Tal Fuchs writes:

An undersized power supply? If it is overloaded, it might overheat
and shut down, and voltages might drop at the output as the overload
develops.

--
Transpose mxsmanic and gmail to reach me by e-mail.

Ther's your problem the one blowing in, is the problem the one blowing out
can not extract fast enough. How do i know i had the same set up with my
fans.

The solution i use is this i got a 120mm fan and made the hole on the side
of the case bigger to accomadate removed the other fan. temp dropped by 5c
so from 58c at idle it is now 53c idle.

Or have both fans extracting should solve your problem as well

My CPU AMD athlon 3200XP

Here is my problem, and I'm not sure it's an over heat one

When I'm trying to play, or load other application that utilize the CPU at
100% for long time the computer is turned off and start to hear a siren.
I have to turn the electric cord off to be able to turn it on again. Turning
the computer off in the button is not enough.
In the BIOS it set up to warn when the CPU temp get to 60c, and turn it off
at 66c. No option to set anything else - System temp, Fan speeds,
Voltages....
Using FanSpeed, MBM, AbitEQ shows a CPU temp of about 58c when the computer
shut off. I get no warnings before. This is lower temp then temp set in the
BIOS.
The normal working temp is about 50c.
Using FanSpeed, MBM I noticed the - volts are wrong. -12V gives -7V only,
and -5V gives -2.5V only. These can't be seen in AbitEQ not in the BIOS.

Here is my system spec:
MB - Abit NF7-S2
CPU - Sempron +2400
PSU - 400W
GraphicCard - nVidia 6800LE AGP


Case Fans - 2* ThermalTake Smart Case Fans, one blow in and the other blow
out.

Any idea ?
Any suggestion ?


Thanks,
Tal
fuchs_t@bezeqint.net

On Sat, 24 Sep 2005 11:26:14 +0200, "Tal Fuchs"
<fuchs_t@bezeqint.net> wrote:

Overheating most likely, check the motherboard manual for
the types of events that are monitored and would produce
this siren alert.

That's way too low. Adjust the temp up higher, at least to
72C. Run Prime95's Torture Test to determine a more
accurate ceiling temp before errors are produced (In Prime95
test).

It's too close, change the temp in bios and see what effect
it has.


Are the fan air passages mostly obstructed by stamped metal
grills? If so you may need to open them up more. It's
possible to keep above parts cool (enough) with two fairly
fast running fans, but only if they're working as
efficiently as possible. The other front intake areas might
be an issue as well, for example IF the front plastic(?)
bezel further restricts air intake.

I'd not worry about those things until after increasing CPU
temp some, as there is no reason to have a system shut off
at a mere 60C when that is within it's regular operating
range. It would be a bit different if the system "never"
got above 50C, then one could expect there was some abnormal
failure to the cooling system.

Check that all fans spin and that there is no excessive dust
buildup. If you suspect poor case ventilation after
checking fans and dust levels, also note the temp of hard
drives and video card, the two other components often
susceptible to elevated case temps. If your intake fan is
on the side of the case instead of the front, it reduces
flow past the hard drives, usually.

You need to get the voltages straighteded out. If your 12 volt supply
is really 7 volts, I don't see how the thing is even running. If you
have a digital voltmeter, you can check the voltages at one of the power
plugs.

As far as heat, take the case off and run an external fan on the box.
If the temps come down quite a bit, you need to change something to get
better cooling. Of course if the voltages are off, it might be making
it run hot.

You can't see the voltages or temps in the bios? Is this because your
bios doesn't support hardware monitoring?

Clark

Tal Fuchs wrote:

On Sat, 24 Sep 2005 13:57:16 GMT, Clark <who@whoknows.com>
wrote:


No voltage could be off by that much, it is merely reporting
the wrong sensor information, which is only a cosmetic (and
use of the software reporting for informational purposes)
issue.

On Sat, 24 Sep 2005 14:41:35 GMT, kony <spam@spam.com>
wrote:


Correction- I had overlooked that these were the negative
rail voltages, and with no load on them (since most modern
systems don't use the -12V or -5V) it is conceivable that
they are in fact correct readings, but since the system is
not using them, it isn't important that they be any closer
to spec as a functionality/use issue.

Darklight wrote:

They don't work that way.

The key here is you got a bigger fan and are moving more air.

It depends on where the fans are located and if it's a typical case then
the front fan blowing in with the rear fan exhausting out is correct.

50 deg C is a tad hot for normal use, get a better H/S yesterday

Work on a basic calculation of + 10 - "15 max" deg C over the ambien
air (room temp) given to the box for an AMD cpu and + 5 - 8 deg C fo
any board in a well cooled box, duals by their nature do run about
deg C warmer than a single

Note; if you take off the side and the system stops beeping and run
cooler, your box needs fixing ASAP, faster / bigger / more fans

The system default temp warning can be dissabled, but don't, fans d
die sometimes and you won't hear a thing but the cpu going POP

All my systems no matter what they are never go over 45 deg C at 100
running 24/7, Thermalright SLK800's do the job OK

There are many good H/S's, just don't use the stock H/S supplied wit
any cpu unless it's an Intel cpu

Consider 50 deg C as MAX temp at 100% system use and work from there
to get it better

Tip; a room a/c is a very worthwhile piece of equipment that cost
less than a vid card, it cools both you and the system/s

On Mon, 26 Sep 2005 13:29:53 GMT,
digisol@optusnet.com-dot-au.no-spam.invalid (digisol) wrote:

CPUs don't go POP. They slow down when they get too hot.

LOL.

