P4C800 Deluxe Woes

Bubba,

Was it giving you a "system failed CPU test message?"

On Wed, 16 Jan 2008 17:02:05 -0500, "Bubba" <bfunke.RemoveThis@nospam.jcs.mil>
wrote:

>I was having issues with my P4C800 DLX not booting. It turned out the PS
>was bad. I purchased an Antec Smart Power 400 and all is now fine. May try
>that and reseating all your connectors.
>
>Bubba
>
 >> Stay informed about: P4C800 Deluxe Woes 

Larry Johnson wrote:
> Hey guys,
>
> I have a P4C800 Deluxe that was giving me the messages "computer
> falied CPU test" and/or "computer falied CPU test due to
> overclocking." However, the computer would boot after I hit the reset
> button sometimes once, sometimes twice, and sometimes more.
>
> I eventually got tired of having to do this. So, I searched the
> archives and found a message stating this problem was caused by a
> solder overrun under the CPU mounting. So I took the board apart, and
> did find that one of the solder mounds under the CPU mounting was
> touching one of the green circuit lines.
>
> I then took an exacto knife and carefully -- I thought -- cut and
> scraped the overrun off. However, after putting the system back
> together again, I get the same message, but the system won't reboot no
> matter how many times I try.
>
> If anybody has any ideas as to how I can possibly save this board, I'd
> greatly appreciate hearing them.
>
> Thanks very much in advance.
>
> Larry

The "solder blob" problem only affected the first few thousand P4C800
motherboards at launch. It looks to be caused by assembling the
CPU heatsink framework, before running the board through the solder
machine.

If you cut away the solder with a knife, check for accidental nicks on
adjacent tracks. If you truly only cut the solder joint in question,
you can always touch it up with a soldering iron. Be aware that
something with a large mass to it (not one of the tiny tracks),
will need an iron with a higher power rating. That will get the joint
hot in a shorter time. For example, I own 15W, 25W, and an 80W iron,
and if this is a large blob of solder, I'd use the 80W to repair it.

You have to be more careful, with things embedded in plastic. For
example, I'd think twice before reheating a pin on a CPU socket, because
it could end up distorting the position of the contact.

When soldering, you want to avoid "joining" two blobs by accident.
If there are two mounds of solder next to one another, it is easy to
bridge them. There are tools, like "solder wick" and "solder sucker",
that can remove excessive amounts of molten solder. Then fresh solder
can be used, to make a proper joint for each mound (without bridging
them). PCBs have "solder mask", which is a finish applied where solder
is not wanted, and solder mask helps prevent bridging.

Solder sucker - spring loaded plunger and trigger creates a temporary vacuum.
Used for removing solder from holes, such as when pulling bad capacitors.

http://www1.istockphoto.com/file_thumbview_approve/550953/2/istockphot...50953_s

Solder wick - for minor touchup work - usability of this stuff varies with gauge.
I use very fine stuff - for me the thick stuff (more available) is a waste
of money. This stuff pictured here is 0.075".

http://www.seattlerobotics.org/encoder/200002/mvc-010f.jpg

I have a couple rolls of #423 in my tool box right now, which is 0.025".
I might use this for cleanup of surface mount pads and the like.

http://www.mgchemicals.com/products/400braid.html

Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Thanks for the detailed response, Paul. I did go back and resolder the
mound, which allowed me to boot up. However, I still get the "system
failed CPU test message," and have to restart the sytem until it
finally boots. So I'm basically back where I started. This leads me to
believe that the problem was in that joint, or that something I did
when I took the board out and put it back in changed its behavior.

In either case, this board has problems. What do you think about
Bubba's suggestion that the power supply might be the problem.

Thanks.

Bill



On Thu, 17 Jan 2008 00:15:48 -0500, Paul <nospam.DeleteThis@needed.com> wrote:

