Motherboard and CPUs Temps Confusing as Reported

System: Asus P5W DH mb - Intel E6600 2 gb ram - Gigabyte 7600 GT fanless
video card, WD 150 gb raptor, Lian Li aluminum case.

I cannot figure out what is up w/my temperature probe. Last week my power
supply (an antec semi-fanless model) blew out, perhaps due to a power surge
during a reboot - (a bulb on my lamp was fried at the same time). I had the
PS replaced with the TriPower Antec 650W PS. I monitor the temperature of
my components with PC Probe II (latest version) and Everest Ultimate. I am
running Vista. There is an admitted problem w/the airflow in my Lian Li case
which I have tried to compensate for.

The problem is instead of a CPU fan, I have a one of those giant heatsinks
that is connected by a rubber duct to the 120 mm back case fan which pulls
the hot CPU air out of the case. This solution seemed very novel at the time
and keep the CPU running at low temperatures for the most part. But that
commandered case fan is no longer fulfilling its role in exhausting other
hot air from the case.

I also have one of those Gigabyte silent pipe GPUs and myLian Li case has
only one other exhaust fan which is far removed from the mb so the hot air
is not exhausted from the case. The result has been very high mb
temperatures. Until last week, I just left the case open and had a small 6"
table fan blowing air directly into the chassis. This kept temperatures at
a reasonable level. After the PC wouldn't turn on, I took it to a repair
tech who wired everything back together and put in the new PS. He connected
my faux CPU fan and the other case fan to the power leads on the new Antec
PS and told me everything would work fine.

With the case closed, the temps skyrocketed. I noticed the mouse cursor
bouncing around which I had seen in the past during overheating. When I
turned the temp monitors back on, the numbers were alarming. It was a very
hot day here as well. The MB temp was 75C, the CPU was 45, the cores were
higher, the hard drive was 48 C. As soon as I took the cover off the case
and turned my small desk fan on the mb, the temps dropped precipitiously.

I then tried putting the case back together and aiming the fan thru the
holes in the 5.25 drive bays. Again, the numbers were OK, very low in fact.
However, when I left the fan on at the higher speed, he motherboard numbers
went down to 1C. I figured this could not be good. On the other hand, when
I left the fan off, the mb numbers climbed within about 10 minutes to over
100C. This just didn't seem right to me and it even occured w/the case open
and the room cool. I started doubting the numbers.

The system is stablilized and I am looking next week to mount a special bay
fan enclosure that pulls air into the case, but I noticed today that the CPU
(not the cores) and MB temp stuck at 24 and 38C respectively. They did not
change for hours. When I rebooted (soft reset), the bios showed 24 and 38
even after I had turned off the small fan blowing air into the front of the
case.

I then shut down the PC completely. On reboot, the mb temp was 50C, what I
would have predicted w/the fan off.

Now I question the reliability of the all the sensors on my system. I think
they are generally correct. However, I would be interested in comments about
what might be happening here. As I write this message, the temp of the CPU
is shown as 66C while core 1 is 32 and core 2 is 31. Is this possible?

After I rebooted, the bios showed the CPU temp as 26C even after it had been
shown as 66C about 60 seconds before in PCProbe and Everest. Is it possible
for the temp to drop that quickly? I left the PC on in the bios for about
30 minutes. The CPU temp stayed at 29-30C the entire time. To me anyway,
this would indicate there is nothing wrong with the way the CPU is mounted,
ie. grease was not spread properly etc.

Again I am at a loss here and could use some insight.

Nothing posted says why the original failure occurred. Parts were
replaced only on speculation. Also not provided is any reason to
believe the new power supply is stable.

Based upon what was posted, every reply can only be speculation.
With so many changes, you must start from scratch to move everything
(component by component) from the 'unknown' category to either
'definitely good' or 'definitely bad'. That means using a 3.5 digit
multimeter to collect numbers from one of the power supplies orange,
purple, red, and yellow wires when the system is accessing all
peripherals simultaneously (multitasking).

Obviously if you can touch each suspected component (motherboard
temperature would be inside one of the motherboard chipsets) and not
leave skin, then temperatures are not problematic. So the question is
why those numbers are so erratic. That begins by confirming voltages
on those four power supply wires (when system is accessing everything
simultaneously) remain above 3.23, 4.87 and 11.7 volts. Voltage
variations can cause massive changes in the A/D circuits that monitor
both temperature and voltage.

