[lm-sensors] About the "CPU temp" output of ADM1025A chip

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi All,

I have an Intel D850MV motherboard. There is an ADM1025A sensor chip on the
motherboard. Also there is an LPC47M142 I/O chip on the motherboard. My CPU
is a Pentium 4 CPU, 1.7GHz.

The following is a sample output for my lm-sensors software.
---------------------------------
$ sensors
adm1025-i2c-0-2d
Adapter: SMBus I801 adapter at efa0

+2.5V:     +2.49 V  (min =  +2.25 V, max =  +2.75 V)
VCore:     +1.68 V  (min =  +1.66 V, max =  +1.84 V)
+3.3V:     +3.30 V  (min =  +2.97 V, max =  +3.63 V)
+5V:       +5.13 V  (min =  +4.50 V, max =  +5.50 V)
VCC:       +3.30 V  (min =  +2.97 V, max =  +3.63 V)
CPU Temp:  +35.0?C  (low  =   +10?C, high =   +60?C)
M/B Temp:  +34.0?C  (low  =   +10?C, high =   +45?C)
vid:      +1.750 V  (VRM Version 9.0)

smsc47m1-isa-0800
Adapter: ISA adapter
fan1:     3233 RPM  (min =  640 RPM, div = 8)
fan2:        0 RPM  (min =  640 RPM, div = 8)
--------------------------------------------------------------

I am interested in the temperature readings from the sensor chip. There are
two temperature readings: CPU temp and M/B temp. According the ADM1025A
datasheet, I think the M/B temp is the on-chip sensor in the ADM1025A chip,
and the CPU temp is the External Temperature Measurement with Remote Diode.

My question is: Where is this remote diode located? Is this remote diode the
internal thermal diode inside the Pentium 4 CPU chip, or a thermal sensor
provided by the motherboard which is attached to the CPU chip very closely?

I downloaded the manual of Intel D850MV motherboard from Intel's website,
but could not get the answer to my question.

I opened the computer case, removed the CPU fan, and saw the CPU and the
ADM1025A chip sit together on the motherboard, but I could not see from
the circuit board whether the PIN 9 and 10 (the two pins connected to
the remote diode) of ADM1025A chip are connected to the internal thermal
diode inside the Pentium 4 CPU chip.

Any suggestions are very welcome.

Thank you!

Regards,
Yongkui
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061021/ddfd7283/attachment.html 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Rudolf Marek]

Hi,

> I have an Intel D850MV motherboard. There is an ADM1025A sensor chip on
> the motherboard. Also there is an LPC47M142 I/O chip on the motherboard.
> My CPU is a Pentium 4 CPU, 1.7GHz.

And here is your motherboard
ftp://download.intel.com/design/motherbd/mv/A6514501.pdf

technical guide ;)


> The following is a sample output for my lm-sensors software.
> ---------------------------------
> $ sensors
> adm1025-i2c-0-2d
> Adapter: SMBus I801 adapter at efa0
>  
> +2.5V:     +2.49 V  (min =  +2.25 V, max =  +2.75 V)
> VCore:     +1.68 V  (min =  +1.66 V, max =  +1.84 V)
> +3.3V:     +3.30 V  (min =  +2.97 V, max =  +3.63 V)
> +5V:       +5.13 V  (min =  +4.50 V, max =  +5.50 V)
> VCC:       +3.30 V  (min =  +2.97 V, max =  +3.63 V)
> CPU Temp:  +35.0?C  (low  =   +10?C, high =   +60?C)
> M/B Temp:  +34.0?C  (low  =   +10?C, high =   +45?C)
> vid:      +1.750 V  (VRM Version 9.0)
>  
> smsc47m1-isa-0800
> Adapter: ISA adapter
> fan1:     3233 RPM  (min =  640 RPM, div = 8)
> fan2:        0 RPM  (min =  640 RPM, div = 8)
> --------------------------------------------------------------
>  
> I am interested in the temperature readings from the sensor chip. There
> are two temperature readings: CPU temp and M/B temp. According the
> ADM1025A datasheet, I think the M/B temp is the on-chip sensor in the
> ADM1025A chip, and the CPU temp is the External Temperature Measurement
> with Remote Diode.
>  
> My question is: Where is this remote diode located? Is this remote diode
> the internal thermal diode inside the Pentium 4 CPU chip, or a
> thermal sensor provided by the motherboard which is attached to the CPU
> chip very closely?

According to technical docs from intel, it is Remote thermal diode sensing for
direct monitoring of processor temperature

inside the CPU.

> I downloaded the manual of Intel D850MV motherboard from Intel's
> website, but could not get the answer to my question.

I did ;)

> I opened the computer case, removed the CPU fan, and saw the CPU and the
> ADM1025A chip sit together on the motherboard, but I could not see from
> the circuit board whether the PIN 9 and 10 (the two pins connected to
> the remote diode) of ADM1025A chip are connected to the internal thermal
> diode inside the Pentium 4 CPU chip.

Well there is a lot of layers PCB so it is impossible to track it down.

The manual also suggest that the fan speeds are somehow controlled. This is
quite interesting. Page 43 Quote:

The temperature sensors that control the fans are initialized by the BIOS at
power-up to turn on only when the sensor reaches 36 C (96.8 F). The fans switch
off when their respective sensors
cool down to 31 C (87.8 F). This prevents the fans from turning off and on when
the ambient air around the sensor fluctuates around 35?36 C. When the fans
switch on, they may appear to rotate
slowly because the fan's duty cycle starts at 70 percent and rises to 100
percent when the sensor reaches 46 C (114.8 F).