>Larry Johnson wrote:
>> Hey guys,
>>
>> I have a P4C800 Deluxe that was giving me the messages "computer
>> falied CPU test" and/or "computer falied CPU test due to
>> overclocking." However, the computer would boot after I hit the reset
>> button sometimes once, sometimes twice, and sometimes more.
>>
>> I eventually got tired of having to do this. So, I searched the
>> archives and found a message stating this problem was caused by a
>> solder overrun under the CPU mounting. So I took the board apart, and
>> did find that one of the solder mounds under the CPU mounting was
>> touching one of the green circuit lines.
>>
>> I then took an exacto knife and carefully -- I thought -- cut and
>> scraped the overrun off. However, after putting the system back
>> together again, I get the same message, but the system won't reboot no
>> matter how many times I try.
>>
>> If anybody has any ideas as to how I can possibly save this board, I'd
>> greatly appreciate hearing them.
>>
>> Thanks very much in advance.
>>
>> Larry
>
>The "solder blob" problem only affected the first few thousand P4C800
>motherboards at launch. It looks to be caused by assembling the
>CPU heatsink framework, before running the board through the solder
>machine.
>
>If you cut away the solder with a knife, check for accidental nicks on
>adjacent tracks. If you truly only cut the solder joint in question,
>you can always touch it up with a soldering iron. Be aware that
>something with a large mass to it (not one of the tiny tracks),
>will need an iron with a higher power rating. That will get the joint
>hot in a shorter time. For example, I own 15W, 25W, and an 80W iron,
>and if this is a large blob of solder, I'd use the 80W to repair it.
>
>You have to be more careful, with things embedded in plastic. For
>example, I'd think twice before reheating a pin on a CPU socket, because
>it could end up distorting the position of the contact.
>
>When soldering, you want to avoid "joining" two blobs by accident.
>If there are two mounds of solder next to one another, it is easy to
>bridge them. There are tools, like "solder wick" and "solder sucker",
>that can remove excessive amounts of molten solder. Then fresh solder
>can be used, to make a proper joint for each mound (without bridging
>them). PCBs have "solder mask", which is a finish applied where solder
>is not wanted, and solder mask helps prevent bridging.
>
>Solder sucker - spring loaded plunger and trigger creates a temporary vacuum.
>Used for removing solder from holes, such as when pulling bad capacitors.
>
>http://www1.istockphoto.com/file_thumbview_approve/550953/2/istockphoto_550953_solder_sucker.jpg
>
>Solder wick - for minor touchup work - usability of this stuff varies with gauge.
>I use very fine stuff - for me the thick stuff (more available) is a waste
>of money. This stuff pictured here is 0.075".
>
>http://www.seattlerobotics.org/encoder/200002/mvc-010f.jpg
>
>I have a couple rolls of #423 in my tool box right now, which is 0.025".
>I might use this for cleanup of surface mount pads and the like.
>
>http://www.mgchemicals.com/products/400braid.html
>
> Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Bill Schmidt wrote:
> Thanks for the detailed response, Paul. I did go back and resolder the
> mound, which allowed me to boot up. However, I still get the "system
> failed CPU test message," and have to restart the sytem until it
> finally boots. So I'm basically back where I started. This leads me to
> believe that the problem was in that joint, or that something I did
> when I took the board out and put it back in changed its behavior.
>
> In either case, this board has problems. What do you think about
> Bubba's suggestion that the power supply might be the problem.
>
> Thanks.
>
> Bill

The Vocal POST messages come from a Winbond chip. It is tied into
one of the audio channels on the built-in sound.

The chip is programmable, and both the program it runs, and the
audio messages it plays, are stored in the same EEPROM. (I think
it is an 8 pin DIP with a serial interface.) The Winbond is
autonomous, and can run its program, even if the CPU socket
is not occupied.

The chip has a couple methods to trigger playback. One is
the use of a timer. For "System failed CPU test", the timer
is set by the Winbond code. If the main processor and its
BIOS code, do not "reach down" and clear the timer, then the
Winbond message gets played during POST. Anything which
prevents the early BIOS code from executing, could cause
it. Including the Vcore regulator for the processor shutting
down (that can sometimes be caused by a power supply that
starts delivering power slowly).

A similar thing happens during RAM test. If the main CPU goes
crazy, while the POST RAM test is running, the timer is again
used to trigger the appropriate message about a failed RAM
test.

For the rest of the message types, they are under the control
of the main processor and BIOS. The processor writes the desired
message number into the Winbond chip, and the Winbond chip then
plays the matching message.