You cannot use motherboard monitor for these measurements.
Motherboard voltage monitor must first be calibrated with that 3.5
digit multimeter. Once calibrated, then the alarm points for that
monitor can be set.

Without numbers from that multimeter, then nobody can say whether
the new power supply is working OK. A defective power supply can
still permit a computer to boot and function. Get and post those
numbers. Your replies will only be as good as information (ie
numbers) provided. Currently, every solution can only be wild
speculation.

On Jun 18, 2:56 am, "John Smith" <j...@smith.com> wrote:
> System: Asus P5W DH mb - Intel E6600 2 gb ram - Gigabyte 7600 GT fanless
> video card, WD 150 gb raptor, Lian Li aluminum case.
>
> I cannot figure out what is up w/my temperature probe. Last week my power
> supply (an antec semi-fanless model) blew out, perhaps due to a power surge
> during a reboot - (a bulb on my lamp was fried at the same time). I had the
> PS replaced with the TriPower Antec 650W PS. I monitor the temperature of
> my components with PC Probe II (latest version) and Everest Ultimate. I am
> running Vista. There is an admitted problem w/the airflow in my Lian Li case
> which I have tried to compensate for.
>
> The problem is instead of a CPU fan, I have a one of those giant heatsinks
> that is connected by a rubber duct to the 120 mm back case fan which pulls
> the hot CPU air out of the case. This solution seemed very novel at the time
> and keep the CPU running at low temperatures for the most part. But that
> commandered case fan is no longer fulfilling its role in exhausting other
> hot air from the case.
>
> I also have one of those Gigabyte silent pipe GPUs and myLian Li case has
> only one other exhaust fan which is far removed from the mb so the hot air
> is not exhausted from the case. The result has been very high mb
> temperatures. Until last week, I just left the case open and had a small 6"
> table fan blowing air directly into the chassis. This kept temperatures at
> a reasonable level. After the PC wouldn't turn on, I took it to a repair
> tech who wired everything back together and put in the new PS. He connected
> my faux CPU fan and the other case fan to the power leads on the new Antec
> PS and told me everything would work fine.
>
> With the case closed, the temps skyrocketed. I noticed the mouse cursor
> bouncing around which I had seen in the past during overheating. When I
> turned the temp monitors back on, the numbers were alarming. It was a very
> hot day here as well. The MB temp was 75C, the CPU was 45, the cores were
> higher, the hard drive was 48 C. As soon as I took the cover off the case
> and turned my small desk fan on the mb, the temps dropped precipitiously.
>
> I then tried putting the case back together and aiming the fan thru the
> holes in the 5.25 drive bays. Again, the numbers were OK, very low in fact.
> However, when I left the fan on at the higher speed, he motherboard numbers
> went down to 1C. I figured this could not be good. On the other hand, when
> I left the fan off, the mb numbers climbed within about 10 minutes to over
> 100C. This just didn't seem right to me and it even occured w/the case open
> and the room cool. I started doubting the numbers.
>
> The system is stablilized and I am looking next week to mount a special bay
> fan enclosure that pulls air into the case, but I noticed today that the CPU
> (not the cores) and MB temp stuck at 24 and 38C respectively. They did not
> change for hours. When I rebooted (soft reset), the bios showed 24 and 38
> even after I had turned off the small fan blowing air into the front of the
> case.
>
> I then shut down the PC completely. On reboot, the mb temp was 50C, what I
> would have predicted w/the fan off.
>
> Now I question the reliability of the all the sensors on my system. I think
> they are generally correct. However, I would be interested in comments about
> what might be happening here. As I write this message, the temp of the CPU
> is shown as 66C while core 1 is 32 and core 2 is 31. Is this possible?
>
> After I rebooted, the bios showed the CPU temp as 26C even after it had been
> shown as 66C about 60 seconds before in PCProbe and Everest. Is it possible
> for the temp to drop that quickly? I left the PC on in the bios for about
> 30 minutes. The CPU temp stayed at 29-30C the entire time. To me anyway,
> this would indicate there is nothing wrong with the way the CPU is mounted,
> ie. grease was not spread properly etc.
>
> Again I am at a loss here and could use some insight.

