Why haven't we seen HSF units that...

[zapato2k]

When you blow air into the top of the HSF, you're basically pushing out all of the other air around the place. When you suck air out of the top, you're basically pulling it in around the place.

Now, when we consider case cooling, I see Fans all over the place, Exhausting and intaking air. I don't see why HSF manufacturers have been restricted to either or.

Bottom line is, i guess... why isn't there a HSF that's like... A large Fan intake on the top, say, 20 CFM, and then two smaller fans on the side of the heatsink which exhaust at 10 CFM each.


What's wrong with this idea, and if there's nothing wrong with it, why aren't manufacturers putting this into play? I'd say it's a lot less expensive than making a silver plated HSF would be.

[Hoot]

Actually, it's not a bad idea at all. In fact it has been brought up before. Using a push-pull setup makes very good sense on the HS as well as it does when cooling a radiator in a water rig. I think the biggest stumbling block is the lack of room around the HS where you can mount two assist fans with some room for the air to move to or from them without immediately running into the PSU or case wall. In a full tower setup, that may not be nearly as restrictive as in a mid tower. The other issue is the height of the HS may be shorter than the minimum size of a fans that would deliver adequate cfm to equal the fan on the top. I seem to recall Joe C. just reviewed a tall HSF on the front page a day or two ago. That would be a good candidate for a push-pull setup. You want the two assist fans CFM to equal the CFM of the fan on top.

Hoot

[AmbientFiction]

Well I posted something like this awhile back. I think it would be something to see but when you look at the fact of haveing the exhaust fan where is it in the back of the case you wouldn't get too mcuh pull from the fan on the back of the haetsink. They should be made this way.
**Note the top fan has the arrow pointing the wrong way sorry about that**

[zapato2k]

Actually, that wasn't quite the idea that I had. My version would be something like this:

[wildbilly2k]

It probably wouldnt even fit on alot of mother boards even if they did make them so they wouldnt be very big seller's.

[cjtune]

The main advantage to using the suck & push fan combo is that all of the fans will maintain pretty steady cfm drop. Normally, axial fans mounted in parallel (like AmbientFriction's) push alot of cfm but mostly in free-delivery situations. Once a fan's output is impeded by, say, dense heatsink fins, the flowrate will fall drastically, unless you use a fan like a Delta38 which delivers high cfm with high pressure (AND noise!) and at a high power requirement (4W?). You can 'emulate' the pressure-supplying characteristic of a Delta38 by combining axial fans in series. More pressure means less significant drops of cfm. Give it a try. However, fans are costly and will add to a HSF's price tag and mfgs would rather just stick only one for every HS.

[zapato2k]

check out the thickness of the fan on your video card.

THAT's the thickness I'm talking about. That shouldn't add too much width...what, 1.5 inches, MAX. Half an inch for each fan, and maybe some more modifications to the body of the HSF to accomodate for these fans, for another half an inch.

[zapato2k]

Originally posted by cjtune
...unless you use a fan like a Delta38 which delivers high cfm with high pressure (AND noise!) and at a high power requirement (4W?). You can 'emulate' the pressure-supplying characteristic of a Delta38 by combining axial fans in series. More pressure means less significant drops of cfm.


Well, the system currently employed is a fan pushing in... that could only be aided by having additional fans, sucking out old air, and creating less pressure inside, so that more air is drawn in.

Notice the 2 * 11 CFM on both sides...and the 20 CFM on the top (btw, these CFM values are arbitrary). That's 20CFM intake, 22CFM exhaust. That should create a negative pressure of air, so and would aid in the intake of air.

But, these additional fans aren't just to aid minorly in the intake of the main fan, it would be too expensive, and too little a change. The main point in this idea is to change the air in the heatsink quickly, so that the air is cooler, and can conduct more heat.... the other effect is merely a positive side effect.

[AmbientFiction]

Well to do that you would either have to have a spacer for the fan or wire it backwards cause of the blades hitting the heatsink

[Über~PhLuBB]

I'd be most worried about having 3 fans tearing my CPU from it's seating. If you use 60MM fans on 3 sides, I suppose this might not be too much of a problem, but is a 60MM fan enough?

[cjtune]

Originally posted by zapato2k



Well, the system currently employed is a fan pushing in... that could only be aided by having additional fans, sucking out old air, and creating less pressure inside, so that more air is drawn in.

Notice the 2 * 11 CFM on both sides...and the 20 CFM on the top (btw, these CFM values are arbitrary). That's 20CFM intake, 22CFM exhaust. That should create a negative pressure of air, so and would aid in the intake of air.

The cfm ratings (note: at free-delivery) do not give much indication of a fan's pressure-supplying capability. As a rule, higher rpm, thicker case fans will have higher pressures than another slower one of the same size. The smaller 11cfm fans you plan to use may actually be overwhelmed by the larger fan's pressure, and this, to the point of view of the larger fan, is impedence. Try matching fans of the same make when using them in series.

[b]But, these additional fans aren't just to aid minorly in the intake of the main fan, it would be too expensive, and too little a change. The main point in this idea is to change the air in the heatsink quickly, so that the air is cooler, and can conduct more heat.... the other effect is merely a positive side effect.

The performance of a fan IN A SYSTEM is subject to all other components in the system, including other fans. You cannot just expect a simple superposition of contributions.

Just to clarify things up again:

Two fans side-by-side will deliver lotsa flow at low pressures but will suffer at high pressures.

Two fans in series (but not stacked!) produces the flow of just one fan but remains about the same even though at high pressures.

Heatsinks with dense fins (eg. Gladiators, SK6) present an impedence to fans and this is reflected by the higher pressure requirements for the fans.

[fuzzba11]

I've got a similar setup to the first picture posted on my CAK 38, with my Delta blowing upward to the top of my case, and a thin default fan from a duron heatsink blowing down. Works perty well, I needed to change my config cause the CAK38 barely fit between my processor and PSU. I've got a new trick up my sleeve, though, and I'm sure everyone will be interested when I post it, but I want to do some testing by myself, first...yes, you should be very very curious, cause this is the best cooling solution I've found. :p

[The Overclocker]

the alpha pal66 or whatever it is called does that,

[TUK101]

I did something similar to that with my golden orb. Broke and haveing a few fans laying about, i screwed on onto each side, and then took the small fan out of the middle and screwed a larger fan on top, therefore pushing air into the heatsink, and pulling it out with the two on the sides of it. It worked pretty well considering it was in the beginning a crappy orb lol. Good idea, and I think that I have saw HSF's working on that same principal as well.

[Intraveinous]

Yeah... hehe... when I first got the old MC462A, I was too lazy to take the mobo out right then since I hadn't gotten the rest of my case modding stuff and wanted to do it all at once. I took the center fan out of my chrome orb and stuck the 68cfm delta 80mm on top of it instead, got a 4C drop in temps on my Duron 700@900.
Needless to say, it performs much better with the MC462 under that delta 80mm than with the orb, but it was a quick fix that gave a decent drop in temps.
Peace
John