shadowscotland said:My 2p worth
after 20sec the case reaches equalibrium: (numbers are unitless)
10 in + 8 out + 2 losses (via holes) = 10 'flow' (high pressure case)
10 out + 8 in + 2 gains (via Holes) = 10 'flow' + dust (low pressure case)
The 'flow' is the same, negative pressure has more dust.
[slightly OT]
Air flow routes are more critical:
A direct 'flow' of 8 will have same cooling potential that a turbulent 'flow'' of 12.
.5 (distance) x 100 (temp gained) / 10 (flow) = +5 Case temp (baseline)
To reduce temps, reduce one or more of the three factors.
Less distance for air to travel (more planning/better case)
.4 x 100 / 10 = 4
Use a cooler chip (AMD / C2D)
.5 x 80 / 10 = 4
More air flow (Bigger/faster fan)
.5 x 10 / 12.5 = 4
Back to my example above to acheive base line temp.
.4 x 100 / 8 = 5 (A direct flow of 8)
.6 x 100 / 12 = 5 (turbulent flow of 12)
Practical usage:
So you've oc'ed your chip, 50% more heat .5 x 150 / 10 = 7.5 (to hot)
You have various options.
.5 x 150 / 15 = 5 (an extra 50% more air flow)
.4 x 150 / 12 = 5 (an extra 20% more air flow well routed)
.4 x 125 / 10 = 5 (Improved cooler plus well routed flow)
.01 x 150 / .5 = 3 (liquid cooling direct on heat source, water flow 1/20 of air)![]()
[/slightly OT]
Strewth I'm sober and I don't get that at all!
![Stick Out Tongue :p :p](/styles/default/xenforo/vbSmilies/Normal/tongue.gif)
Seriously though folks, like said earlier, trial and error is the best bet. I personally prefer more pulling out with negative internal pressure. From what I've seen the difference is minimal anyways. I do think its agreed that a coupla 120mm fans are gonna be better than a coupla 80mm fans though, more air and quieter running.