Re: P4C800-E Deluxe, BIOS and Fan Speed

4 views
Skip to first unread message
Message has been deleted

Paul

unread,
May 8, 2006, 6:25:01 PM5/8/06
to
In article <Xns97BD9506D...@news.supernews.com>, Mark
<Mark@nolocation!.com> wrote:

> Hi,
>
> I'm a lot more relaxed after venting some frustration with Antec in here
> recently. Although my opinion of them hasn't changed, I'd like to thank
> those that replied and allowed me to vent.
>
> The solution to the power supply fan speed has raised a very
> interesting, yet confusing circumstance, the disabling in BIOS of the
> Power Fan Speed sensor.
>
> When the Power Fan Speed sensor was enabled in BIOS, ASUS PC Probe
> failed to read the speed of the power supply fan below 1328 RPM. Now
> that the Power Fan Speed sensor is disabled in BIOS, it not only reads
> below 1328 RPM accurately, but that it's readable at all.
>
> Anyone have any thoughts on why disabling the Power Fan Speed sensor in
> BIOS still allows PC Probe to read the Power Fan Speed sensor and cure
> the less than 1328 RPM measurement problem?

PC Probe is likely using a different scaling divisor. That
changes the range of RPMs that can be handled. Maybe PC Probe
respects the value the BIOS is using, and won't overwrite it.
Or, the BIOS continuously refreshes its preferred divisor.
Plenty of possibilities.

You might also experiment with Speedfan from almico.com . Remove
Asus Probe, and run one monitor program at a time.

All of this could largely be avoided, if the Super I/O chip
that houses the monitoring function, used wider counters for
the fan RPMs.

Paul

Message has been deleted

Paul

unread,
May 8, 2006, 11:02:06 PM5/8/06
to

Mark wrote:
>
> You wrote in news:nospam-0805...@192.168.1.178 thusly:


>
> >> The solution to the power supply fan speed has raised a very
> >> interesting, yet confusing circumstance, the disabling in BIOS of the
> >> Power Fan Speed sensor.
> >>
> >> When the Power Fan Speed sensor was enabled in BIOS, ASUS PC Probe
> >> failed to read the speed of the power supply fan below 1328 RPM. Now
> >> that the Power Fan Speed sensor is disabled in BIOS, it not only
> >> reads below 1328 RPM accurately, but that it's readable at all.
> >>
> >> Anyone have any thoughts on why disabling the Power Fan Speed sensor
> >> in BIOS still allows PC Probe to read the Power Fan Speed sensor and
> >> cure the less than 1328 RPM measurement problem?
> >
> > PC Probe is likely using a different scaling divisor. That changes the
> > range of RPMs that can be handled. Maybe PC Probe respects the value
> > the BIOS is using, and won't overwrite it. Or, the BIOS continuously
> > refreshes its preferred divisor. Plenty of possibilities.
>

> Please correct me if I'm wrong, but I would have thought that if the PC
> Probe is respecting the BIOS, with it set to disable, PC Probe would not
> be reading anything for the power supply fan. In fact, I would have
> expected PC Probe's GUI to remove the Power Fan history completely from
> the PC Probe window.


>
> > You might also experiment with Speedfan from almico.com . Remove Asus
> > Probe, and run one monitor program at a time.
>

> While I didn't uninstall Asus PC Probe, I did run Everest and SpeedFan.
> Each one without the other running and the Power Fan Speed sensor
> enabled in BIOS.
>
> Everest didn't show a power supply fan entry under Cooling Fans at all
> until the speed went above 1328 RPM, then the entry appeared in the GUI
> and the RPM was measured. SpeedFan just showed zero RPM, until 1328 RPM
> was reached.


>
> > All of this could largely be avoided, if the Super I/O chip that

> > houses the monitoring function, used wider counters for the fan RPM.
>
> FWIW, this all came about since the power supply exchange. Prior to
> this, I never had PC Probe (same version) popping up and beeping at me
> to inform me that the power supply fan was at zero RPM. And it would
> never have outright occurred to me that the solution was to disable the
> Power Supply Fan sensor in BIOS to get it to work correctly. :)
>
> --
>
> Mark

I have a motherboard with an 1800RPM minimum. The power supply on that
thing, always starts at about 1750RPM or so, causing the BIOS to flag
the slow fan and stop the boot. By the time I get into the BIOS, the
power supply is running just over 1800 RPM, and everything is back to
normal. Being lazy, I never bothered to do anything about that,
and on a warm boot, it doesn't cause problems, so there isn't a lot
of incentive to fix it.