For example, a flag of some kind is maintained by the BIOS.
If the motherboard does an orderly shutdown, the BIOS has a
chance to clear the flag. If the motherboard does a disorderly
shutdown (where the BIOS has no chance to do anything), the
flag used remains set. On the *next* POST, the BIOS notices
that the previous shutdown was not clean. The BIOS then writes
the desired message number into the Winbond chip. The Winbond
chip plays back something like "System failed due to CPU overclock".
It doesn't actually mean that an overclock was involved - merely
that since the last time the system was running, the shutdown was
not clean. I don't know what is used for the flag (I'd have to be
able to decipher BIOS code to figure that out).

I learned most of this, by downloading the Winbond Editor package
(there were a couple versions on the Asus site). The source
code for the Winbond chip, is in the install folder for that
tool. Between the Winbond datasheet, and the source code, I
pieced together the above picture.

It could be the motherboard that is at fault. And there is a
slim chance it could be the power supply. But for the cases
I can remember for power supply, the victims were never able
to get it to post. So it wasn't flaky. Which leaves the
motherboard as a potential cause. Processors don't fail too
often, which is why I wouldn't immediately assume that as
a cause. And testing the processor on another board, would
tell you soon enough, whether that was the problem.

I think the surprising thing about motherboards, is how often
the problems are able to be figured out. There should be
a lot more "great unknown" type failures, due to the
complexity of the hardware and interconnect wiring.

Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Paul,

Thanks again for the response: I'm amazed at how much you know about
this.

I, too, really doubted whether it's a processor problem, since the
board runs fine once it has booted, and I'll defer to your doubts
about it being caused by the PS.

At this point, then, is there anything I can do to try to resolve the
problem, short of getting a new board. And I don't suppose that Abit
is going to do anything for me, or will even acknowledge that the
board was defective?

Regards,

Bill



On Thu, 17 Jan 2008 10:01:01 -0500, Paul <nospam.DeleteThis@needed.com> wrote:

>Bill Schmidt wrote:
>> Thanks for the detailed response, Paul. I did go back and resolder the
>> mound, which allowed me to boot up. However, I still get the "system
>> failed CPU test message," and have to restart the sytem until it
>> finally boots. So I'm basically back where I started. This leads me to
>> believe that the problem was in that joint, or that something I did
>> when I took the board out and put it back in changed its behavior.
>>
>> In either case, this board has problems. What do you think about
>> Bubba's suggestion that the power supply might be the problem.
>>
>> Thanks.
>>
>> Bill
>
>The Vocal POST messages come from a Winbond chip. It is tied into
>one of the audio channels on the built-in sound.
>
>The chip is programmable, and both the program it runs, and the
>audio messages it plays, are stored in the same EEPROM. (I think
>it is an 8 pin DIP with a serial interface.) The Winbond is
>autonomous, and can run its program, even if the CPU socket
>is not occupied.
>
>The chip has a couple methods to trigger playback. One is
>the use of a timer. For "System failed CPU test", the timer
>is set by the Winbond code. If the main processor and its
>BIOS code, do not "reach down" and clear the timer, then the
>Winbond message gets played during POST. Anything which
>prevents the early BIOS code from executing, could cause
>it. Including the Vcore regulator for the processor shutting
>down (that can sometimes be caused by a power supply that
>starts delivering power slowly).
>
>A similar thing happens during RAM test. If the main CPU goes
>crazy, while the POST RAM test is running, the timer is again
>used to trigger the appropriate message about a failed RAM
>test.
>
>For the rest of the message types, they are under the control
>of the main processor and BIOS. The processor writes the desired
>message number into the Winbond chip, and the Winbond chip then
>plays the matching message.
>
>For example, a flag of some kind is maintained by the BIOS.
>If the motherboard does an orderly shutdown, the BIOS has a
>chance to clear the flag. If the motherboard does a disorderly
>shutdown (where the BIOS has no chance to do anything), the
>flag used remains set. On the *next* POST, the BIOS notices
>that the previous shutdown was not clean. The BIOS then writes
>the desired message number into the Winbond chip. The Winbond
>chip plays back something like "System failed due to CPU overclock".
>It doesn't actually mean that an overclock was involved - merely
>that since the last time the system was running, the shutdown was
>not clean. I don't know what is used for the flag (I'd have to be
>able to decipher BIOS code to figure that out).
>
>I learned most of this, by downloading the Winbond Editor package
>(there were a couple versions on the Asus site). The source
>code for the Winbond chip, is in the install folder for that
>tool. Between the Winbond datasheet, and the source code, I
>pieced together the above picture.
>
>It could be the motherboard that is at fault. And there is a
>slim chance it could be the power supply. But for the cases
>I can remember for power supply, the victims were never able
>to get it to post. So it wasn't flaky. Which leaves the
>motherboard as a potential cause. Processors don't fail too
>often, which is why I wouldn't immediately assume that as
>a cause. And testing the processor on another board, would
>tell you soon enough, whether that was the problem.
>
>I think the surprising thing about motherboards, is how often
>the problems are able to be figured out. There should be
>a lot more "great unknown" type failures, due to the
>complexity of the hardware and interconnect wiring.
>
> Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Bill Schmidt wrote:
> Paul,
>
> Thanks again for the response: I'm amazed at how much you know about
> this.
>
> I, too, really doubted whether it's a processor problem, since the
> board runs fine once it has booted, and I'll defer to your doubts
> about it being caused by the PS.
>
> At this point, then, is there anything I can do to try to resolve the
> problem, short of getting a new board. And I don't suppose that Abit
> is going to do anything for me, or will even acknowledge that the
> board was defective?
>
> Regards,
>
> Bill

