ASUS A7N8X-E Fan RPM Reads Zero Problem RESOLVED (Sort-Of)

Hello,

I have a new ASUS A7N8X-E motherboard which shows fan RPM as ZERO
except for the case that the fan is actually being POWERED from the +12V
terminal of the three-pin fan header. Now this is not a problem for the chassis
fan; but it IS an issue with the Power Supply Fan (draws power internally
from the PS) and the CPU fan which is a high-performance model that exceeds
the load capacity available from the fan header (350mA ~ 740mA).

Both the PS and CPU fans provide a fan speed sense connector for the purpose
of RPM monitoring. The PS has wires for both ground and RPM sense. The CPU
fan has only a wire for the RPM sense. Until I ran across this board, the RPM of
both of these could be read out without difficulty despite the fact that they drew
power from another source.

I flashed the board to the LATEST BIOS and the problem still persists. Anybody
know of a way to get around this problem? What would happen if I were to put some
kind of artificial load on the +12V fan header output? Would the ground line also have
to be connected?

*** THESE THINGS USED TO WORK JUST FINE "AS IS" IN THE PAST!!! ***


TO ASUS:
-----------------------
Is this a design flaw or was it done on purpose? Obviously, I MUST be able to
monitor these CRITICAL fans - otherwise I could fry an expensive PS and CPU should
a fan go down and I'd not get any alarm (because I'd have to disable it in order to run
with this board). Yes, over-temp should catch the CPU but by then it may be too late!!
I don't know what protection the PS has built into it (an Antec True 430-watt).

TO AJUMP.COM:
-----------------------
If this cannot be repaired (meaning this is a design flaw - EVERY board will exhibit
the problem) how do I get my money back on it? I also am worried that the ASUS
A7V880 I just had you guys replace under RMA may have the same issue!! I am going
to test this theory as soon as possible. What happens if I have to RMA BOTH boards
requesting refund of the purchase price? Reply to my e-mail as per the instructions in my
signature (or look up customer #11A048238). Thanks.

=====
You may send e-mail if you wish provided it is NOT SPAM. To decode my valid
E-mail address, you will need to remove the NOSPAM. and the 666. I had
to get complicated as spammers are now doing the obvious removal of the word
"NOSPAM" to compile their lists of e-mails...

- G.L. Cross

On Tue, 28 Sep 2004 04:00:28 -0500, "G.L. Cross"
wrote:

Hello,

I have a new ASUS A7N8X-E motherboard which shows fan RPM as ZERO
except for the case that the fan is actually being POWERED from the +12V
terminal of the three-pin fan header. Now this is not a problem for the chassis
fan; but it IS an issue with the Power Supply Fan (draws power internally
from the PS) and the CPU fan which is a high-performance model that exceeds
the load capacity available from the fan header (350mA ~ 740mA).

Use a different fan. There is no CPU you could run in that
board that needs 350-740mA fan. Even 250mA is excessive
unless you have very poor heatsink.

Let's suppose you're trying for a potentially-damaging,
extreme overclock... In that situation, you'd need water
cooling or other exotic measures, and a board that uses 12V
for CPU power. That board uses 5V, and will drop the
voltage too much before you'd ever begin to need 350mA fan.


Both the PS and CPU fans provide a fan speed sense connector for the purpose
of RPM monitoring. The PS has wires for both ground and RPM sense. The CPU
fan has only a wire for the RPM sense. Until I ran across this board, the RPM of
both of these could be read out without difficulty despite the fact that they drew
power from another source.

Maybe your board is defective?
Maybe you've blown out the fan header somehow?
Maybe someone has falsely lead you to believe your goal is
more important than it is?

The number one way to combat fan failure is to use a good
quality fan at moderate RPM. If you fear yours will fail,
you should replace the fans now with something better...
regardless of whether it works with RPM monitoring on your
board.



I flashed the board to the LATEST BIOS and the problem still persists. Anybody
know of a way to get around this problem? What would happen if I were to put some
kind of artificial load on the +12V fan header output? Would the ground line also have
to be connected?

Artificial load will get you nothing.



*** THESE THINGS USED TO WORK JUST FINE "AS IS" IN THE PAST!!! ***


TO ASUS:
-----------------------
Is this a design flaw or was it done on purpose? Obviously, I MUST be able to
monitor these CRITICAL fans - otherwise I could fry an expensive PS and CPU should
a fan go down and I'd not get any alarm (because I'd have to disable it in order to run
with this board). Yes, over-temp should catch the CPU but by then it may be too late!!
I don't know what protection the PS has built into it (an Antec True 430-watt).


