Re: Problem with M2N32 WS Pro
- From: Paul <nospam@xxxxxxxxxx>
- Date: Wed, 18 Jul 2007 04:02:38 -0400
chinaski69@xxxxxxxxx wrote:
I have seen a lot of positive feedback but I have had a lot of
problems with mine. What I have is that motherboard plus the AMD
6000+, 2MB of Corsair RAM, one PCIE video card, one PCIX SCSI
adapter. No other add-in cards.
Lately, what happens is one of two things:
1) It POSTS, then I get a warning in BIOS that I need to enter setup
to check hardware settings, which tells me everything is overvoltage.
For example, 12 is 16. I can boot from there where my Asus Probe
program tells me the CPU is running at something like 213 C and the MB
is at something like 200 C. Obviously crazy numbers.
The CPU fan runs about 3200rpm.
2) I get one beep, fans spin up, then it dies. No video output, no
POST.
What I have done to get around that so far is to reseat the video
adapter and the SCSI adapter. Usually reseating the video adapter
does it. My video adapter is a RADEON X1600 (I think). I've tried an
NVIDIA 7300 (which blue screened XP) and get inconsistent results. I
can't rule in or out the video adapter as the source of the problem.
Seems as likely it's an electrical problem on the board. Help?
I presume you've already checked that the BIOS is up to date. There
is a 6000+ rev F3 listed here, as needing BIOS 1102 or later.
http://support.asus.com.tw/cpusupport/cpu_support_right_master.aspx?type=1&name=M2N32%20WS%20Professional&SLanguage=en-us&cache=1
The hardware monitor chip (Super I/O plus monitor) is an ITE8716F-S,
http://www.motherboards.org/images/reviews/motherboards/1666_p3_8.jpg?
http://www.iteusa.com/product_info/PC/Brief-IT8716_2.asp
http://www.iteusa.com/product_info/file/pc/IT8716F_V0.3.zip (datasheet)
On page 100 of the datasheet, you can see the connection scheme for
voltages. 3.3V would be connected directly to the chip (no divider).
+12V is measured by means of a resistive divider. If "Vref" is bad
(i.e. not at 4.096 volts), the measured value will be thrown off,
as Vref would be used internally by the ADC. A thermistor temperature
sensor (Rt) near the top of the diagram, would not be thrown off,
because that measurement is ratiometric (i.e. the input voltage
tracks what happens to Vref - if Vref is cut in half, the voltage
at the point labeled Tin would also be cut in half). The diode based
temperature measurements would be thrown off by a Vref error, in the
same way that a voltage measurement would. I would say, for the most
part, all temps and voltages would go nuts, if Vref was partially shorted.
Another way for the hardware monitor to fail, would be if data values
on the bus were corrupted coming back from the chip. But if the bus
was messed up, you'd think programming the chip to do anything, would
be pretty difficult. The Super I/O functions wouldn't work right if
that was the case. (Serial port, parallel port, floppy interface etc.)
Unless Vref is defective, I don't see any other easy answers.
For the machine to switch off, the CPU overheat function could do
that. And depending on what chip is being used to handle the
THERMTRIP logic signal from the processor, if that logic chip
were to fail, that could cause the computer to switch off. Since
the thing beeps once, a more remote possibility, is an actual
instruction is being executed, that actuates the power-off
feature. Maybe the BIOS is clever enough, to take the out
of range readings from the BIOS, as evidence of trouble ?
If it was my board, my first test would be the "cardboard test".
Since the board would likely require warranty service (RMA), you'd
be pulling the motherboard anyway. The "cardboard test" involves
running the motherboard outside the computer case. Doing so,
eliminates the possibility of a standoff or screw, shorting
the motherboard. Sometimes, an extra standoff is left behind,
when changing motherboards. Each standoff should only mate with a
screw hole on the motherboard - a metal standoff should not be
allowed to touch the copper traces on the bottom of the board.
The plated hole intended for a standoff, is actually at ground
potential, so they are meant to touch. (On one Asus board, when
customers reported one channel of their audio didn't work, the
Tech Support knew right away, there was a standoff underneath an
audio signal. Removing the standoff fixed it.)
The "cardboard" part of the test, involves making sure the
motherboard is sitting on an insulator. I use a telephone book
with a cardboard cover. (Note - some people will rightly point
out, that this is bad practice from an antistatic point of view.
They could well be right, but the odds that most people know
what materials are right for the job, is also limited. *Don't*
stick tin foil underneath the board. That would be really
bad, and would give an interesting light show and cloud of
smoke. The lesser of these evils, is to use cardboard. And
touch the case of the PSU, before handling or installing stuff
on the motherboard. That will drain at least some static
off your fingers.)
My second test, would involve getting out a multimeter. Checking
all supply voltages, to see if they are within 5% of nominal. I'd
also try checking Vref on the SuperI/O - hardware monitor, but
experience has shown me that to do so is dangerous. I've worked with
those fine pitch chips before, and you can bend the leads by poking
them with a voltmeter. If you're clumsy like me, stuff gets shorted
by doing so. To work on chips with fine pitch leads, you
really need some magnification to see what you're doing.
Good luck,
Paul
.
- References:
- Problem with M2N32 WS Pro
- From: chinaski69
- Problem with M2N32 WS Pro
- Prev by Date: Re: Memory Advice Please - ASUS P5B Deluxe
- Next by Date: Re: Memory Advice Please - ASUS P5B Deluxe
- Previous by thread: Problem with M2N32 WS Pro
- Next by thread: Re: Problem with M2N32 WS Pro
- Index(es):
Relevant Pages
|
