HardwareBanter

HardwareBanter (http://www.hardwarebanter.com/index.php)
-   Gigabyte Motherboards (http://www.hardwarebanter.com/forumdisplay.php?f=14)
-   -   System vs. CPU temperature (7VAXP) (http://www.hardwarebanter.com/showthread.php?t=38077)

Orange Barrel April 20th 04 05:39 PM

System vs. CPU temperature (7VAXP)
 
The system temperature on my computer is always higher than the CPU
temperature. For example, the current CPU temp is 41.1C, while the system
temp is 47C. The relative difference remains constant even under load.

From a thermodynamic standpoint, is that even possible?

I asked GB tech support if the sensor display was reversed, and they
assured me that the temperature difference would be "normal if you got a
slower CPU and better heat sink fan."

I'm running an XP 2700+ (stock timings) with a CoolerMaster HHC-001 H/F on
a 7VAXP Rev 1.2. It's inside an Antec Sonata case (380W TruePower P/S)
with a 60GB and a 200GB hard drive. All three mem slots are occupied
(Kingston PC 2700 512 MB).

I've searched for an answer to this and gotten nothing but conflicting
information. Any help would be appreciated.

While I'm on the subject, how often should the H/F be reseated and thermal
paste reapplied? Two months ago, my CPU temp was 36C idle, and now it
idles at 40-41C. I work in a relatively dusty environment, so some of it
is attributable to dust accumulation. Should I be concerned about the idle
temp? Load temp now peaks at 48C CPU and 52C system.

Thanks in advance,
Orange Barrel

Henrik Dissing April 20th 04 06:27 PM

On Tue, 20 Apr 2004 11:39:57 -0500, Orange Barrel wrote:

The system temperature on my computer is always higher than the CPU
temperature. For example, the current CPU temp is 41.1C, while the system
temp is 47C. The relative difference remains constant even under load.

From a thermodynamic standpoint, is that even possible?


No, that would be in conflict with one of natures most basic laws: the
second law of thermodynamics. Heat always passes from a warmer body to a
colder body.

If your readings were correct, the surrounding air would actually heat up
the CPU in stead of cooling it down, so you would be better off without a
CPU cooler, which would only serve to speed up that process. Don't try to
remove the cooler, though! :-)

Something's wrong with your readings.
--
Best regards,
Henrik Dissing

(e-mail: hendis AT post DOT tele DOT dk)

Heckler ²°°³ April 23rd 04 01:31 AM

Orange Barrel wrote:
The system temperature on my computer is always higher than the CPU
temperature. For example, the current CPU temp is 41.1C, while the
system temp is 47C. The relative difference remains constant even
under load.

From a thermodynamic standpoint, is that even possible?

I asked GB tech support if the sensor display was reversed, and they
assured me that the temperature difference would be "normal if you
got a slower CPU and better heat sink fan."

I'm running an XP 2700+ (stock timings) with a CoolerMaster HHC-001
H/F on a 7VAXP Rev 1.2. It's inside an Antec Sonata case (380W
TruePower P/S) with a 60GB and a 200GB hard drive. All three mem
slots are occupied (Kingston PC 2700 512 MB).

I've searched for an answer to this and gotten nothing but conflicting
information. Any help would be appreciated.

While I'm on the subject, how often should the H/F be reseated and
thermal paste reapplied? Two months ago, my CPU temp was 36C idle,
and now it idles at 40-41C. I work in a relatively dusty environment,
so some of it is attributable to dust accumulation. Should I be
concerned about the idle temp? Load temp now peaks at 48C CPU and 52C
system.


I've had similar readings on my board (7VAXP Ultra), CPU temps (XP2800) were
39º and system temps were 42º. However after making some changes to the
case, adding soundproofing and ambient temps in this region have risen due
to a change in season... CPU temps are now 45º, but system temps remain
constant at 42º.

This would indicate that there is something wrong with the readings I'm
getting. I used to have an old Matsonic cheapo MB, and the system temps on
that thing used to jump around between 30-50º, and the CPU temps always read
56-58º, and upto 62º under load... I never believed them though, simply
because after swapping the CPU and Heatsink into another board in an
identicle case... case and cpu temps were 34 and 48º respectively... the
same heasink in my other PC on a faster CPU gives an even lower 45º... and
swapping heatsinks out for faster and more efficient ones (going from AMD
standard HS (alluminium) to an all copper coolermaster Delta heatpipe, and
then a coolermaster all copper aero 7) all gave identical temps... and
there's no way on earth that that's possible.