John Smith wrote:
> System: Asus P5W DH mb - Intel E6600 2 gb ram - Gigabyte 7600 GT fanless
> video card, WD 150 gb raptor, Lian Li aluminum case.
>
> I cannot figure out what is up w/my temperature probe. Last week my
> power supply (an antec semi-fanless model) blew out, perhaps due to a
> power surge during a reboot - (a bulb on my lamp was fried at the same
> time). I had the PS replaced with the TriPower Antec 650W PS. I monitor
> the temperature of my components with PC Probe II (latest version) and
> Everest Ultimate. I am running Vista. There is an admitted problem w/the
> airflow in my Lian Li case which I have tried to compensate for.
>
> The problem is instead of a CPU fan, I have a one of those giant
> heatsinks that is connected by a rubber duct to the 120 mm back case fan
> which pulls the hot CPU air out of the case. This solution seemed very
> novel at the time and keep the CPU running at low temperatures for the
> most part. But that commandered case fan is no longer fulfilling its
> role in exhausting other hot air from the case.
>
> I also have one of those Gigabyte silent pipe GPUs and myLian Li case
> has only one other exhaust fan which is far removed from the mb so the
> hot air is not exhausted from the case. The result has been very high mb
> temperatures. Until last week, I just left the case open and had a
> small 6" table fan blowing air directly into the chassis. This kept
> temperatures at a reasonable level. After the PC wouldn't turn on, I
> took it to a repair tech who wired everything back together and put in
> the new PS. He connected my faux CPU fan and the other case fan to the
> power leads on the new Antec PS and told me everything would work fine.
>
> With the case closed, the temps skyrocketed. I noticed the mouse cursor
> bouncing around which I had seen in the past during overheating. When I
> turned the temp monitors back on, the numbers were alarming. It was a
> very hot day here as well. The MB temp was 75C, the CPU was 45, the
> cores were higher, the hard drive was 48 C. As soon as I took the cover
> off the case and turned my small desk fan on the mb, the temps dropped
> precipitiously.
>
> I then tried putting the case back together and aiming the fan thru the
> holes in the 5.25 drive bays. Again, the numbers were OK, very low in
> fact. However, when I left the fan on at the higher speed, he
> motherboard numbers went down to 1C. I figured this could not be good.
> On the other hand, when I left the fan off, the mb numbers climbed
> within about 10 minutes to over 100C. This just didn't seem right to me
> and it even occured w/the case open and the room cool. I started
> doubting the numbers.
>
> The system is stablilized and I am looking next week to mount a special
> bay fan enclosure that pulls air into the case, but I noticed today that
> the CPU (not the cores) and MB temp stuck at 24 and 38C respectively.
> They did not change for hours. When I rebooted (soft reset), the bios
> showed 24 and 38 even after I had turned off the small fan blowing air
> into the front of the case.
>
> I then shut down the PC completely. On reboot, the mb temp was 50C, what
> I would have predicted w/the fan off.
>
> Now I question the reliability of the all the sensors on my system. I
> think they are generally correct. However, I would be interested in
> comments about what might be happening here. As I write this message,
> the temp of the CPU is shown as 66C while core 1 is 32 and core 2 is 31.
> Is this possible?
>
> After I rebooted, the bios showed the CPU temp as 26C even after it had
> been shown as 66C about 60 seconds before in PCProbe and Everest. Is it
> possible for the temp to drop that quickly? I left the PC on in the
> bios for about 30 minutes. The CPU temp stayed at 29-30C the entire
> time. To me anyway, this would indicate there is nothing wrong with the
> way the CPU is mounted, ie. grease was not spread properly etc.
>
> Again I am at a loss here and could use some insight.
>
>
>
>
>
>
>
>
What an interesting problem! Let me comment on the two areas of the
problem, the sensing of the temperature and the airflow in the box.

1. You are apparently using a sensing device (in addition to what the
system provides to the BIOS) you don't trust. A simple idea to provide
a sanity check would be to get one of those indoor/outdoor thermostats
where the wire goes out the window and there is a plastic-covered sensor
you can stick on the outside of your house. Only, put it somewhere
strategic inside your case, away from fans, of course. And be sure it
has no metallic parts showing. Don't use an oven thermometer: its
sensor is a big metal thing that you don't want floating around inside
your box.

