test of thermal pad on AMD

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Tue, 18 May 2004 17:01:18 GMT, kony <spam@spam.com> wrote:

>...
>Remember that they are not concerned about "optimal", only (proven)
>adequate cooling. AMD provides a 'sink and interface that is adequate
>if chassis airflow and ambient temp is reasonable. People report 55C CPU
>temp but that is not too hot for a CPU running at stock speed. Through
>more effort the CPU temp may be lowered to 40C, but without specific
>reason for doing so it was just a waste of time. However there are good,
>specific reasons for doing so, for example if the thermal interface is
>more efficient then it allows somewhat lower airflow, lower
>noise/wear/dust for same resulting CPU temp... so for that example the
>goal wasn't even lower temp, just quieter operation with temp remaining
>within margin for stability. Another reason may be overclocking, but
>that's a whole topic in itself.

I just put a thermalright heatsink and 92mm panaflo high speed fan on
an amd64 2800 (1.8ghz). Can you say overkill ? LOL. Luckily the
Aopen AK86-L motherboard does have silent tek that monitors cpu
temperature and can automatically adjust the cpu fan speed. The
latest bios also enables cool n quiet technology. I get sub 30C idle
temps and turn down the fan to about 80% with just the silent tek.
Haven't played with the cool and quiet yet, but not bad for plain
white goop.

 

[Stacey]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

kony wrote:

> Through
> more effort the CPU temp may be lowered to 40C, but without specific
> reason for doing so it was just a waste of time. However there are good,
> specific reasons for doing so, for example if the thermal interface is
> more efficient then it allows somewhat lower airflow, lower
> noise/wear/dust for same resulting CPU temp...

That was my goal. After tweaking the heatsink I was able to drop the fan
speed 1000 RPM with no cooling problems.
--

Stacey

 

[Guest]

Archived from groups: alt.comp.hardware,alt.comp.hardware.amd.thunderbird,alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

> Such measures are applied in industrial installations (like in electricity
> cabinets of large machines in a factory - where water sprinklers are not a
> wise anti-flame option). In non electrical / non chemical environements
> water sprinklers remotely and automaticly switched open (for EE guys, open
> means there's a current of water) by a relay connected to the fire / heat
> alarm. Im sure there are trip-switches that also disconnect elevators and
> such when there's a building fire...
>

Seen a LOT of panels/Other for the Nuclear & PetroChem industry's (cant say
I ever seen any automated fire ext kit for a sole peice of electronics)..

In reality, If something is allowed to get hot enough to burn your house
through an electrical fault (dead short) then it means that the consumer
unit is not properly set up. Once a fire is started removing electricity
(for most things that we have) will not stop the fire. Normally IF there is
a fire you would combat the whole area and not just the specific fire point
of detection (speaking about automated fire systems here).

But I would need to wonder HOW his pc burning (probably have 5-10 fuses
before it gets to the power strip) will spread to other things in the house.

Metal (normally metal) cases generally dont burn well, nor does the things
inside of a case.

My suggestion, Get proper electrical wiring.

 

[john]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Mon, 31 May 2004 00:50:45 +0100, "rstlne" <.@text.news.virgin.net>
wrote:


>
>But I would need to wonder HOW his pc burning (probably have 5-10 fuses
>before it gets to the power strip) will spread to other things in the house.
>
>Metal (normally metal) cases generally dont burn well, nor does the things
>inside of a case.
>
>My suggestion, Get proper electrical wiring.
>

I read an article about the accumulation of cat5 wire in office
buildings as successive occupants add their own cables, but leave the
old stuff in place. Apparently the insulation on the wires spreads
fires all over the place. I guess they need better insulation fire
ratings.

a bit off topic, just a FYI 😉

 

[Guest]

Archived from groups: alt.comp.hardware,alt.comp.hardware.amd.thunderbird,alt.comp.hardware.homebuilt,alt.comp.hardware.pc-homebuilt (More info?)

"rstlne" <.@text.news.virgin.net> wrote in message
news:ljvuc.595$cs4.259@newsfe4-gui...
> > Such measures are applied in industrial installations (like in
electricity
> > cabinets of large machines in a factory - where water sprinklers are not
a
> > wise anti-flame option). In non electrical / non chemical environements
> > water sprinklers remotely and automaticly switched open (for EE guys,
open
> > means there's a current of water) by a relay connected to the fire /
heat
> > alarm. Im sure there are trip-switches that also disconnect elevators
and
> > such when there's a building fire...
> >
>
> Seen a LOT of panels/Other for the Nuclear & PetroChem industry's (cant
say
> I ever seen any automated fire ext kit for a sole peice of electronics)..
>
> In reality, If something is allowed to get hot enough to burn your house
> through an electrical fault (dead short) then it means that the consumer
> unit is not properly set up. Once a fire is started removing electricity
> (for most things that we have) will not stop the fire. Normally IF there
is
> a fire you would combat the whole area and not just the specific fire
point
> of detection (speaking about automated fire systems here).
Electrical fault can also be an arc - which means there's an ongoing spark
and heat source. Generally speaking, most of electrical components are in
some way or another flame retraded (all PCBs have UL-94 V-0 marking),
electric outlets should be FR-HIPS or melamine, cables are (intrinsically)
flame retarded PVC etc...
>
> But I would need to wonder HOW his pc burning (probably have 5-10 fuses
> before it gets to the power strip) will spread to other things in the
house.
>
> Metal (normally metal) cases generally dont burn well, nor does the things
> inside of a case.
Aluminum burns like hell(fire) ;-) steel (or copper) doesn't. Think about
airplane burning - it's not just the fuel, it's the whole fuselage.
>
> My suggestion, Get proper electrical wiring.
Agreed. And standard components (UL, CE, TUV marks)
>
>