The problem sounds like a power/timing problem, during the initial
startup of the board. If it was something digital (processor crash
during startup), there might be more issues with it failing while
it is running later. But so far, all your listed symptoms seem
to happen right at startup. I can't think of too many things you
can do to change that. (Maybe try pulling the RAM, and look for
different symptoms, but I wouldn't be holding my breath. I expect
it will fail to start, just as many times as before.)

The "power train" involves lots of little bits of analog, and no
two motherboard manufacturers do things the same way. (I'm referring
to companies that actually design motherboards, not EVGA.) For example,
if you compare the Intel reference schematic for 875P, to my
Asus P4C800-E board, they aren't even close to doing power the
same way. Intel has plenty of "backfeed cut" circuits, and
Asus by comparison, has nothing that I could see. So looking
at the Intel schematic, and trying to find something similar
on an Asus board, isn't going to work. I can't even begin
to guess what things Asus might have tied into their power_good
logic. The digital connections (like Northbridge to processor
or Northbridge to RAM) would be much more predictable.

Paul
 >> Stay informed about: P4C800 Deluxe Woes 

In article <u4tuo3lqnoinr3mm1ghbge26bduj351vsi.DeleteThis@4ax.com>, Bill Schmidt <schmitty.DeleteThis@comcast.net> wrote:
>Paul,
>
>Thanks again for the response: I'm amazed at how much you know about
>this.
>
>I, too, really doubted whether it's a processor problem, since the
>board runs fine once it has booted, and I'll defer to your doubts
>about it being caused by the PS.
>
>At this point, then, is there anything I can do to try to resolve the
>problem, short of getting a new board. And I don't suppose that Abit
>is going to do anything for me, or will even acknowledge that the
>board was defective?
>
>Regards,
>
>Bill
>

I doubt abit would, its an ASUS board (Evil grin)