I just installed Speedfan on this machine (my antique gaming system),
and in Configure:Advanced there is a pulldown menu so you can select
the monitor chip. On this motherboard, the monitor is ASB100. If I set
the divisors of the three fans to 8, I can measure the slowest fan in
the system, which is the PSU fan. When Speedfan started, the defaults
appeared to be 2. It would seem that Speedfan is not forcing the issue,
at least until you use Configure:Advanced to change it.

The reason I noticed that the value made a difference, is I started
MBM5 running, while Speedfan was also running. MBM5 seems to have
set the three divisors to 8, and suddenly I noticed my PSU fan
speed starting to register in Speedfan. So it looks like a higher
divisor value, like the maximum of 8, allows lower fan speeds to be
measured. Maybe the divisor values listed are a function of the
monitor chip ? I hope Speedfan is not just stopping at 8 on a whim
of the programmer.

So it looks like the initial value for the divisor might be 2, and
Speedfan doesn't change it unless you use the Configure option. MBM5
on the other hand, seems to have selected a value for itself.

Paul

Message has been deleted

Paul

unread,
May 9, 2006, 3:24:09 AM5/9/06
to

Mark wrote:
>
> You wrote in news:4460062D...@needed.com thusly:


>
> > I have a motherboard with an 1800RPM minimum. The power supply on that
> > thing, always starts at about 1750RPM or so, causing the BIOS to flag
> > the slow fan and stop the boot. By the time I get into the BIOS, the
> > power supply is running just over 1800 RPM, and everything is back to
> > normal. Being lazy, I never bothered to do anything about that, and on
> > a warm boot, it doesn't cause problems, so there isn't a lot of
> > incentive to fix it.
>

> Yep, I've seen the red flashing RPM indicator for the power supply fan
> sensor in BIOS before, but up until recently once in WinXP that has
> never prevented Asus PC Probe software from reading the RPM of the power
> supply's fan.


>
> > I just installed Speedfan on this machine (my antique gaming system),
> > and in Configure:Advanced there is a pulldown menu so you can select
> > the monitor chip. On this motherboard, the monitor is ASB100. If I set
> > the divisors of the three fans to 8, I can measure the slowest fan in
> > the system, which is the PSU fan. When Speedfan started, the defaults
> > appeared to be 2. It would seem that Speedfan is not forcing the
> > issue, at least until you use Configure:Advanced to change it.
> >
> > The reason I noticed that the value made a difference, is I started
> > MBM5 running, while Speedfan was also running. MBM5 seems to have set
> > the three divisors to 8, and suddenly I noticed my PSU fan speed
> > starting to register in Speedfan. So it looks like a higher divisor
> > value, like the maximum of 8, allows lower fan speeds to be measured.
> > Maybe the divisor values listed are a function of the monitor chip ? I
> > hope Speedfan is not just stopping at 8 on a whim of the programmer.
> >
> > So it looks like the initial value for the divisor might be 2, and
> > Speedfan doesn't change it unless you use the Configure option. MBM5
> > on the other hand, seems to have selected a value for itself.
>

> Upgraded to version 4.28 from version 4.24 of Speedfan. In both versions
> of Speedfan, and both settings for FAN3 (ie: mult and div) are as
> follows:
>
> With the Power Fan Speed sensor disable, the values are 1
> With the Power Fan Speed sensor enabled, the values are 1
>
> And I know a value of one works fine with the Power Fan Speed sensor
> disabled in BIOS and RPM lower than 1328. Because of this, I don't think
> that this explains why disabling the Power Speed sensor fixes the
> problem, since both mult and div are set to one in all instances.
>
> Does Speedfan, PC Probe or whatever you run still report a rotating fan
> when you disable the Power Fan Speed sensor in BIOS?
>
> Thank you for replying,
>
> Mark

The "mult" and "div" appear to be scale factors applied to the results.
For example, my CPU fan reads 5200 RPM or thereabouts. The correct speed
is actually 2600 RPM. If I set "div" to "2" in Advanced:Config, Speedfan
then reads correctly as 2600 RPM. So mult and div are simple scale
factors
applied to the measurement.

What I was talking about is the "divisor". Increasing the "divisor",
reduces the minimum readable RPM. If my divisor is set to "2", the
BIOS written default, I cannot read my power fan. It reads 0 in
Speedfan. If I set the "divisor" to 8, suddenly I can read the fan.