To user: We do not support this configuration. Attach
normal lower amperage fans to each header to see if the
function works. Clear CMOS and try again. If fan headers
then do not work your board is defective.


TO AJUMP.COM:
-----------------------
If this cannot be repaired (meaning this is a design flaw - EVERY board will exhibit
the problem) how do I get my money back on it? I also am worried that the ASUS
A7V880 I just had you guys replace under RMA may have the same issue!! I am going
to test this theory as soon as possible. What happens if I have to RMA BOTH boards
requesting refund of the purchase price? Reply to my e-mail as per the instructions in my
signature (or look up customer #11A048238). Thanks.

Multiple boards having same problem?
I'm beginning to suspect that you're wiring something wrong
or that the fans don't actually have RPM output, maybe a
stall sensor lead instead, which isn't compatible with
motherboard headers. On the other hand you report that it
worked previously, so two possibilities I see:

Power supply fan is thermal-controlled by power supply, and
it's RPM is too low to register (not uncommon). One hopes
for a bios update to address this, but it may or may not
ever come. Fan without RPM (thermal) control and more
common RPM (like 2200-5000) should be tried on that header
to rule out damage or defect.

CPU fan above monitor's tolerable range, RPM TOO high.
Perhaps not, it's far less common but then it's not common
to use such a high-powered fan either. When a fan header is
calibrated to register smaller range of values, it should
have higher accuracy(?) so it's not necessarily a bad thing
for most people.


Use high quality fans and moderate RPM and don't worry about
fan failure till it happens, OR replace motherboard, hoping
for luck from Asus or vendor. I'd choose the former.

"kony" wrote in message news
On Tue, 28 Sep 2004 04:00:28 -0500, "G.L. Cross"
wrote:

Hello,

.... snipped ...

Use a different fan. There is no CPU you could run in that
board that needs 350-740mA fan. Even 250mA is excessive
unless you have very poor heatsink.

The number one way to combat fan failure is to use a good
quality fan at moderate RPM. If you fear yours will fail,
you should replace the fans now with something better...
regardless of whether it works with RPM monitoring on your
board.

Duh?? Ever heard of Swiftech?? 1/4" solid copper heatsink base with
100+ milled aluminum fins. Fan came with it and runs at 870mA when
full bore but it is rheostat controlled. At full power, it can keep an
Athlon-XP 3000+ running Prime95 non-stop to about 86 deg F in a
70 deg room with the case sealed. If I shut down the system, something
about that copper chills it to 4 - 6 degrees BELOW the ambient room
temperature (with no active cooling processes operating). This verified
with a temperature probe sensitive to 1/100 of a degree borrowed from
the University here. Don't ask: I have not figured out why the copper metal
stays colder than the room (it doesn't make logical sense)...

And yes, I verified that all three headers register provided the fan is receiving
its power from the MB header.


- G

On Tue, 28 Sep 2004 16:35:36 -0500, "G.L. Cross"
wrote:

"kony" wrote in message news
On Tue, 28 Sep 2004 04:00:28 -0500, "G.L. Cross"
wrote:

Hello,

... snipped ...

Use a different fan. There is no CPU you could run in that
board that needs 350-740mA fan. Even 250mA is excessive
unless you have very poor heatsink.

The number one way to combat fan failure is to use a good
quality fan at moderate RPM. If you fear yours will fail,
you should replace the fans now with something better...
regardless of whether it works with RPM monitoring on your
board.

Duh?? Ever heard of Swiftech?? 1/4" solid copper heatsink base with
100+ milled aluminum fins. Fan came with it and runs at 870mA when
full bore but it is rheostat controlled.

Actually it doesn't, actual current is below the 870mA
rating in use but it's beside the point.
There is no useful purpose to putting a 870mA fan on an
XP3000.. even my overvolted XP @ 2.5 GHz uses a 120 mA fan.

At full power, it can keep an
Athlon-XP 3000+ running Prime95 non-stop to about 86 deg F in a
70 deg room with the case sealed. If I shut down the system, something
about that copper chills it to 4 - 6 degrees BELOW the ambient room
temperature (with no active cooling processes operating).

Magic copper chills it below room temp? Nope.


