Need better air cooling in my HAF912

[BonzaiDuck]

My HAF 912 will still fit a 120mm...Better than nothing :-/

Generally speaking the HAF cases all get great reviews on newegg. I'm assuming these people are running tests to see what their temps are.

Also Charlie, you've got to take into account the air temp that you're actually pulling in as well. That will make a large difference..But it seems like in your example, switching cases helped.

I actually work with a bunch of engineers (I'm one myself) and I can ask them if theoretically more open space would be worse. Based on my limited knowledge, I would say it depends. It's more about how impeded the airflow is...and probably less about the "openness" of the case.

No disagreement there. You have any number of options that won't "ruin" your case if proceed with care. For instance, you could pick up a spring-loaded punch to mark the points to drill numerous, evenly spaced holes that would widen your side-panel vent and convert the 912 to a 922. Pick the right 200mm fan and drill mounting holes for it.

On the size or interior capacity of a case: Larger doesn't necessarily mean better if you've merely created a large reservoir of air with dead zones. Whatever the case, whatever the cooling strategy (air or water), you want to push air over components that generate heat. The narrower the aperture, the greater the velocity of air-flow, the better the cooling. And that can be done in very close confines or a more compact case. But with less space, it takes perhaps more thought, planning -- trouble.

 

[aigomorla]

Cliffs for the AT article:
-CPU water cooling didn't provide substancially more cooling than a good HSF (in this case, Noctua U14 I believe)
-GPU water cooling very effective; halfed the temps under load

I found reading some of the comments helpful too to understanding maybe why these test results seemed odd. Interesting to me that air cooling the CPU in this case was almost as good as water. I'll have to read more articles.

Water is based on the principle of equilibrium. However...
Water can only hold so much before the temp rises...

Math/Physics states:
1.5Gal/Min ~ 350W of heat water can contact before it goes up 1C.

780GTX states it has a TDP of 250W, which means, 2 in SLI would be doing 500W.

The average loop can only do about .7 Gal/Min on a single pump, for a multi component LC system.. (CPU+GPU+Radiator)

The Point of diminished returns on a loop is around 1.5GPM, which is means anymore flow past that point = a exponentially small return for the exponentially expensive cost.
(looking at non overkill heat loads)

So what did he have? Something called a heat gradient in which equilibrium was trying to keep up but couldn't.

Size of Reservior has no impact on performance outside Holding capacity.
Meaning the more coolant u have, the longer the system takes to reach equilibrium.

Id honestly like to see an Air setup keep up with my LC setup, which is physic/math/thermodynamic impossibility.
There is no air sink in existence which has the surface area equal to PA120.3 which i have dedicated sole to my cpu.

However my CPU loop is and was very expensive... (+/-) ~ $400 dollars just for that single loop.

Yeah I felt like the set up wasn't as good as it could be. I kept researching and saw a similar hardware setup, except the user had a 1080mm rad and a 240mm rad. There was also a seperate pump so I'm guessing that was pretty beefy. Looked very costly though..I'm guessing that set up would run you $600+.

I am not a fan of a single loop with a heat load which exceeds 350W for reasons stated above.
I am also running Tri SLI setup with all 3 of my GPU's never breaking 37C at a ambient of 22-24C.
However I have 4 radiators (1xPA120.3(CPU)+2xMCR320(GPU)+PA160(Motherboard)), 6 pumps (4xDDC's 2xD5) on 3 independent loops (all which are @ or above 1.5GPM),

More math for you, the Delta from coolant vs ambient delta on my cpu loops is 1-2C.
The coolant delta on my gpu loop is no greater then 3-4C.
The coolant delta on my motherboard is no greater then 2C.
(this means the temp at air going inside the radiator vs the actual water temp)

Coolant delta is what we call the efficiency on a system.. the average mid tier system will do about 7-10C delta, while the high performance do about 5C.
Then you have your extreme setups which do below 3C.

Well i dont want to derail the thread anymore as its not pertaining to LC, so i'll stop here... but i really dont like the results neophyte had on his system.

 

[BonzaiDuck]

[I'm SURE there's a forum member named neophyte. He just didn't post to this thread. So . . . what were his results?]