Datasheet is here http://www.datasheets.org.uk/datasheet.php?article 40255

It seems it somehow changes the PWM via software (system management interrupt)

I hope it helps,

Regards
Rudolf

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Rudolf Marek]

Hi Yongkui,

> Thanks very much for your help.

Well always good to hear it. Unfortunately I have not so much free time recently 
for this kind of lm-sensors support stuff.

> The pin numbers for the thermal diode in the Pentium 4 CPU are H38 and 
> E39. I think the PIN 9 and 10 of ADM1025A chip are connected to the PIN 
> H38 and E39 of the Pentium 4 CPU. So the temperature output "CPU Temp" 
> from lm-sensors should be the temprature reading of this thermal diode.

Good.

> The datasheet of my Pentium 4 CPU is here:
> http://www.intel.com/design/pentium4/datashts/249198.htm
> ftp://download.intel.com/design/Pentium4/datashts/24919805.pdf
>  
>  
> A picture for my CPU chip and the ADM1025A chip is attached. You can see 
> in the picture that the ADM1025A chip sits close to the Pentium 4 CPU 
> chip (the top-right location).

It should be so - the analog lines are as shortest as possible.

> The followings are some observations from my experiments with lm-sensors:
> -----------------------------------
> The operating system is idle: CPU5C, M/B4C.
> I run many CPU-intensive programs to see how high the CPU temperature 
> can reach.
> The highest temperature I observed are: CPUUC, M/BGC
> The largest temperature difference between CPU and M/B is 8C.
> It takes about 8 minutes for the CPU temp to reach 55C, and then stay at 
> 55C. This means it reached its steady state temperature.

Hmm this indicates perhaps the thermistor instead of the diode. It takes too 
long. The core temperature should grow very very fast. But if the pins are 
connected to the socket it must be diode. (and it seems that the chip is for 
diode only)

> I expect the CPU temperature could reach 80C or 90C for some 
> CPU-intensive programs. But unfortunatly, 55C is the highest temperature 
> I observed so far. 55C is not so high in my opinion, why the CPU 
> temperature cannot reach 80C or 90C? The critical temperature for the 
> Pentium 4 CPU chip to shut down is 135C. 

Yes true.

> Now I am wondering where this thermal diode is located inside the 
> Pentium 4 CPU. Is this thermal diode located near the Integer Register 
> unit, or near the Floating-point Multiplication unit,  or near the Level 
> 2 Cache unit? You know, there is a large temperature difference between 
> two different locations on the CPU chip.

Well this is not known. If some parts get too hot the CPU  might throttle a bit 
to cool down. Maybe you could try the "cpuburn" software.
http://packages.debian.org/unstable/misc/cpuburn.html
(for debian)

> Do you have any hints about the thermal diode location in the Pentium 4 
> CPU chip?

No but you may take a look to: Intel? 64 and IA-32 Architectures Software 
Developer's Manual, Volume 3A: System Programming Guide describes the 
operating-system support environment of an IA-32 and Intel? 64 architectures, 
including: memory management, protection, task management, interrupt and 
exception handling, multi-processor support, and thermal and power management 
features.

There are some notes and stuff about the thermal monitor (the digital sensor 
have core/core2 cpus so perhaps not yours)

http://www.intel.com/design/pentium4/manuals/index_new.htm

All in all there is more then one temperature sensor, sometimes there is digital 
temperature sensor, and even the diode is locate on other place...


You may check the "IA32_THERM_STATUS MSR" via MSR reads (/dev/msr/...)
http://www.google.com/codesearch?q=+rdmsr.c+show:EynqVEYBclg:kdySSoK4L6Y:chKAG6mOivc&sa=N&cd=1&ct=rc&cs_p=http://www.stillhq.com/extracted/x86info-wx.tgz&cs_f=x86info-wx/x86info-1.11/rdmsr.c#a0

To sum it up. You need some cpuburn to burn the cpu. Then the temp should grow 
quickly. Using the MSR you may check if the CPU defend itself when too hot.

I hope it will help you during your experiments. If you burn the CPU to dead ;), 
buy Core 2 CPU it has the embedded temperature sensor (and I wrote the driver 
recently)

Regards
Rudolf

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

Thanks so much for your reply.

I downloaded the CPU burn program, and run it to see how high the CPU temp
can reach.

I run the CPU burn program for 10 minutes = 600 seconds. The temperature
curve for the CPU burn program is attached for your reference. The highest
temperature reached is 53C. I also drew the temperature curve for the 900
seconds before running cpuburn, and for the 300 seconds after cpuburn
program exited.

In the attachments, the file cpuburn.pdf is the complete temperature curve
for nearly 1800 seconds. The file cpuburn.100s.pdf is for the 100
seconds time interval of [850, 950] only.

The CPU temp is increased 4C (from 39C to 43C) during the first 8 seconds
when cpuburn program starts running. But it takes about 300 seconds for the
CPU temp to reach 53C. Does the CPU temp grow fast or slowly?

Based on the temperature curve, which one do you think the CPU temp is for?
The thermal diode inside the CPU chip, or just a thermistor?

Thanks!