This verified
with a temperature probe sensitive to 1/100 of a degree borrowed from
the University here. Don't ask: I have not figured out why the copper metal
stays colder than the room (it doesn't make logical sense)...

It's impossible.
Recheck everything, either your ambient temp is wrong, the
copper/CPU temp is wrong, or the temp probe is
miscalibrated.

It's all pointless though, the only thing such a high RPM
fan accomplishes is to reduce it's own lifespan, increase
dust accumulation, and make a silly amount of noise. There
is no useful purpose to attaining room temp. Further, the
return on higher fan speed is quite low, the copper base of
the 'sink plays by far the largest role in temp.


And yes, I verified that all three headers register provided the fan is receiving
its power from the MB header.

So essentially the board is working properly, it just
doesn't support the 3rd party mods you're trying.
The system seems to work fine, enjoy it instead of putting
yourself though these hassles.

"kony" wrote in message ...
On Tue, 28 Sep 2004 16:35:36 -0500, "G.L. Cross"
wrote:
If I shut down the system, something
about that copper chills it to 4 - 6 degrees BELOW the ambient room
temperature (with no active cooling processes operating).

Magic copper chills it below room temp? Nope.

Search me... I did e-mail my friend at the physics department and he said that one
way this could happen is via evaporative cooling. Probably moisture left on the metal
after handling it: evaporation is a cooling process and the metal is a great thermal
conductor and could hold the chilled state for some time depending on the conditions.
(I did take the temp readings within about 30 minutes after setting up for the "test").
Hell, it's like UFOs: I see the temp drop below the ambient temp and can't explain it
walking away wondering "wierd!" If I had come back and measured again in a couple
of hours, he said it would have equalized with the room. Anyway your "challenge" to
me did induce me to solve the mystery: "Bad experimental technique due to an unanticipated
effect," he said.

And yes, I verified that all three headers register provided the fan is receiving
its power from the MB header.

So essentially the board is working properly, it just
doesn't support the 3rd party mods you're trying.
The system seems to work fine, enjoy it instead of putting
yourself though these hassles.

Exactly what "3rd party mods" do yoou think I'm trying? The fan was factory-made with
only the speed-sense wire for the MB header. The PS was factory made in the same manner.
The RPM values are supposed to be available even if the speed-sense wire is the only
one connected to the MB header. There should be no requirement that the fan be
powered from the header +12V source. And yes, it would be the case that fan speed
control technology like ASUS Q-Fan would be unable to control a fan powered
from an external source...

If what you say is by design, then it doesn't make sense that the board provides for
monitoring of the PS fan. I've never seen a PS whose fan was powered and controlled
by a source exterior to the PS itself (most PS use their own form of fan speed control).
Can you name a PS that incorporates an externally-controlled fan (allowing the MB to
set the speed)?

I did do an experiment and determined that the BIOS reads out ZERO RPM if the
rotation drops below about 2200 or goes above about 4800 as someone had suggested
could happen. Setting cutoffs here doesn't make any sense: why do they not just display
the count and be done with it (or put a low bound at say 100 RPM which would be about
as effective as ZERO). 2200 RPM (or even half that) can still cool much better than a
non-operative fan! However if I connect only the speed-sense wire (of this very same fan)
to the header while powering it from another source at RPM within the range above, the BIOS
still reads it as ZERO (that's sort of like claiming you are not moving just because you took
your foot off the accelerator when, in fact, you're actually still coasting at 60 mph).

The ****er in my opinion is that the manual does not discuss any high/low cutoffs or any
other description about the RPM readout or even Q-Fan. Of course the manuals never seem to
explain anything anyway... 8-()


- G

On Wed, 29 Sep 2004 03:09:10 -0500, "G.L. Cross"
wrote:


"kony" wrote in message ...
On Tue, 28 Sep 2004 16:35:36 -0500, "G.L. Cross"
wrote:
If I shut down the system, something
about that copper chills it to 4 - 6 degrees BELOW the ambient room
temperature (with no active cooling processes operating).

Magic copper chills it below room temp? Nope.

Search me... I did e-mail my friend at the physics department and he said that one
way this could happen is via evaporative cooling. Probably moisture left on the metal
after handling it: evaporation is a cooling process and the metal is a great thermal
conductor and could hold the chilled state for some time depending on the conditions.

Except that:

1) There was never a "chilled state".
2) There is no moisture buildup on a piece of metal heated
above room temp with (any, but especially) that tornado
speed fan on it).