H



Dimitris April 23rd 04 10:59 AM

Orange Barrel wrote in message ...
The system temperature on my computer is always higher than the CPU
temperature. For example, the current CPU temp is 41.1C, while the system
temp is 47C. The relative difference remains constant even under load.

From a thermodynamic standpoint, is that even possible?

I think it is possible , rather very logical i would say from a
thermodynamic and fluiddynamic point of view. That is because the
processor has larger fan and it dissipates the heat away, but the hot
air goes and stays around the system chip(northbridge) which has much
smaller fan, and it is very near to the proccesor. Mine temperatures
are about the same, and yes many times the system temperature can be
up to 4-5 degrees higher(i have seen 48C CPU and 53 System after alot
of gaming.)

You dont have to worry about temperatures if they dont go much higher
than 50C under load always. Of course the summer season you should add
7-10C unless you have good air condition.

Henrik Dissing April 24th 04 03:12 PM

On 23 Apr 2004 02:59:16 -0700, Dimitris wrote:

From a thermodynamic standpoint, is that even possible?


I think it is possible , rather very logical i would say from a
thermodynamic and fluiddynamic point of view. That is because the
processor has larger fan and it dissipates the heat away, but the hot
air goes and stays around the system chip(northbridge) which has much
smaller fan, and it is very near to the proccesor.


I suppose it comes down to what is meant by "system temperature". If it
reflects the temperature of something which itself produces heat, such as
the north bridge chip, you could be right.

However, all the system temperature reporting motherboards that I have come
about had a sensor mounted in safe distance from and significant heat source
and did therefore give a good indication of the air temperature in case. The
CPU temperature was always reported significantly higher than the system
temperature.
--
Best regards,
Henrik Dissing

(e-mail: hendis AT post DOT tele DOT dk)

Dimitris April 26th 04 12:29 PM

Henrik Dissing wrote in message . ..
On 23 Apr 2004 02:59:16 -0700, Dimitris wrote:

From a thermodynamic standpoint, is that even possible?


I think it is possible , rather very logical i would say from a
thermodynamic and fluiddynamic point of view. That is because the
processor has larger fan and it dissipates the heat away, but the hot
air goes and stays around the system chip(northbridge) which has much
smaller fan, and it is very near to the proccesor.


I suppose it comes down to what is meant by "system temperature". If it
reflects the temperature of something which itself produces heat, such as
the north bridge chip, you could be right.

However, all the system temperature reporting motherboards that I have come
about had a sensor mounted in safe distance from and significant heat source
and did therefore give a good indication of the air temperature in case. The
CPU temperature was always reported significantly higher than the system
temperature.

Even if the system temperature isnt that of the northbridge chip, if
the temperature sensor is located at a point with bad air flow(out of
reach of external cool air), then heat leaves other components (cpu,
hard disks) and is accumulated to that point(this can be done with
internal flow patterns of hot air ).
I believe that the system temperature is made up as an average of
northbridge temp and another one or two temperatures sensors across
the board, or maybe other important chips like southbridge. But
because northbridge temp becomes high(due to the reason i mention in
my first message), the average becomes also high.

Henrik Dissing April 26th 04 01:32 PM

On 26 Apr 2004 04:29:12 -0700, Dimitris wrote:

Even if the system temperature isnt that of the northbridge chip, if
the temperature sensor is located at a point with bad air flow(out of
reach of external cool air), then heat leaves other components (cpu,
hard disks) and is accumulated to that point(this can be done with
internal flow patterns of hot air ).


The point is that, at a certain point, heat will stop passing from the CPU
and harddisks to that point with bad airflow.

In a closed system, nothing can become warmer than the hottest heat source.
You cannot accumulate heat just by not transporting it away, you need a heat
source hotter than the current temperature.

The temperature of that point with bad air flow will never become hotter
than the hottest of the heat sources in the cabinet, which is likely to be
the CPU.

I'm sorry, but what you propose is physically impossible.
--
Best regards,
Henrik Dissing

(e-mail: hendis AT post DOT tele DOT dk)

Dimitris April 27th 04 05:57 AM

Henrik Dissing wrote in message . ..
On 26 Apr 2004 04:29:12 -0700, Dimitris wrote:

Even if the system temperature isnt that of the northbridge chip, if
the temperature sensor is located at a point with bad air flow(out of
reach of external cool air), then heat leaves other components (cpu,
hard disks) and is accumulated to that point(this can be done with
internal flow patterns of hot air ).