We have a 3.2 KHz P4 Prescott Retail Box and when it got to 70C we
decided to get a Zalman 7700.

Intel claims 67C is the max temp.


--

Greatest Movie Line Ever
http://home.houston.rr.com/rkba/MovieLine.wmv

"What is history but the story of how politicians have
squandered the blood and treasure of the human race?"
--Thomas Sowell

On Mon, 26 Sep 2005 13:29:53 GMT,
digisol@optusnet.com-dot-au.no-spam.invalid (digisol) wrote:

Complete and total waste of time and money.

There is no benefit whatsoever to getting the CPU below 50C,
unless you had some reasonable expectation to believe you'd
need to use it 20 years from now- to extend the lifespan
beyond that reasonably needed.



Consider that you have no reason nor data to support this
argument. Getting temps as low as possible is mostly a
tool used by overclockers to squeeze that last few dozen MHz
overclock out of a chip. There is NO CPU that cannot run at
50C, 24/7 for years at stock speed. There is no risk of
damage nor instability at 50C.

My two cents for what it's worth, having got my idle temp down to 37
from 52C in two days

Check the CPU fan is working properly. Maybe replace the therma
paste
Check your PSU fans are working (One of mine was dying - new PSU
with an extra 50 watts reduced temps dramatically

If you've ever taken your case fans off, check to make sure you pu
them on the correct way around. I hadn't ops: - correcting th
problem produced 8 - 9 degree drop

Just benchmarked my temps - Idle 37C - under stress (Prime 95 set fo
max power, max temps) 48C

Oh, and under normal load my diode temp is 4 degrees less than th
socket temp

/me one happy bunny

On Wed, 28 Sep 2005 02:25:28 -0500,
Steve@hargreavess.freeserve.co-dot-uk.no-spam.invalid (D-Dan) wrote:

What is the way you put them in correctly?

What is the diode temp and what is the socket temp?

What is the "motherboard temp"?


--

Greatest Movie Line Ever
http://home.houston.rr.com/rkba/MovieLine.wmv

"What is history but the story of how politicians have
squandered the blood and treasure of the human race?"
--Thomas Sowell

For example the rear exhaust fan on mine, after I removed it fo
cleaning, I re attached back to front, so it was sucking air in, no
blowing it out. I use a cigarette paper to test the direction of ai
flow (If it got sucked to the back of the tower the fan was on th
wrong way

The diode temp is the CPU itself, socket temp is ambient temp wher
the CPU plugs in (the two are usually about the same at idle - diod
will go higher under stress), and the motherboard temp is generall
the amibient temp in the case

On Fri, 30 Sep 2005 02:25:23 -0500,
Steve@hargreavess.freeserve.co-dot-uk.no-spam.invalid (D-Dan) wrote:

OK, so you put the front fans in to blow air into the case and the

I have seen that one - it's a small device on a spring-loaded mount.

I wonder why my MB temp rises considerably when one particular high
stress program (DVD Shrink) is run but not with any other high-stress
programs. The disk does not even get hot but the MB temps sets off an
MBM5 alarm that is never triggered except when I run this one program.




--

Greatest Movie Line Ever
http://home.houston.rr.com/rkba/MovieLine.wmv

"What is history but the story of how politicians have
squandered the blood and treasure of the human race?"
--Thomas Sowell

On Fri, 30 Sep 2005 02:25:23 -0500,
Steve@hargreavess.freeserve.co-dot-uk.no-spam.invalid
(D-Dan) wrote:

Not sure which you mean, but a rear case fan should exhaust
out of the system, not blow in. Reason being that the CPU
is not the only part that needs airflow, turning a fan
alters the entire chassis flow.

On Fri, 30 Sep 2005 13:42:04 GMT, kony <spam@spam.com> wrote:

My earlier post was cut off so I want to reiterate here:

1) All front fans blow into the case
2) All rear fans blow out of the case.

Which way should the side fan blow, and why?


--

Greatest Movie Line Ever
http://home.houston.rr.com/rkba/MovieLine.wmv

"What is history but the story of how politicians have
squandered the blood and treasure of the human race?"
--Thomas Sowell

On Fri, 30 Sep 2005 15:00:17 GMT, spam@uce.gov (Bob) wrote:


Depends on the case and reason for side-fan, which parts
need more cooling. I generally don't recommend adding fans
just for the heck of it unless the case was already known to
be insufficient otherwise.

If the side fan blows in, low on the side, it will help
video card most. Higher up, CPU, or in the middle both
parts some. It can be a necessary config if the case is
poor (cheap or old design meant for early ATX systems a
decade ago) and has little intake, and only a PSU exahust.
In this latter situation it can be useful to make sure the
side panel fan produces significant positive pressurization
of the case so the HDD rack airflow goes the opposite
direction, out the bottom front of the case. This latter
altenative is not idea to say the least but some people have
already built up a system and simply refuse to strip case
down to currect it or buy another case, so the remaining
alternative is a barely-effective slot fan or moddin' only
the side panel.

Any of the 3 (blowing in) options tend to reduce flow past
HDD rack.

Blowing out will help HDDs most, video card a little (in
general, again the specific case would need considered).

Blowing in it usually more useful but I would do it only if
the HDD rack has a pusher fan in front of it, instead of
only passive intake.

Bob wrote:
The way that provides the best cooling -- really. It seems that each
computer case is different enough that it requires testing the fans in
every possible way to find the combination that provides the best
cooling. How many fans do you have? And how many hours of testing do
you want to do?

On 30 Sep 2005 22:47:12 -0700, "larry moe 'n curly"
<larrymoencurly@my-deja.com> wrote:

I accept the direction that the case designers chose.