Regards,
Yongkui


On 10/22/06, Rudolf Marek <r.marek at assembler.cz> wrote:
>
> Hi Yongkui,
>
> > Thanks very much for your help.
>
> Well always good to hear it. Unfortunately I have not so much free time
> recently
> for this kind of lm-sensors support stuff.
>
> > The pin numbers for the thermal diode in the Pentium 4 CPU are H38 and
> > E39. I think the PIN 9 and 10 of ADM1025A chip are connected to the PIN
> > H38 and E39 of the Pentium 4 CPU. So the temperature output "CPU Temp"
> > from lm-sensors should be the temprature reading of this thermal diode.
>
> Good.
>
> > The datasheet of my Pentium 4 CPU is here:
> > http://www.intel.com/design/pentium4/datashts/249198.htm
> > ftp://download.intel.com/design/Pentium4/datashts/24919805.pdf
> >
> >
> > A picture for my CPU chip and the ADM1025A chip is attached. You can see
> > in the picture that the ADM1025A chip sits close to the Pentium 4 CPU
> > chip (the top-right location).
>
> It should be so - the analog lines are as shortest as possible.
>
> > The followings are some observations from my experiments with
> lm-sensors:
> > -----------------------------------
> > The operating system is idle: CPU5C, M/B4C.
> > I run many CPU-intensive programs to see how high the CPU temperature
> > can reach.
> > The highest temperature I observed are: CPUUC, M/BGC
> > The largest temperature difference between CPU and M/B is 8C.
> > It takes about 8 minutes for the CPU temp to reach 55C, and then stay at
> > 55C. This means it reached its steady state temperature.
>
> Hmm this indicates perhaps the thermistor instead of the diode. It takes
> too
> long. The core temperature should grow very very fast. But if the pins are
> connected to the socket it must be diode. (and it seems that the chip is
> for
> diode only)
>
> > I expect the CPU temperature could reach 80C or 90C for some
> > CPU-intensive programs. But unfortunatly, 55C is the highest temperature
> > I observed so far. 55C is not so high in my opinion, why the CPU
> > temperature cannot reach 80C or 90C? The critical temperature for the
> > Pentium 4 CPU chip to shut down is 135C.
>
> Yes true.
>
> > Now I am wondering where this thermal diode is located inside the
> > Pentium 4 CPU. Is this thermal diode located near the Integer Register
> > unit, or near the Floating-point Multiplication unit,  or near the Level
> > 2 Cache unit? You know, there is a large temperature difference between
> > two different locations on the CPU chip.
>
> Well this is not known. If some parts get too hot the CPU  might throttle
> a bit
> to cool down. Maybe you could try the "cpuburn" software.
> http://packages.debian.org/unstable/misc/cpuburn.html
> (for debian)
>
> > Do you have any hints about the thermal diode location in the Pentium 4
> > CPU chip?
>
> No but you may take a look to: Intel(r) 64 and IA-32 Architectures Software
> Developer's Manual, Volume 3A: System Programming Guide describes the
> operating-system support environment of an IA-32 and Intel(r) 64
> architectures,
> including: memory management, protection, task management, interrupt and
> exception handling, multi-processor support, and thermal and power
> management
> features.
>
> There are some notes and stuff about the thermal monitor (the digital
> sensor
> have core/core2 cpus so perhaps not yours)
>
> http://www.intel.com/design/pentium4/manuals/index_new.htm
>
> All in all there is more then one temperature sensor, sometimes there is
> digital
> temperature sensor, and even the diode is locate on other place...
>
>
> You may check the "IA32_THERM_STATUS MSR" via MSR reads (/dev/msr/...)
>
> http://www.google.com/codesearch?q=+rdmsr.c+show:EynqVEYBclg:kdySSoK4L6Y:chKAG6mOivc&sa=N&cd=1&ct=rc&cs_p=http://www.stillhq.com/extracted/x86info-wx.tgz&cs_f=x86info-wx/x86info-1.11/rdmsr.c#a0
>
> To sum it up. You need some cpuburn to burn the cpu. Then the temp should
> grow
> quickly. Using the MSR you may check if the CPU defend itself when too
> hot.
>
> I hope it will help you during your experiments. If you burn the CPU to
> dead ;),
> buy Core 2 CPU it has the embedded temperature sensor (and I wrote the
> driver
> recently)
>
> Regards
> Rudolf
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061022/eca992a8/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cpuburn.pdf
Type: application/pdf
Size: 3367 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061022/eca992a8/attachment-0002.pdf 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cpuburn.100s.pdf
Type: application/pdf
Size: 1728 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061022/eca992a8/attachment-0003.pdf 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

I calculated the time constant based on the temperature curve of "CPU temp".
It is about 20 seconds.

In my opinion, the time constant of the silicon chip is about 10
milliseconds.

Does this suggest that the "CPU temp" is not the temperature of the thermal
diode inside the Pentium 4 CPU?


Oh, I just thought of one possible reason for this. Perhaps the current or
the voltage drop of thermal diode is not changing so fast as the temperature
change of the silicon area around it. So the temperature readings from the
diode will lag. If this is true, this could explain why the readings from
the thermal diode is not changing as rapidly as the temperature of the
silicon area around it.

But this explanation contradicts with the purpose of the thermal diode. If
the readings from the thermal diode cannot catch up with the rapid
temperature change, the thermal diode is not so useful.

I am confused here. I still think the "CPU temp" should be the readings from
the thermal diode inside the Pentium 4 CPU chip. But the fact that this "CPU
temp" is changing slowly in seconds and it could not reach a high
temperature makes me doubt it.

Any suggestions are very appreciated!