So all my "div" values should be 2, as my Speedfan results are 2x
too fast without it. And by using 8 for the divisor, I'm able to
read slower fans, than with divisor values of 4 or 2. Motherboard
Monitor (MBM5 - mbm.livewiredev.com) is smart enough to set the divisors
all to 8 for me.

I went into the BIOS, as you requested, and in fact I cannot tell the
BIOS to do anything! My monitor output on this motherboard, just
gives readouts, and there is no ability to toggle any fields. The
only setting on the page I can change, is to enable or disable
Q-fan. So this is not really a good motherboard to experiment with
(A7N8X-E).

This is from a previous post I wrote a while back:

***************************************************************
A fan pulses twice per revolution. The monitor chip uses a high
speed clock, to measure the number of counts between pulse edges

0123456789 <--- register contains 0x09
|<-------->|
__ __
| | | |
____| |_______| |______

As the fan speed drops, the pulses move further apart. The
measurement register is getting a little bit fuller.

0123456789ABC <--- register contains 0x0C
|<----------->|
__ __
| | | |
____| |__________| |______

The inverse of the measured period, yields the frequency
and thus the RPMs.

If the fan goes slow enough, the register contains 0xFF
That is the minimum speed that can be measured.
If the fan stops running altogether, the register will also
read 0xFF. The software has no way of knowing what the speed
is, once the speed drops below the minimum, and the measurement
register is full.

There is a programmable divisor. It changed the rate of
the measurement clock. In this example, the sampling clock
is running at one half the speed of the examples above

0 1 2 3 4 5 6 <--- register contains 0x06
|<----------->|
__ __
| | | |
____| |__________| |______

Some monitoring programs allow changing that divisor, and
then the minimum fan speed can be improved. The BIOS
uses divide by 2 as the default (AFAIK), and the monitoring
chip can do better than that.
***************************************************************

Paul

Paul

unread,
May 9, 2006, 5:12:23 AM5/9/06
to
In article <Xns97BDDD350...@news.supernews.com>, Mark
<Mark@nolocation!.com> wrote:

> You wrote in news:4460062D...@needed.com thusly:

>
> > I have a motherboard with an 1800RPM minimum. The power supply on that
> > thing, always starts at about 1750RPM or so, causing the BIOS to flag
> > the slow fan and stop the boot. By the time I get into the BIOS, the
> > power supply is running just over 1800 RPM, and everything is back to
> > normal. Being lazy, I never bothered to do anything about that, and on
> > a warm boot, it doesn't cause problems, so there isn't a lot of
> > incentive to fix it.
>

> Yep, I've seen the red flashing RPM indicator for the power supply fan
> sensor in BIOS before, but up until recently once in WinXP that has
> never prevented Asus PC Probe software from reading the RPM of the power
> supply's fan.
>

> > I just installed Speedfan on this machine (my antique gaming system),
> > and in Configure:Advanced there is a pulldown menu so you can select
> > the monitor chip. On this motherboard, the monitor is ASB100. If I set
> > the divisors of the three fans to 8, I can measure the slowest fan in
> > the system, which is the PSU fan. When Speedfan started, the defaults
> > appeared to be 2. It would seem that Speedfan is not forcing the
> > issue, at least until you use Configure:Advanced to change it.
> >
> > The reason I noticed that the value made a difference, is I started
> > MBM5 running, while Speedfan was also running. MBM5 seems to have set
> > the three divisors to 8, and suddenly I noticed my PSU fan speed
> > starting to register in Speedfan. So it looks like a higher divisor
> > value, like the maximum of 8, allows lower fan speeds to be measured.
> > Maybe the divisor values listed are a function of the monitor chip ? I
> > hope Speedfan is not just stopping at 8 on a whim of the programmer.
> >
> > So it looks like the initial value for the divisor might be 2, and
> > Speedfan doesn't change it unless you use the Configure option. MBM5
> > on the other hand, seems to have selected a value for itself.
>

> Upgraded to version 4.28 from version 4.24 of Speedfan. In both versions
> of Speedfan, and both settings for FAN3 (ie: mult and div) are as
> follows:
>
> With the Power Fan Speed sensor disable, the values are 1
> With the Power Fan Speed sensor enabled, the values are 1
>

> And I know a value of one works fine with the Power Fan Speed sensor

> disabled in BIOS and RPM lower than 1328. Because of this, I don't think
> that this explains why disabling the Power Speed sensor fixes the
> problem, since both mult and div are set to one in all instances.
>
> Does Speedfan, PC Probe or whatever you run still report a rotating fan
> when you disable the Power Fan Speed sensor in BIOS?
>
> Thank you for replying,

OK, I dug out my P4C800-E and set it up.