(I did take the temp readings within about 30 minutes after setting up for the "test").
Hell, it's like UFOs: I see the temp drop below the ambient temp and can't explain it
walking away wondering "wierd!" If I had come back and measured again in a couple
of hours, he said it would have equalized with the room. Anyway your "challenge" to
me did induce me to solve the mystery: "Bad experimental technique due to an unanticipated
effect," he said.

It is a curiosity, but one that if it went unsolved, would
not be a great loss... it really doesn't matter to the
extent that it effects the system, unless you had
overheating problems and needed to know what actual temps
were, since one of the readings is clearly wrong.


And yes, I verified that all three headers register provided the fan is receiving
its power from the MB header.

So essentially the board is working properly, it just
doesn't support the 3rd party mods you're trying.
The system seems to work fine, enjoy it instead of putting
yourself though these hassles.

Exactly what "3rd party mods" do yoou think I'm trying? The fan was factory-made with
only the speed-sense wire for the MB header. The PS was factory made in the same manner.
The RPM values are supposed to be available even if the speed-sense wire is the only
one connected to the MB header. There should be no requirement that the fan be
powered from the header +12V source. And yes, it would be the case that fan speed
control technology like ASUS Q-Fan would be unable to control a fan powered
from an external source...

Did fan come with motherboard? Nope. It is non-standard, a
3rd party device. So is PSU.
A standard fan with RPM monitor plugs into a single
pin-header. As I wrote previously, it's a choice, Q-Fan and
other speed control technologies or not... often new
features mean compatibility quirks.


If what you say is by design, then it doesn't make sense that the board provides for
monitoring of the PS fan.

It does if you plug fan into it, not into another power
source.

I've never seen a PS whose fan was powered and controlled
by a source exterior to the PS itself (most PS use their own form of fan speed control).
Can you name a PS that incorporates an externally-controlled fan (allowing the MB to
set the speed)?

Generally a PC system does not monitor PSU fan at all. It
may be that a mobo has a header labeled as such, after all
they have a bunch of headers and to differentiate for their
software monitor (like PCProbe) the headers ought to be
labeled such that CPU, had reading for fan plugged into that
header.

Asus has had Q tech for a few years now, if it don't like
it, don't buy boards supporting it or any similar function.


I did do an experiment and determined that the BIOS reads out ZERO RPM if the
rotation drops below about 2200 or goes above about 4800 as someone had suggested
could happen. Setting cutoffs here doesn't make any sense: why do they not just display
the count and be done with it (or put a low bound at say 100 RPM which would be about
as effective as ZERO). 2200 RPM (or even half that) can still cool much better than a
non-operative fan!

I'd expect it boils down to polling interval and precision
of readout. It's not like Asus is the only company to have
products with this issue.


However if I connect only the speed-sense wire (of this very same fan)
to the header while powering it from another source at RPM within the range above, the BIOS
still reads it as ZERO (that's sort of like claiming you are not moving just because you took
your foot off the accelerator when, in fact, you're actually still coasting at 60 mph).

.... or it's like claiming that if you're standing still on a
bus that's moving, that _relatively_ you're not moving even
if the bus is.


The ****er in my opinion is that the manual does not discuss any high/low cutoffs or any
other description about the RPM readout or even Q-Fan. Of course the manuals never seem to
explain anything anyway... 8-()


Well they COULD provide extensive detail of all subsystems
similarly, and the volumes of books necessary would cost
more than the board itself. I doubt that's the tradeoff you
really want, and it can't be expected that they'd just add
whatever feature descriptions that benefit your particular
non-typical use by forture-telling what your needs will be,
contrasted with anyone else's non-standard needs.

So sure, in an ideal world it'd work, but ultimately if you
used the better arrangement for CPU fan that reading would
be resolved, and not having PSU fan reading merely makes
your system like the majority out there... boring to know
perhaps, but in the end the system runs, you USE it and move
on, keeping speed-control technology in mind next time a
purchase is made.

Then again, if you really want the PSU fan reading, you know
how to do it, just wire whole fan out to motherboard, or use
3rd party fan since it probably has insufficient leads and
would be too loud at full speed (but then, perhaps you like
full-speed fans, it makes a lot more sense to have the PSU
exhaust fan loud/hi-flow than CPU fan, if cooling is REALLY
that important to you).