Thanks,
Yongkui


On 10/22/06, Yongkui Han <hanyongkui99 at gmail.com> wrote:
>
> Hi Rudolf,
>
> Thanks so much for your reply.
>
> I downloaded the CPU burn program, and run it to see how high the CPU temp
> can reach.
>
> I run the CPU burn program for 10 minutes = 600 seconds. The temperature
> curve for the CPU burn program is attached for your reference. The highest
> temperature reached is 53C. I also drew the temperature curve for the 900
> seconds before running cpuburn, and for the 300 seconds after cpuburn
> program exited.
>
> In the attachments, the file cpuburn.pdf is the complete temperature curve
> for nearly 1800 seconds. The file cpuburn.100s.pdf is for the 100
> seconds time interval of [850, 950] only.
>
> The CPU temp is increased 4C (from 39C to 43C) during the first 8 seconds
> when cpuburn program starts running. But it takes about 300 seconds for the
> CPU temp to reach 53C. Does the CPU temp grow fast or slowly?
>
> Based on the temperature curve, which one do you think the CPU temp is
> for? The thermal diode inside the CPU chip, or just a thermistor?
>
> Thanks!
>
> Regards,
> Yongkui
>
>
> On 10/22/06, Rudolf Marek <r.marek at assembler.cz> wrote:
> >
> > Hi Yongkui,
> >
> > > Thanks very much for your help.
> >
> > Well always good to hear it. Unfortunately I have not so much free time
> > recently
> > for this kind of lm-sensors support stuff.
> >
> > > The pin numbers for the thermal diode in the Pentium 4 CPU are H38 and
> > > E39. I think the PIN 9 and 10 of ADM1025A chip are connected to the
> > PIN
> > > H38 and E39 of the Pentium 4 CPU. So the temperature output "CPU Temp"
> >
> > > from lm-sensors should be the temprature reading of this thermal
> > diode.
> >
> > Good.
> >
> > > The datasheet of my Pentium 4 CPU is here:
> > > http://www.intel.com/design/pentium4/datashts/249198.htm
> > > ftp://download.intel.com/design/Pentium4/datashts/24919805.pdf
> > >
> > >
> > > A picture for my CPU chip and the ADM1025A chip is attached. You can
> > see
> > > in the picture that the ADM1025A chip sits close to the Pentium 4 CPU
> > > chip (the top-right location).
> >
> > It should be so - the analog lines are as shortest as possible.
> >
> > > The followings are some observations from my experiments with
> > lm-sensors:
> > > -----------------------------------
> > > The operating system is idle: CPU5C, M/B4C.
> > > I run many CPU-intensive programs to see how high the CPU temperature
> > > can reach.
> > > The highest temperature I observed are: CPUUC, M/BGC
> > > The largest temperature difference between CPU and M/B is 8C.
> > > It takes about 8 minutes for the CPU temp to reach 55C, and then stay
> > at
> > > 55C. This means it reached its steady state temperature.
> >
> > Hmm this indicates perhaps the thermistor instead of the diode. It takes
> > too
> > long. The core temperature should grow very very fast. But if the pins
> > are
> > connected to the socket it must be diode. (and it seems that the chip is
> > for
> > diode only)
> >
> > > I expect the CPU temperature could reach 80C or 90C for some
> > > CPU-intensive programs. But unfortunatly, 55C is the highest
> > temperature
> > > I observed so far. 55C is not so high in my opinion, why the CPU
> > > temperature cannot reach 80C or 90C? The critical temperature for the
> > > Pentium 4 CPU chip to shut down is 135C.
> >
> > Yes true.
> >
> > > Now I am wondering where this thermal diode is located inside the
> > > Pentium 4 CPU. Is this thermal diode located near the Integer Register
> > > unit, or near the Floating-point Multiplication unit,  or near the
> > Level
> > > 2 Cache unit? You know, there is a large temperature difference
> > between
> > > two different locations on the CPU chip.
> >
> > Well this is not known. If some parts get too hot the CPU  might
> > throttle a bit
> > to cool down. Maybe you could try the "cpuburn" software.
> > http://packages.debian.org/unstable/misc/cpuburn.html
> > (for debian)
> >
> > > Do you have any hints about the thermal diode location in the Pentium
> > 4
> > > CPU chip?
> >
> > No but you may take a look to: Intel(r) 64 and IA-32 Architectures
> > Software
> > Developer's Manual, Volume 3A: System Programming Guide describes the
> > operating-system support environment of an IA-32 and Intel(r) 64
> > architectures,
> > including: memory management, protection, task management, interrupt and
> > exception handling, multi-processor support, and thermal and power
> > management
> > features.
> >
> > There are some notes and stuff about the thermal monitor (the digital
> > sensor
> > have core/core2 cpus so perhaps not yours)
> >
> > http://www.intel.com/design/pentium4/manuals/index_new.htm
> >
> > All in all there is more then one temperature sensor, sometimes there is
> > digital
> > temperature sensor, and even the diode is locate on other place...
> >
> >
> > You may check the "IA32_THERM_STATUS MSR" via MSR reads (/dev/msr/...)
> >
> > http://www.google.com/codesearch?q=+rdmsr.c+show:EynqVEYBclg:kdySSoK4L6Y:chKAG6mOivc&sa=N&cd=1&ct=rc&cs_p=http://www.stillhq.com/extracted/x86info-wx.tgz&cs_f=x86info-wx/x86info-1.11/rdmsr.c#a0
> >
> > To sum it up. You need some cpuburn to burn the cpu. Then the temp
> > should grow
> > quickly. Using the MSR you may check if the CPU defend itself when too
> > hot.
> >
> > I hope it will help you during your experiments. If you burn the CPU to
> > dead ;),
> > buy Core 2 CPU it has the embedded temperature sensor (and I wrote the
> > driver
> > recently)
> >
> > Regards
> > Rudolf
> >
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061023/a889892e/attachment-0001.html 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Rudolf Marek]

Hi  Yongkui

> I calculated the time constant based on the temperature curve of "CPU 
> temp". It is about 20 seconds.