The indoor/outdoor thermometer idea would tell you what device is
telling you the truth.

2. My idea for a properly-cooled case is to follow the AMD spec for
power supplies. They want the fan at the bottom of the power supply,
not at the back. That draws the heat from the CPU and the MB uniformly.
There is usually nothing that needs the fan to be in the back of the
power supply unit, so why do they put them there?

I had a heat buildup problem some years ago which I solved by adding a
fan in the place provided at the lower front of the case, blowing in. I
still had the fan in the back of the case in addition to the one in the
power supply unit. The case had a place for a fan there. It also houses
the speaker for the system. I went to Hobby Lobby and bought some
plastic sheets and industrial strength glue and made a house for a
couple of big SCSI drives that ran very hot. I got two pieces of
aluminum to hold the drives and to anchor them to this box. I ran a
wire from the aluminum to the case for grounding. The back of the box
opened to the bottom of the case, cooling the lower components. The
SCSI drives ran so hot you ordinarily couldn't touch them, but with this
box and the fan flowing on them, they were only slightly warm.

I still had a heat problem. The back case fan died and I had nothing in
place to tell me I had a problem. That may be what you have the PC
Probe II for. I don't know what that is. But I think we do need to
have something in the case to make an audible/visual alarm when the
numbers get out of range. Sometimes the BIOS/MB would do this. My last
MB was a Gigabyte, so forget it. My last Asus had something that would
put the alarm on the desktop.

I now use an OCX power supply unit which does follow the AMD guideline
and has a 120mm fan on the under side and that runs very much cooler
than my previous 80mm fan. The OCX does run to the edge of the usually
5% acceptable voltage range, so I'm not too happy with it. But it has a
good fan.

John

On Jun 18, 9:40 pm, John <J...@TheStewartWeb.com> wrote:
> 2. My idea for a properly-cooled case is to follow the AMD spec for
> power supplies. They want the fan at the bottom of the power supply,
> not at the back. That draws the heat from the CPU and the MB uniformly.
> There is usually nothing that needs the fan to be in the back of the
> power supply unit, so why do they put them there?

Location of a fan inside a power supply will have zero consequences
for chassis cooling. Aiflfow is a most fundamental concept in
cooling. Only significant parameter is the CFM number - how much air
is moved. Same air is moved with fan on bottom of supply or at rear.
Fans have unique CFM and dB numbers.

Some supply manufacturers may locate that fan for reasons internal -
so as to eliminate 'dead spots' or reduce noise. That is unique to
'their' design. For chassis cooling, the only significant number is
CFM. CFM is same whether can is on bottom of supply or in rear.

And obviously, none of these cooling concepts have any relationship
to John Smith's original problem.

Mystery solved:

The bizarre temp readings, (even after blowing air thru the front bays),
were the result of the Asus utility, PC Probe II. After uninstalling PC
Probe II and Everest, and then reinstalling Everest, all of my temp
regisiters are working normally. Very clearly PC Probe II threw a monkey
wrench into the monitoring.

The first person who replied to my post here was well-intentioned, but
proposed a strategy more appropriate to investigation of a plane a crash.
When a PC fails to turn on and the power supply is determined to be
defective, common sense tells you it's extraordinarily unlikely that the new
PS is defective as that poster asserted it might be. I complicated the
situation by introducing a new element into the equation, PC Probe II, which
I installed after the new PS was put into my PC. The bizarre readings it
produced led to the erroneous assumption that the PS might not be the entire
problem or other components were damaged by the alleged power surger.

I posted a problem description, not as extensive as the one here, on the
Asus tech support forum for my mb, at Anantech, again for my mb, and at
Everest's site. Someone just happened to have had this same problem w/PC
Probe II - CPU temp exceeeding core temps by 40C and bouncing around wildly.

For those who offered cooling solutions, in retrospect, my selection of the
Thermalright HR-01-775 fan duct
http://www.thermalright.com/product_default.htm was not a smart one because
it doesn't exhaust any heat off the mb, only the CPU. It starves the mb for
air while serving as a ventilator for the CPU only - that is where I have no
other fan cutouts for exhaust. My solution will be in install the SYS
COOLER525 SCYTHE|SCKB-1000SL RT which is a 120mm low noise fan that can
blow air in thru open front bays. It's an interesting product and was highly
rated on Newegg.