The summary is, there is no difference in behavior, whether
the Power Fan monitoring is disabled in the BIOS or not. I'm
using BIOS 1014 (as shipped from the factory).

I did get to see some interesting behavior.

1) With just the CPU fan (all that is normally connected on that
machine), the divisor settings go

Fan1 128
Fan2 4 (2657 RPM)
Fan3 128

What that means, is SpeedFan in this case is "autoranging".
It increases the divisor a step at a time, then checks to
see if the register stops overflowing. It went all the way
to 128, hoping to see some pulses on Fan1 and Fan3, but as
there was no fan on the header, no pulses were found.

2) I connected my Antec power fan monitor cable to the P4C800-E
PWR fan header. I connected a spare Vantec Stealth to
the CHA fan header. After a fresh boot, upon looking in
SpeedFan, I get

Fan1 (CHA) 8 (1442 RPM)
Fan2 (CPU) 4 (2657 RPM)
Fan3 (PWR) 8 (1318 RPM)

Now, if I start MBM5, the divisor settings seem to be
changed by MBM5.

Fan1 (CHA) 4
Fan2 (CPU) 4
Fan3 (PWR) 4

The setting of 4 for the divisor, appears to be sufficient
to prevent the measurement register from overflowing. So,
the SpeedFan autoranging algorithm seems to want to leave
a bit more dynamic range in the register (not trying to
run it quite as close to full).

I don't know if MBM5 is attempting to autorange, or if the
value was randomly picked by the MBM5 author. Because the
ratio of Fan2 to Fan3 is slightly more than 2, MBM5 shouldn't
have picked the same settings for all three fans, if its
divisor sleection algorithm was smart.

I didn't see an anomaly when the BIOS monitoring was disabled.
What was I supposed to see ? Whether the divisors were 8,4,8
from Speedfan itself, or modified to 4,4,4 by MBM5, I got
readings in both cases. Only when I set a divisor on the two
slower fans to 2, did I get a zero reading.

In looking at the MBM5 configuration options, MBM5 has both
an option to change the divisor and an option to set a
scale factor for the computed result. If you change the
MBM5 divisor setting, MBM5 is too stupid to compute the
RPMs correctly. I had to move the scaling slider when I
moved the divisor slider. At least SpeedFan gets that part
right.

My experiences so far leave me less than impressed. Between
the authors of those two programs, they should have
released source, so we could fix this stuff right :-)
For example, on SpeedFan, it lists some "Temps", but I
cannot tell which is which. Having the source would at
least allow me to determine which chunk of hardware
that the readings are coming from.

Paul

Message has been deleted
Message has been deleted

Paul

unread,
May 17, 2006, 9:04:29 PM5/17/06
to
In article <Xns97C660FC7...@news.supernews.com>, Mark
<Mark@nolocation!.com> wrote:

> You wrote in news:44604399...@needed.com thusly:

>
> > I went into the BIOS, as you requested, and in fact I cannot tell the
> > BIOS to do anything! My monitor output on this motherboard, just gives
> > readouts, and there is no ability to toggle any fields. The only
> > setting on the page I can change, is to enable or disable Q-fan. So
> > this is not really a good motherboard to experiment with (A7N8X-E).
>

> Please see my other reply to your other post where you describe using
> your P4C800-E Deluxe for more. With regards to this, I'm surprised that
> option isn't available, isn't that a newer board?
>
> Here's an update on what I've tried so far.
>
> I had a hell of a time finding the ASUS PC Probe software on the ASUS
> site. The way the ASUS site is organized, with regard to some things, is
> atrocious. I finally found the "trick" to get the download page to
> display current and pass releases of PC Probe.
>
> There was a PC Probe with a newer upload _date_, but it had an older
> version number. I downloaded that version and version 22410 at a head
> twisting speed of 2K/s, hmm. As I had expected, the contents of this
> newer uploaded PC Probe had much older file dates inside the archive.
> Another poor design of the ASUS site, sort by date only.
>
> I tried some ASUS forum searches, but their search engine is broken.
>
> Anyway, I uninstalled PC Probe 22410 and reinstalled the newly
> downloaded PC Probe 22410, the results were the same - fan goes below
> 1328 RPM and PC Probe starts warning that the rotation is at zero RPM.
>
> I attempted to contact ASUS, but after being on hold for 25 minutes,
> someone came on the line, took my name and number and said someone would
> call back in 24 hours. That was 6 days ago! No wonder the site is
> designed the way it is.
>
> I hope you had a chance to read my other reply. I don't want you to
> think I ignored you when you were kind enough to dig out your P4C800-E
> Deluxe motherboard for testing.