Ok. Important question: Is the fan switched off - disconnected? I think we see 
the "slow" start because the system cools down. The board has some kind of 
automatic cooling. Of course without the fan it might be bit dangerous if system 
temperature grows, but I think this is necessary to test it ;) Or if you are 
curious and want to test it ;)

>  
> In my opinion, the time constant of the silicon chip is about 10 
> milliseconds.

Well what constant?

> Does this suggest that the "CPU temp" is not the temperature of the 
> thermal diode inside the Pentium 4 CPU?

Also from the previous mail with graphs, I think it is but system is cooled

> Oh, I just thought of one possible reason for this. Perhaps the current 
> or the voltage drop of thermal diode is not changing so fast as the 
> temperature change of the silicon area around it. So the temperature 
> readings from the diode will lag. If this is true, this could explain 
> why the readings from the thermal diode is not changing as rapidly as 
> the temperature of the silicon area around it.

Yes but not in range of hundreds of seconds.


> I am confused here. I still think the "CPU temp" should be the readings 
> from the thermal diode inside the Pentium 4 CPU chip. But the fact that 
> this "CPU temp" is changing slowly in seconds and it could not reach a 
> high temperature makes me doubt it.

I think because of that hidden autoregulation. I would suggest to disconnect the 
fan connector (of course not the heatspreader ;) and try it without that.

Sorry for delay,

Regards
Rudolf

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

Thank you for your reply.

On 10/23/06, Rudolf Marek <r.marek at assembler.cz> wrote:
>
> Hi  Yongkui
>
> > I calculated the time constant based on the temperature curve of "CPU
> > temp". It is about 20 seconds.
>
> Ok. Important question: Is the fan switched off - disconnected? I think we
> see
> the "slow" start because the system cools down. The board has some kind of
> automatic cooling. Of course without the fan it might be bit dangerous if
> system
> temperature grows, but I think this is necessary to test it ;) Or if you
> are
> curious and want to test it ;)


Yes, the fan is connected and running all the time.  I plan to test it
without the fan this weekend. I will let you know the new high temperature
without the fan as soon as possible.


> >
> > In my opinion, the time constant of the silicon chip is about 10
> > milliseconds.
>
> Well what constant?


The thickness of the silicon chip is about 0.5mm, the thermal resistance is
about R = 0.4 K/W, the thermal capacitance is about C=0.025 J/K, so the time
constant = R*C = 0.01 second.
The time constant of the heat sink is about several minutes.
So it is my expectation that although the temperature of the silicon chip
changes rapidly, the temperature of the heat sink changes very slowly (in
minutes). So if the thermal diode is not inside the silicon layer but near
the heat sink or heat spreader, then the temperature change will be as slow
as the temperature curve of my experiments (or even slower).

> Does this suggest that the "CPU temp" is not the temperature of the
> > thermal diode inside the Pentium 4 CPU?
>
> Also from the previous mail with graphs, I think it is but system is
> cooled
>
> > Oh, I just thought of one possible reason for this. Perhaps the current
> > or the voltage drop of thermal diode is not changing so fast as the
> > temperature change of the silicon area around it. So the temperature
> > readings from the diode will lag. If this is true, this could explain
> > why the readings from the thermal diode is not changing as rapidly as
> > the temperature of the silicon area around it.
>
> Yes but not in range of hundreds of seconds.
>
>
> > I am confused here. I still think the "CPU temp" should be the readings
> > from the thermal diode inside the Pentium 4 CPU chip. But the fact that
> > this "CPU temp" is changing slowly in seconds and it could not reach a
> > high temperature makes me doubt it.
>
> I think because of that hidden autoregulation. I would suggest to
> disconnect the
> fan connector (of course not the heatspreader ;) and try it without that.


Will the hidden autoregulation (cpu fan speed change) prevent the
temperature of the silicon chip from increasing rapidly (from tens of
milliseconds to tens of seconds)? I have doubt on that.
Anyway, I plan to try it without fan working to see what will happen.

Sorry for delay,
>
> Regards
> Rudolf
>

Thank you very much for your ideas on this!

Regards,
Yongkui
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061024/b199ac84/attachment.html 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

I drew the curve of the fan speed together with the "CPU Temp" and "M/B
Temp" outputs from lm-sensors. The pdf file is attached for your reference.

The fan speed is very stable before the cpuburn program starts running, the
speed is switching between 3150 RPM and 3323 RPM.

After the cpubrun program starts running, the fan speed goes up: 3323 RPM
---> 3413 RPM ---> 3510 RPM ---> 3614 RPM ---> 3723 RPM. And after the
cpuburn program stops running, the fan speed goes down: 3614 RPM ---> 3510
RPM --->  3413 RPM.

Although the fan speed goes up when the CPU temperature is increasing, the
difference of the fan speed is not so much: 3723 RPM vs. 3150 RPM. The
difference is only 573 RPM.  Will this difference in the fan speed prevent
the temperature of the silicon chip from increasing rapidly (from tens of
milliseconds to tens of seconds)?

Do you have any insights about this?

