Hello
  While I agree that turbulence can conceivably be a problem (tho not
always... sometimes it is even a benefit especially around the fins of the
hs) 1) Even w/o going to the trouble of using smoke and a clear cover much
can be done to avoid this issue just by logic ie tying down cables, clean
layout etc. and 2) Turbulence is far less fearful than lack of CFM, moving
quantity of air.  I realize this is not Felix's point since he is properly
touting real world testing as the final arbiter but I just wanted to make
sure it wasn't taken for granted to simply avoid moving lots of air.

  Additionally, I'm reasonably certain that Thermal Resistance
(Conductivity) is one of the rare specs that can basically stand alone
since it simply measures a physical attribute and condition.  It's
application can be used regardless of materials used, including coolant,
mass etc since those are taken into account by this form of spec which is
not about intensity, temperature directly, but rather about the quantity of
any thermal units which *can* move as a result of the thermal resistance
which is assumed to be a negative function, almost like transistor
"holes".  What we do in application of that thermal resistance or
ability to "pull" calories, BTUs whatever a small distance away
from the CPU into the heatsink and then whisk it away so that ambient
doesn't continue to rise until unity is reached is what ultimately counts.
Witness the deleterious effects of Peltier's which have amazing thermal
conductivity but "pollute" the thermal environment so badly that
they cancel out what
ever value it may have in most systems and often *cause* overheating. 

  Personally I find it easier to deal with these issues separately as a 2
step process  I tend to seek out the lowest thermal resistance heatsink I
can find that fits size and cost constraints and then use as phsically
large a fan as I can make fit to be able to keep rpms down at reasonable
CFMs.  Most manufacturers publish fairly useful specs regarding sound
pressure levels in db.  However it is also imperative to get as close to
ideal transfer as possible in the real world so mating surfaces and proper
application of a good "thermal grease" is essential.  Recently I
read that Apple recommended using an entire tube of grease on *each* of
CPU, GPU, and chipset!!! This is counter productive since grease has lousy
thermal conductivity, it's just better  than air pockets.

  And, like Felix says, ultimately I bask in low temps and the stability
that provides.  For example I have a Slackware Linux box that I use daily
(this one, in fact) which sports a 1.4Ghz Tualatin Celeron overclocked (and
slightly over "cored") up to 1.68GHz which has an uptime of over
39 weeks, and were it not for a power failure that lasted over a half-hour
a few weeks ago, it'd be more than that.  That may not be awesome uptime
but it ain't bad especially considering it's usage and overclocking.  BTW
it runs 42C at idle and only rises to 50C under Prime 95 Torture Test even
if left running for days.  Decent, I think. It was worth it to me to polish
surfaces, use good grease and application techniques, and maximize CFM
while keeping noise reasonable.
Jimmy

 -------------- Original message ----------------------
From: inkleput{at}isp.com
> Felix Miata  said:
> 
> >The proof is in the CPU temp, not the CFM. Excessive airspeed can
> >result in turbulence that reduces effectiveness. Until you measure
> >temp, you can't know for sure whether any cooler is any better than any
> >other.
> 

> I think this large cooler is meant for a different BIOS.  It lists
> online as having a 300 RPM - 2500 RPM range.  This BIOS shows it at 2812
> to 2860 all the time and I know it doesn't whine nearly so high pitched
> (thank God) as the last one that was supposed to vary 2100 to 2700.

As long as it does a proper job, it's fine regardless of what the bios
"expects".  If it provides insufficient heat tansfer reflected in
higher temps or too much noise, once you are certain of mating surfaces and
cleanliness of air path, the fix is obvious - larger, slightly slower fan
or better made equivalent.
> 
> Jim L, via eCS 1.24 version of OS/2
> -- 
>



[Non-text portions of this message have been removed]



 
Yahoo! Groups Links

 To visit your group on the web, go to:
    http://groups.yahoo.com/group/os2hardware/

 To unsubscribe from this group, send an email to:
    os2hardware-unsubscribe{at}yahoogroups.com

 Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 



---