I agree with your comments about the download web page. I tried to
find the most recent version of Asus Probe, and they've been spread
all over the place. Using the download page, normally a search
for "Tools" brings you to the most recent version.

Long story short, this is the best I could find today, for someone
else. Note that there are two development streams of Asus Probe.
There is Asus Probe and Asus Probe II. Apparently the latter
version is even worst than the original, in terms of a feature set.
I don't use Asus Probe that often myself, so I cannot tell you
what the differences between the two of them are.

ftp://ftp.asus.com/pub/ASUS/misc/utils/PCProbe22502.zip

Paul

John Malone

unread,
May 18, 2006, 8:35:02 PM5/18/06
to
I have been lurking watching and intrigued by what you have been going
through.
I to have a P4C800-E Deluxe and haven't used the PCProbe in a long time SO.
I downloaded the 22410 version and gave it a whirl. To my surprise it said
the CPU fan was 0.
I knew that the probe worked when I first got the board so I took the CD
that came with the motherboard and installed that version of PCProbe. Same
results zero 0 for the CPU fan.

Now this is the interesting part.... I had the probe open and recording
while I was doing some editing of videos (a very processor intensive
operation) and watching the temp of the CPU. The temp seem to go up very
fast until it got to 60C and then it started back down... What? Why did it
start downward. I checked on the fans and low and behold the CPU fan was
running at 4600 rpm.

After the processing of the video and the CPU usage went down the computer
started to cool and the fan speed dropped. When it got to 3200 rpm it then
went to 0 again. I am having the same problem? Well I ripped open the case
to investigate. I fired it back up and the CPU fan was spinning - PCProbe
said 3800 rpm and a CPU temp of 58C (I guess shutting it down so fast it was
still kind of hot). So OK the fan works... NOT! While the temp came down
to about 48C The fan slowed down to a stop! WOW! I couldn't believe it.
As the temp went back up the fan started again.

Needless to say I shut it down and replaced the heat sink and fan. Now all
is well and life is back to normal.

I feel good that the impending disaster was avoided and I didn't loose my
computer.

Thank You.....

--
John Malone
==============
"Mark" <Mark@nolocation!.com> wrote in message
news:Xns97C660FC7...@news.supernews.com...


|
| You wrote in news:44604399...@needed.com thusly:
|

| > I went into the BIOS, as you requested, and in fact I cannot tell the
| > BIOS to do anything! My monitor output on this motherboard, just gives
| > readouts, and there is no ability to toggle any fields. The only
| > setting on the page I can change, is to enable or disable Q-fan. So
| > this is not really a good motherboard to experiment with (A7N8X-E).
|

| Please see my other reply to your other post where you describe using
| your P4C800-E Deluxe for more. With regards to this, I'm surprised that
| option isn't available, isn't that a newer board?
|
| Here's an update on what I've tried so far.
|
| I had a hell of a time finding the ASUS PC Probe software on the ASUS
| site. The way the ASUS site is organized, with regard to some things, is
| atrocious. I finally found the "trick" to get the download page to
| display current and pass releases of PC Probe.
|
| There was a PC Probe with a newer upload _date_, but it had an older
| version number. I downloaded that version and version 22410 at a head
| twisting speed of 2K/s, hmm. As I had expected, the contents of this
| newer uploaded PC Probe had much older file dates inside the archive.
| Another poor design of the ASUS site, sort by date only.
|
| I tried some ASUS forum searches, but their search engine is broken.
|
| Anyway, I uninstalled PC Probe 22410 and reinstalled the newly
| downloaded PC Probe 22410, the results were the same - fan goes below
| 1328 RPM and PC Probe starts warning that the rotation is at zero RPM.
|
| I attempted to contact ASUS, but after being on hold for 25 minutes,
| someone came on the line, took my name and number and said someone would
| call back in 24 hours. That was 6 days ago! No wonder the site is
| designed the way it is.
|
| I hope you had a chance to read my other reply. I don't want you to
| think I ignored you when you were kind enough to dig out your P4C800-E
| Deluxe motherboard for testing.
|

| --
|
| Mark
|


Reply all
Reply to author
Forward
0 new messages