Thanks!
Yongkui


On 10/24/06, Yongkui Han <hanyongkui99 at gmail.com> wrote:
>
> Hi Rudolf,
>
> Thank you for your reply.
>
> On 10/23/06, Rudolf Marek <r.marek at assembler.cz > wrote:
> >
> > Hi  Yongkui
> >
> > > I calculated the time constant based on the temperature curve of "CPU
> > > temp". It is about 20 seconds.
> >
> > Ok. Important question: Is the fan switched off - disconnected? I think
> > we see
> > the "slow" start because the system cools down. The board has some kind
> > of
> > automatic cooling. Of course without the fan it might be bit dangerous
> > if system
> > temperature grows, but I think this is necessary to test it ;) Or if you
> > are
> > curious and want to test it ;)
>
>
> Yes, the fan is connected and running all the time.  I plan to test it
> without the fan this weekend. I will let you know the new high temperature
> without the fan as soon as possible.
>
>
> > >
> > > In my opinion, the time constant of the silicon chip is about 10
> > > milliseconds.
> >
> > Well what constant?
>
>
> The thickness of the silicon chip is about 0.5mm, the thermal resistance
> is about R = 0.4 K/W, the thermal capacitance is about C=0.025 J/K, so the
> time constant = R*C = 0.01 second.
> The time constant of the heat sink is about several minutes.
> So it is my expectation that although the temperature of the silicon chip
> changes rapidly, the temperature of the heat sink changes very slowly (in
> minutes). So if the thermal diode is not inside the silicon layer but near
> the heat sink or heat spreader, then the temperature change will be as slow
> as the temperature curve of my experiments (or even slower).
>
> > Does this suggest that the "CPU temp" is not the temperature of the
> > > thermal diode inside the Pentium 4 CPU?
> >
> > Also from the previous mail with graphs, I think it is but system is
> > cooled
> >
> > > Oh, I just thought of one possible reason for this. Perhaps the
> > current
> > > or the voltage drop of thermal diode is not changing so fast as the
> > > temperature change of the silicon area around it. So the temperature
> > > readings from the diode will lag. If this is true, this could explain
> > > why the readings from the thermal diode is not changing as rapidly as
> > > the temperature of the silicon area around it.
> >
> > Yes but not in range of hundreds of seconds.
> >
> >
> > > I am confused here. I still think the "CPU temp" should be the
> > readings
> > > from the thermal diode inside the Pentium 4 CPU chip. But the fact
> > that
> > > this "CPU temp" is changing slowly in seconds and it could not reach a
> > > high temperature makes me doubt it.
> >
> > I think because of that hidden autoregulation. I would suggest to
> > disconnect the
> > fan connector (of course not the heatspreader ;) and try it without
> > that.
>
>
> Will the hidden autoregulation (cpu fan speed change) prevent the
> temperature of the silicon chip from increasing rapidly (from tens of
> milliseconds to tens of seconds)? I have doubt on that.
> Anyway, I plan to try it without fan working to see what will happen.
>
> Sorry for delay,
> >
> > Regards
> > Rudolf
> >
>
> Thank you very much for your ideas on this!
>
> Regards,
> Yongkui
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061025/4d6f4876/attachment.html 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

Forgot to include the attachment. Please check the attachment.

Thank you!
Yongkui


On 10/25/06, Yongkui Han <hanyongkui99 at gmail.com> wrote:
>
> Hi Rudolf,
>
> I drew the curve of the fan speed together with the "CPU Temp" and "M/B
> Temp" outputs from lm-sensors. The pdf file is attached for your reference.
>
> The fan speed is very stable before the cpuburn program starts running,
> the speed is switching between 3150 RPM and 3323 RPM.
>
> After the cpubrun program starts running, the fan speed goes up: 3323 RPM
> ---> 3413 RPM ---> 3510 RPM ---> 3614 RPM ---> 3723 RPM. And after the
> cpuburn program stops running, the fan speed goes down: 3614 RPM ---> 3510
> RPM --->  3413 RPM.
>
> Although the fan speed goes up when the CPU temperature is increasing, the
> difference of the fan speed is not so much: 3723 RPM vs. 3150 RPM. The
> difference is only 573 RPM.  Will this difference in the fan speed prevent
> the temperature of the silicon chip from increasing rapidly (from tens of
> milliseconds to tens of seconds)?
>
> Do you have any insights about this?
>
> Thanks!
> Yongkui
>
>
> On 10/24/06, Yongkui Han <hanyongkui99 at gmail.com> wrote:
> >
> > Hi Rudolf,
> >
> > Thank you for your reply.
> >
> > On 10/23/06, Rudolf Marek <r.marek at assembler.cz > wrote:
> > >
> > > Hi  Yongkui
> > >
> > > > I calculated the time constant based on the temperature curve of
> > > "CPU
> > > > temp". It is about 20 seconds.
> > >
> > > Ok. Important question: Is the fan switched off - disconnected? I
> > > think we see
> > > the "slow" start because the system cools down. The board has some
> > > kind of
> > > automatic cooling. Of course without the fan it might be bit dangerous
> > > if system
> > > temperature grows, but I think this is necessary to test it ;) Or if
> > > you are
> > > curious and want to test it ;)
> >
> >
> > Yes, the fan is connected and running all the time.  I plan to test it
> > without the fan this weekend. I will let you know the new high temperature
> > without the fan as soon as possible.
> >
> >
> > > >
> > > > In my opinion, the time constant of the silicon chip is about 10
> > > > milliseconds.
> > >
> > > Well what constant?
> >
> >
> > The thickness of the silicon chip is about 0.5mm, the thermal resistance
> > is about R = 0.4 K/W, the thermal capacitance is about C=0.025 J/K, so
> > the time constant = R*C = 0.01 second.
> > The time constant of the heat sink is about several minutes.
> > So it is my expectation that although the temperature of the silicon
> > chip changes rapidly, the temperature of the heat sink changes very slowly
> > (in minutes). So if the thermal diode is not inside the silicon layer but
> > near the heat sink or heat spreader, then the temperature change will be as
> > slow as the temperature curve of my experiments (or even slower).
> >
> > > Does this suggest that the "CPU temp" is not the temperature of the
> > > > thermal diode inside the Pentium 4 CPU?
> > >
> > > Also from the previous mail with graphs, I think it is but system is
> > > cooled
> > >
> > > > Oh, I just thought of one possible reason for this. Perhaps the
> > > current
> > > > or the voltage drop of thermal diode is not changing so fast as the
> > > > temperature change of the silicon area around it. So the temperature
> > > > readings from the diode will lag. If this is true, this could
> > > explain
> > > > why the readings from the thermal diode is not changing as rapidly
> > > as
> > > > the temperature of the silicon area around it.
> > >
> > > Yes but not in range of hundreds of seconds.
> > >
> > >
> > > > I am confused here. I still think the "CPU temp" should be the
> > > readings
> > > > from the thermal diode inside the Pentium 4 CPU chip. But the fact
> > > that
> > > > this "CPU temp" is changing slowly in seconds and it could not reach
> > > a
> > > > high temperature makes me doubt it.
> > >
> > > I think because of that hidden autoregulation. I would suggest to
> > > disconnect the
> > > fan connector (of course not the heatspreader ;) and try it without
> > > that.
> >
> >
> > Will the hidden autoregulation (cpu fan speed change) prevent the
> > temperature of the silicon chip from increasing rapidly (from tens of
> > milliseconds to tens of seconds)? I have doubt on that.
> > Anyway, I plan to try it without fan working to see what will happen.
> >
> > Sorry for delay,
> > >
> > > Regards
> > > Rudolf
> > >
> >
> > Thank you very much for your ideas on this!
> >
> > Regards,
> > Yongkui
> >
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061025/43125769/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cpuburn.fan.pdf
Type: application/pdf
Size: 5454 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061025/43125769/attachment-0001.pdf 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

I did some experiments without the CPU fan. The enclosed pdf file is for the
cpuburn benchmark.

I run the CPU burn benchmark for only 1 minute (plus 15 minutes system idle
before running cpuburn, and 15 minutes system idle after cpuburn stops
running), I even got the following message when the cpuburn program started
running for less than 1 minute:
---------------------------------------------
Message from syslogd at localhost at Mon Oct 30 11:39:21 2006 ...
localhost kernel: CPU0: Temperature above threshold

Message from syslogd at localhost at Mon Oct 30 11:39:21 2006 ...
localhost kernel: CPU0: Running in modulated clock mode

Message from syslogd at localhost at Mon Oct 30 11:39:26 2006 ...
localhost kernel: CPU0: Temperature above threshold

Message from syslogd at localhost at Mon Oct 30 11:39:26 2006 ...
localhost kernel: CPU0: Running in modulated clock mode

Message from syslogd at localhost at Mon Oct 30 11:54:30 2006 ...
localhost kernel: CPU0: Temperature above threshold

Message from syslogd at localhost at Mon Oct 30 11:54:30 2006 ...
localhost kernel: CPU0: Running in modulated clock mode
----------------------------------------------
This message means the Pentium 4 is running in modulated clock mode, I
forgot which temperature value will trigger this mechanism, perhaps it is
88C. I am not sure about this value, I will find this value in Pentium 4's
manual later and let you know.

Without the CPU fan, the CPU Temp cannot be cooled down rapidly. The CPU
Temp is staying at high temperature (>80C, CPU Temp = 84C, M/B = 64C) for
long time.

When I first turned on the computer with the CPU fan is turned off, the CPU
Temp is about 39C, the M/B Temp is 35C. They are both increasing even if
the linux operating system is idle. CPU Temp: 39C --> 40C --> 41C ......
--> 51C.  When you just run some small program such as vi, the CPU Temp will
go up: 49C --> 52C --> ...... --> 62C --> 63C.


I am wondering whether I can control the speed of the CPU fan. I would like
the CPU fan to run at a fixed speed, such as 3130 RPM or maximum RPM.

My CPU fan is Intel Fan & Heatsink A65061-001, with 3 wires, green, yellow,
and black.

Any suggestions on controlling the CPU fan speed?

