p
 
You must log in or register to reply here.
Back
Top