The point is that, at a certain point, heat will stop passing from the CPU
and harddisks to that point with bad airflow.

In a closed system, nothing can become warmer than the hottest heat source.
You cannot accumulate heat just by not transporting it away, you need a heat
source hotter than the current temperature.

The temperature of that point with bad air flow will never become hotter
than the hottest of the heat sources in the cabinet, which is likely to be
the CPU.

I'm sorry, but what you propose is physically impossible.


You forgot the fans. They blow air and remove the heat from the cpu to
other points in the internal of the case. So things like what i
suggest can happen. Your view of the situation is rather simplistic
i.e two points A and B , A is hotter, B is cooler, and B no matter
cant become hotter than A. If you have a device which with mechanical
work transfer heat, then anything can happened, like transferring heat
from a hot to a hottest. This is physically possible and in accordance
with the 2nd thermodynamic law. Also you tried to view only by
thermoquasistatic-dynamic point of view. You forgot the fluid dynamic
point of view(fans and forced airflow regardless of temperature
difference.)

guym May 29th 04 07:54 PM

The system temperature is measured on the Northbridge with a VAXP. My sys
temp always runs lower than the CPU temp. I can't understand how the sys
temp could be higher. Use that as a relative temp and if it starts to get
unstable or averages higher over time, then I would be concerned.


"Orange Barrel" wrote in message
...
The system temperature on my computer is always higher than the CPU
temperature. For example, the current CPU temp is 41.1C, while the system
temp is 47C. The relative difference remains constant even under load.

From a thermodynamic standpoint, is that even possible?

I asked GB tech support if the sensor display was reversed, and they
assured me that the temperature difference would be "normal if you got a
slower CPU and better heat sink fan."

I'm running an XP 2700+ (stock timings) with a CoolerMaster HHC-001 H/F on
a 7VAXP Rev 1.2. It's inside an Antec Sonata case (380W TruePower P/S)
with a 60GB and a 200GB hard drive. All three mem slots are occupied
(Kingston PC 2700 512 MB).

I've searched for an answer to this and gotten nothing but conflicting
information. Any help would be appreciated.

While I'm on the subject, how often should the H/F be reseated and thermal
paste reapplied? Two months ago, my CPU temp was 36C idle, and now it
idles at 40-41C. I work in a relatively dusty environment, so some of it
is attributable to dust accumulation. Should I be concerned about the idle
temp? Load temp now peaks at 48C CPU and 52C system.

Thanks in advance,
Orange Barrel




Dimitris May 31st 04 08:16 PM

"guym" wrote in message ...
The system temperature is measured on the Northbridge with a VAXP. My sys
temp always runs lower than the CPU temp. I can't understand how the sys
temp could be higher. Use that as a relative temp and if it starts to get
unstable or averages higher over time, then I would be concerned.


It depends on which is your cpu, and how good the cpu fan is. A better
fan removes the heat from the cpu and spread it to other components in
the system, and the northbridge is the most likely to be affected, and
the one getting the biggest portion of that heat.

To clear it up i will speak with concrete numbers and give a more
detailed analysis. Say that your CPU temp is about 48C.. This means
that the air above the cpu fan has, on average, somewhat lower
temperature than 48(say 45), and the air after passing through the cpu
heatsink about equals that of the cpu(48C). Now because of the big
airflow that the cpu fan induces, this 48C temp air will dominate the
area above the northbridge fan. Now the amount of heat per second
removed from the northbridge chip is proportional to the temp
difference of the air above and under the fan, and proportional to the
quantity of air that passes through, per second. Since the latter
quantity is small(cause of the small northbridge fan) we can say that
almost solely depend on the temperature difference.. Since the temp is
48C above the northbridge fan, after blowing and passing through the
northbridge heatsink, the air temperature goes a bit higher than
48C(How much higher depends on the quantity of airflow/sec that passes
through, hence on the fan size,rotation speed e.t.c). But the
temperature of air that leaves the northbridge heatsink(which is as
explained before a bit higher than 48C) about equals that of the
northbridge chip. Hence the temp of the northbridge chip is a bit
higher than 48Ci.e a bit higher than cpu temp.

Now where this analysis goes wrong in your case so that you always
have lower system temperature. Simply you must have better ventilation
in the case so that most of the hot air from cpu goes outside and
doesnt dominate the area above the northbridge fan.


All times are GMT +1. The time now is 10:17 PM.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
HardwareBanter.com