Thanks!
Yongkui


On 10/25/06, Yongkui Han <hanyongkui99 at gmail.com> wrote:
>
> Hi Rudolf,
>
> Forgot to include the attachment. Please check the attachment.
>
> Thank you!
> Yongkui
>
>
>  On 10/25/06, Yongkui Han <hanyongkui99 at gmail.com> wrote:
> >
> > Hi Rudolf,
> >
> > I drew the curve of the fan speed together with the "CPU Temp" and "M/B
> > Temp" outputs from lm-sensors. The pdf file is attached for your reference.
> >
> > The fan speed is very stable before the cpuburn program starts running,
> > the speed is switching between 3150 RPM and 3323 RPM.
> >
> > After the cpubrun program starts running, the fan speed goes up: 3323
> > RPM ---> 3413 RPM ---> 3510 RPM ---> 3614 RPM ---> 3723 RPM. And after the
> > cpuburn program stops running, the fan speed goes down: 3614 RPM ---> 3510
> > RPM --->  3413 RPM.
> >
> > Although the fan speed goes up when the CPU temperature is increasing,
> > the difference of the fan speed is not so much: 3723 RPM vs. 3150 RPM. The
> > difference is only 573 RPM.  Will this difference in the fan speed prevent
> > the temperature of the silicon chip from increasing rapidly (from tens of
> > milliseconds to tens of seconds)?
> >
> > Do you have any insights about this?
> >
> > Thanks!
> > Yongkui
> >
> >
> > On 10/24/06, Yongkui Han <hanyongkui99 at gmail.com > wrote:
> > >
> > > Hi Rudolf,
> > >
> > > Thank you for your reply.
> > >
> > > On 10/23/06, Rudolf Marek <r.marek at assembler.cz > wrote:
> > > >
> > > > Hi  Yongkui
> > > >
> > > > > I calculated the time constant based on the temperature curve of
> > > > "CPU
> > > > > temp". It is about 20 seconds.
> > > >
> > > > Ok. Important question: Is the fan switched off - disconnected? I
> > > > think we see
> > > > the "slow" start because the system cools down. The board has some
> > > > kind of
> > > > automatic cooling. Of course without the fan it might be bit
> > > > dangerous if system
> > > > temperature grows, but I think this is necessary to test it ;) Or if
> > > > you are
> > > > curious and want to test it ;)
> > >
> > >
> > > Yes, the fan is connected and running all the time.  I plan to test it
> > > without the fan this weekend. I will let you know the new high temperature
> > > without the fan as soon as possible.
> > >
> > >
> > > > >
> > > > > In my opinion, the time constant of the silicon chip is about 10
> > > > > milliseconds.
> > > >
> > > > Well what constant?
> > >
> > >
> > > The thickness of the silicon chip is about 0.5mm, the thermal
> > > resistance is about R = 0.4 K/W, the thermal capacitance is about C> > > 0.025 J/K, so the time constant = R*C = 0.01 second.
> > > The time constant of the heat sink is about several minutes.
> > > So it is my expectation that although the temperature of the silicon
> > > chip changes rapidly, the temperature of the heat sink changes very slowly
> > > (in minutes). So if the thermal diode is not inside the silicon layer but
> > > near the heat sink or heat spreader, then the temperature change will be as
> > > slow as the temperature curve of my experiments (or even slower).
> > >
> > > > Does this suggest that the "CPU temp" is not the temperature of the
> > > > > thermal diode inside the Pentium 4 CPU?
> > > >
> > > > Also from the previous mail with graphs, I think it is but system is
> > > > cooled
> > > >
> > > > > Oh, I just thought of one possible reason for this. Perhaps the
> > > > current
> > > > > or the voltage drop of thermal diode is not changing so fast as
> > > > the
> > > > > temperature change of the silicon area around it. So the
> > > > temperature
> > > > > readings from the diode will lag. If this is true, this could
> > > > explain
> > > > > why the readings from the thermal diode is not changing as rapidly
> > > > as
> > > > > the temperature of the silicon area around it.
> > > >
> > > > Yes but not in range of hundreds of seconds.
> > > >
> > > >
> > > > > I am confused here. I still think the "CPU temp" should be the
> > > > readings
> > > > > from the thermal diode inside the Pentium 4 CPU chip. But the fact
> > > > that
> > > > > this "CPU temp" is changing slowly in seconds and it could not
> > > > reach a
> > > > > high temperature makes me doubt it.
> > > >
> > > > I think because of that hidden autoregulation. I would suggest to
> > > > disconnect the
> > > > fan connector (of course not the heatspreader ;) and try it without
> > > > that.
> > >
> > >
> > > Will the hidden autoregulation (cpu fan speed change) prevent the
> > > temperature of the silicon chip from increasing rapidly (from tens of
> > > milliseconds to tens of seconds)? I have doubt on that.
> > > Anyway, I plan to try it without fan working to see what will happen.
> > >
> > > Sorry for delay,
> > > >
> > > > Regards
> > > > Rudolf
> > > >
> > >
> > > Thank you very much for your ideas on this!
> > >
> > > Regards,
> > > Yongkui
> > >
> > >
> >
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061030/252e1828/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cpuburn.nocpufan.pdf
Type: application/pdf
Size: 3772 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061030/252e1828/attachment.pdf 

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Rudolf Marek]

> I run the CPU burn benchmark for only 1 minute (plus 15 minutes system 
> idle before running cpuburn, and 15 minutes system idle after cpuburn 
> stops running), I even got the following message when the cpuburn 
> program started running for less than 1 minute:
> ---------------------------------------------
> Message from syslogd at localhost <mailto:syslogd at localhost> at Mon Oct 30 
> 11:39:21 2006 ...
> localhost kernel: CPU0: Temperature above threshold

I think it is 90C so now you know that the temp is measured correctly. If you 
have 88C this may be just error of measurement. So your cooling cools quite well.

  > Any suggestions on controlling the CPU fan speed?

Please check the other thread.

Regards
Rudolf

[lm-sensors] About the "CPU temp" output of ADM1025A chip

[Yongkui Han]

Hi Rudolf,

Thank you very much!

You are very nice, and you really helped a lot.

Yongkui


On 11/5/06, Rudolf Marek <r.marek at assembler.cz> wrote:
>
> > I run the CPU burn benchmark for only 1 minute (plus 15 minutes system
> > idle before running cpuburn, and 15 minutes system idle after cpuburn
> > stops running), I even got the following message when the cpuburn
> > program started running for less than 1 minute:
> > ---------------------------------------------
> > Message from syslogd at localhost <mailto:syslogd at localhost> at Mon Oct 30
> > 11:39:21 2006 ...
> > localhost kernel: CPU0: Temperature above threshold
>
> I think it is 90C so now you know that the temp is measured correctly. If
> you
> have 88C this may be just error of measurement. So your cooling cools
> quite well.
>
> > Any suggestions on controlling the CPU fan speed?
>
> Please check the other thread.
>
> Regards
> Rudolf
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20061106/9c5bfaa1/attachment.html