I would not recommend the particular Lian Li case I purchased to anyone.
Cases that have more exhaust fan openings like the Antec P180 are better
designed. I would also not recommend the Antec Phantom silent PS. It failed
and the Antec PS I replaced it w/which has a fan makes no more noise.

I really pushed the envelope of the silent PC paradigm: fanless PS, fanless
video card, "fanless" CPU cooler (appropriating back fan for exhaust) and
paid a steep price in terms of time consumed dealing w/this problem. I also
reinjured my back carting around a broken PC for repair. There are clearly
better solutions are out there.

Thanks for all of your responses.



"w_tom" <> wrote in message
news: ups.com...
> On Jun 18, 9:40 pm, John <J...@TheStewartWeb.com> wrote:
>> 2. My idea for a properly-cooled case is to follow the AMD spec for
>> power supplies. They want the fan at the bottom of the power supply,
>> not at the back. That draws the heat from the CPU and the MB uniformly.
>> There is usually nothing that needs the fan to be in the back of the
>> power supply unit, so why do they put them there?
>
> Location of a fan inside a power supply will have zero consequences
> for chassis cooling. Aiflfow is a most fundamental concept in
> cooling. Only significant parameter is the CFM number - how much air
> is moved. Same air is moved with fan on bottom of supply or at rear.
> Fans have unique CFM and dB numbers.
>
> Some supply manufacturers may locate that fan for reasons internal -
> so as to eliminate 'dead spots' or reduce noise. That is unique to
> 'their' design. For chassis cooling, the only significant number is
> CFM. CFM is same whether can is on bottom of supply or in rear.
>
> And obviously, none of these cooling concepts have any relationship
> to John Smith's original problem.
>

doh, I forgot to repost, I sent you an email by accident and not to group
your email address was wrong in the header tho (doh) .. I ment to re-post to
group but I never did lol

----- Original Message -----
From: xxx email removed
To: "John Smith" email removed
Sent: Monday, June 18, 2007 9:49 AM
Subject: Re: Motherboard and CPUs Temps Confusing as Reported


> > I monitor the temperature of
> > my components with PC Probe II (latest version) and Everest Ultimate. I
am
> > running Vista. There is an admitted problem w/the airflow in my Lian Li
> case
> > which I have tried to compensate for.
> >
>
> Running more than 1 program that will monitor the same sensor chip can
cause
> errors on all sensors I have ever seen. Try with just pc probe?
> If the bios has a automatic speedfan setting then that could also be
pulling
> the temp and helping to cause erronious error..
>
> > However, when I left the fan on at the higher speed, he motherboard
> numbers
> > went down to 1C. I figured this could not be good. On the other hand,
> when
> > I left the fan off, the mb numbers climbed within about 10 minutes to
over
> > 100C. This just didn't seem right to me and it even occured w/the case
> open
> > and the room cool. I started doubting the numbers.
>
> 1c and 100c could be a scaling issue with the software or just a faulty
> sensor.. a bit like rpm sensors for fans.. If the motherboard temp was
> 100c then the board would be on the edge of death .. so it's a good sign
> that there is a problem (the good part being that ur board didnt actually
> get to 100c)
>
> > Now I question the reliability of the all the sensors on my system. I
> think
> > they are generally correct. However, I would be interested in comments
> about
> > what might be happening here. As I write this message, the temp of the
CPU
> > is shown as 66C while core 1 is 32 and core 2 is 31. Is this possible?
> >
> > After I rebooted, the bios showed the CPU temp as 26C even after it had
> been
> > shown as 66C about 60 seconds before in PCProbe and Everest. Is it
> possible
> > for the temp to drop that quickly? I left the PC on in the bios for
about
> > 30 minutes. The CPU temp stayed at 29-30C the entire time. To me anyway,
> > this would indicate there is nothing wrong with the way the CPU is
> mounted,
> > ie. grease was not spread properly etc.
>
> I would make sure that speedfan or whatever in the bios is off, COMPLETLY
> disable pc probe II .. and then do a cold boot and see what everest
> reads..
>
> and yes, the cpu could possibly cool from 66 to 30 in a matter of seconds
> (depending on the heatsink).. I like the vent options but I wouldnt have
> thought it a requirement for your cpu :/
>