On Mon, 21 Jun 2004 19:54:43 GMT, in alt.comp.periphs.mainboard.asus,
"Travis King" wrote:

I have three questions regarding case fans even though I know that this is
an Asus newsgroup. Should the back case fan blow in air or blow out air?
Should the case fan in the side panel blow in air or blow out air? Should
the case fan on the top blow in or blow out? Help would be appreciated.
Thanks.


Do you want the "conventional approach" answer, the "in a perfect world"
answer, or the *true* answer?

Taking the last one first:

There are a lot of wrong ways to cool a computer; but there is no one "right"
way. Too much depends on the specifics of the particular case, the equipment
(including the fans themselves) housed in that case, and the conditions under
which it will be used.

Ideally ("perfect world"), you would not need *any* fans, nor any case vents,
to keep the internal system components adequately cool for reliable operation
and long life. The case would in fact be hermetically sealed, thus best
protecting those internal components from such environmental hazards as dust,
moisture, corrosive pollutants in the atmosphere, etc.

But we don't live in a perfect world; so the next-best thing to that is to
have the minimum number/size/etc. of fans needed to accomplish the following
goals:

1. A constant and adequate intake supply of fresh, clean, ambient-temperature
air -- emphasis on "clean", which strongly implies proper filtration.

2. Efficient evacuation of air warmed by convection/radiation from the
heat-producing components (note that "efficient" does *NOT* necessarily mean
"more is better").

3. An unrestricted and more-or-less even distribution of airflow throughout
the case, especially around the most heat-producing components, with no "dead
spots" where the air is permitted to stagnate.

4. Positive internal pressure maintained in the case at all times, so as to
not suck in dirty air through misc. unfiltered vents and gaps.

5. Silent or near-silent operation.

Now, this might seem like a simple and straight-forward list; but in practice,
it is actually rather difficult to fully accomplish all of these goals
simultaneously. The old original IBM PC and PC/AT case/fan designs did in
fact directly violate several of these principles; yet, because it was Big
Blue who did it, for a long time most everybody mindlessly copied them and
hence repeated the same mistakes. More recently, it has become fashionable
(especially in the "I think I know everything" youth market) to stick as many
fans in/on a case as the designer can manage to cram in; but typically, little
thought is given to what is really being accomplished (besides making lots of
racket) with all this monkey motion.

A reasonable compromise, given a typical "tower style" case, is to place one
large-capacity (or sometimes two medium-capacity) intake fan(s) as low on the
front panel as feasible, behind a good high-efficiency (meaning it will trap a
high percentage of the very smallest particles) washable/replacable-element
filter. Then use the fan(s) built into the power supply unit (mounted at the
top-rear of the case) to forcibly exhaust the now-heated air (as you probably
know, the warmest air will naturally migrate to the top of any enclosed
vessel), plus a small fan mounted directly on the CPU's heat sink to ensure
that air does not stagnate around this most critical (and high heat-producing)
item. If the case, fans, and various heat sinks scattered throughout the box
are *properly* designed (which inescapably implies recognition of the *whole*
system design), this should be all you *really* need. If you need to add
still more fans in various other locations, it is a tacit admission that
something is not right with the basic thermal design of the system.

Further notes: The addition of "extra" fans may in fact be counterproductive,
depending on the specific system/case design. It is quite possible to create
am "opposing forces cancel" situation, where two fans are effectively trying
to blow air into the same area from different directions; this can cause the
actual airflow through the subject area to *decrease* (with a concomitant
localized increase in both pressure and temperature). Secondly, all other
things being equal, more fans == more noise, every time. Some thoughtlessly
"hot rodded" systems with a half-dozen or more case fans sound like a freakin'
jetliner on take-off. Think long and hard before adding more fans than you
*really* need, if you want to be able to live with the system long-term.
Also, the faster you spin a fan, the more air it will move -- but also, the
more noise it will make. As a general rule, a larger fan spinning slower will
move more air than a smaller fan spinning faster, and be MUCH quieter while
doing it. Hence, I prefer *one* good 120mm intake fan at the lower front, as
opposed to two 80mm or 60mm fans. Similarly, most power supply units use two
smallish (80mm or 60mm) fans, one on the bottom of the PSU case, one at the
rear; but other PSUs, such as some from Fortron Source, use a single 120mm fan
on the bottom, then simply allow the air to be exhausted through a highly
ventilated rear panel. This latter approach should always produce more
airflow per unit of noise, and probably better overall cooling as well (tho'
this latter issue is somewhat dependant on the specific supply designs).
Finally, in order to maintain a slight positive internal pressure, you MUST
blow more air into the case than you're sucking out of it. The benefits to
doing it this way are manifold, especially in the less-than-pristine
environments (like under your desk) that most systems tend to be used in.

--

Jay T. Blocksom
--------------------------------
Appropriate Technology, Inc.
usenet01[at]appropriate-tech.net


"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
-- Benjamin Franklin, Historical Review of Pennsylvania, 1759.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Unsolicited advertising sent to this E-Mail address is expressly prohibited
under USC Title 47, Section 227. Violators are subject to charge of up to
$1,500 per incident or treble actual costs, whichever is greater.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -