[NTLUG:Discuss] PC Hardware Issue

[Gilbert Morrow]

Wow , I think you missed the point , trouble shooting over the net is hard
to say the least , if you were where you could be in front of the computer
then the problem would some what be easy to see or trouble shoot . In my
opinion I saw nothing in his pictures that was to wrong looked normal case
setup , front fans pull , rear push , pretty
simple sticker on fans face air flow indicator , sticker facing
in pushing sticker facing out pulling .
I found in my OCing days Winders would crash and SuSE would boot and run ,
so in my experience
Linux (Linux tends to put much less load on key components) is more stable
when some insane OCing is going on .
ATI drivers in Linux are still crap , I tell everyone get Nvidia if you can
, but since AMD has ATI now ATI may get a little more Linux friendly (hope)
.


On 12/24/06, steve <sjbaker1 at airmail.net> wrote:
>
> Robert Pearson wrote:
> > On 12/21/06, Neil Aggarwal <neil at jammconsulting.com> wrote:
> >> Robert:
> >>
> >> I have never heard that too much airflow could be a
> >> problem.  Please explain.
> >>
> >> Thanks,
> >>         Neil
> >
> > I should have said that "too fast" airflow could be a problem.
> > If you have a lot of airflow through a case, there is a chance that
> > the air may not stay in contact with the heated surfaces long enough
> > to remove the heat.
>
> This is urban legend...there are no possible laws of physics that
> support such a claim!  Read about Newton's law of cooling...no
> mention there of speeds.  Rate of heat loss is proportional to the
> square of the temperature difference...period.  The air in contact
> with the surface warms up thus reducing the temperature differential
> and because it's a square-law dramatically reducing the heat flow
> out of the surface.  Replacing that warm air with cooler air helps
> the heat flow.  The faster you replace it, the less heat it has
> absorbed before it's replaced and therefore the more heat you can
> suck out of the surface.
>
> > A lesser problem this can cause is turbulence,
> > which can create "hot spots".
>
> ...or lack of turbulence which can cause stagnation between
> components or close-packed boards.  The aerodynamics of each
> component and between boards and around cables is vastly complex
> and predicting the location of hot spots as a function of air
> speed is almost impossible.
>
> Places where the air stagnates can be a problem - but generally, if
> they get hot then convection will replace the air with cooler air
> from elsewhere.  The question is whether that's enough - sometimes
> it is, sometimes it isn't.
>
> It's very hard to tell.
>
> But we're getting way ahead of ourselves here.
>
> Right now though we don't even know that this is an overheating
> problem at all!
>
> You need to do the test of getting a big household fan turned up
> as fast as it'll go - and direct it right into the case (with the
> lid off).  If the system still craps out with that much airflow
> then you'll KNOW it's not to do with overheating.
>
> Buying silver paste and messing around with moving boards and
> repositioning cables is what you do AFTER you know you have a
> heating problem.  But you are still in a position where it could
> be:
>
> 1) Software issues with crappy ATI drivers.
> 2) Bad memory stick(s)
> 3) Bad graphics card
> 4) Bad motherboard
> 5) Insufficient power supply to cope with peak demand.
> 6) ...oh, yeah...and overheating.
>
>
> _______________________________________________
> http://www.ntlug.org/mailman/listinfo/discuss
>