>
>
>On Thu, 17 Jan 2008 10:01:01 -0500, Paul <nospam.DeleteThis@needed.com> wrote:
>
>>Bill Schmidt wrote:
>>> Thanks for the detailed response, Paul. I did go back and resolder the
>>> mound, which allowed me to boot up. However, I still get the "system
>>> failed CPU test message," and have to restart the sytem until it
>>> finally boots. So I'm basically back where I started. This leads me to
>>> believe that the problem was in that joint, or that something I did
>>> when I took the board out and put it back in changed its behavior.
>>>
>>> In either case, this board has problems. What do you think about
>>> Bubba's suggestion that the power supply might be the problem.
>>>
>>> Thanks.
>>>
>>> Bill
>>
>>The Vocal POST messages come from a Winbond chip. It is tied into
>>one of the audio channels on the built-in sound.
>>
>>The chip is programmable, and both the program it runs, and the
>>audio messages it plays, are stored in the same EEPROM. (I think
>>it is an 8 pin DIP with a serial interface.) The Winbond is
>>autonomous, and can run its program, even if the CPU socket
>>is not occupied.
>>
>>The chip has a couple methods to trigger playback. One is
>>the use of a timer. For "System failed CPU test", the timer
>>is set by the Winbond code. If the main processor and its
>>BIOS code, do not "reach down" and clear the timer, then the
>>Winbond message gets played during POST. Anything which
>>prevents the early BIOS code from executing, could cause
>>it. Including the Vcore regulator for the processor shutting
>>down (that can sometimes be caused by a power supply that
>>starts delivering power slowly).
>>
>>A similar thing happens during RAM test. If the main CPU goes
>>crazy, while the POST RAM test is running, the timer is again
>>used to trigger the appropriate message about a failed RAM
>>test.
>>
>>For the rest of the message types, they are under the control
>>of the main processor and BIOS. The processor writes the desired
>>message number into the Winbond chip, and the Winbond chip then
>>plays the matching message.
>>
>>For example, a flag of some kind is maintained by the BIOS.
>>If the motherboard does an orderly shutdown, the BIOS has a
>>chance to clear the flag. If the motherboard does a disorderly
>>shutdown (where the BIOS has no chance to do anything), the
>>flag used remains set. On the *next* POST, the BIOS notices
>>that the previous shutdown was not clean. The BIOS then writes
>>the desired message number into the Winbond chip. The Winbond
>>chip plays back something like "System failed due to CPU overclock".
>>It doesn't actually mean that an overclock was involved - merely
>>that since the last time the system was running, the shutdown was
>>not clean. I don't know what is used for the flag (I'd have to be
>>able to decipher BIOS code to figure that out).
>>
>>I learned most of this, by downloading the Winbond Editor package
>>(there were a couple versions on the Asus site). The source
>>code for the Winbond chip, is in the install folder for that
>>tool. Between the Winbond datasheet, and the source code, I
>>pieced together the above picture.
>>
>>It could be the motherboard that is at fault. And there is a
>>slim chance it could be the power supply. But for the cases
>>I can remember for power supply, the victims were never able
>>to get it to post. So it wasn't flaky. Which leaves the
>>motherboard as a potential cause. Processors don't fail too
>>often, which is why I wouldn't immediately assume that as
>>a cause. And testing the processor on another board, would
>>tell you soon enough, whether that was the problem.
>>
>>I think the surprising thing about motherboards, is how often
>>the problems are able to be figured out. There should be
>>a lot more "great unknown" type failures, due to the
>>complexity of the hardware and interconnect wiring.
>>
>> Paul
>
 >> Stay informed about: P4C800 Deluxe Woes 

Paul,

Could this timing problem also be related to the system's recent
inability to maintain the correct time and date; i.e., everytime I
boot up, the clock resets to 12 AM, Jan. 2002, even though I recently
changed the CMOS batttery.

In fact, this morning when I booted up, the system spit out a host of
messages, including failure to detect any HD's, boot devices, CMOS,
etc. Then, when I rebooted, it allowed me to load BIOS defaults, and
went right into Xp.

So, does this means that I should scrap the board and, if I do, what
would you recommend as a replacement. I see that these boards are
going for about 150 on Ebay, but I'd hate to have the same problem
with a new board.

OTOH, I really like having the on-board RAID. ANd I would like the
ability to just move my Socket 478 P4, hardware and SATA drives into a
new board with the least amount of compatiblity issues.

Any suggestions?

Thanks






On Thu, 17 Jan 2008 12:16:58 -0500, Paul <nospam RemoveThis @needed.com> wrote:

>Bill Schmidt wrote:
>> Paul,
>>
>> Thanks again for the response: I'm amazed at how much you know about
>> this.
>>
>> I, too, really doubted whether it's a processor problem, since the
>> board runs fine once it has booted, and I'll defer to your doubts
>> about it being caused by the PS.
>>
>> At this point, then, is there anything I can do to try to resolve the
>> problem, short of getting a new board. And I don't suppose that Abit
>> is going to do anything for me, or will even acknowledge that the
>> board was defective?
>>
>> Regards,
>>
>> Bill
>
>The problem sounds like a power/timing problem, during the initial
>startup of the board. If it was something digital (processor crash
>during startup), there might be more issues with it failing while
>it is running later. But so far, all your listed symptoms seem
>to happen right at startup. I can't think of too many things you
>can do to change that. (Maybe try pulling the RAM, and look for
>different symptoms, but I wouldn't be holding my breath. I expect
>it will fail to start, just as many times as before.)
>
>The "power train" involves lots of little bits of analog, and no
>two motherboard manufacturers do things the same way. (I'm referring
>to companies that actually design motherboards, not EVGA.) For example,
>if you compare the Intel reference schematic for 875P, to my
>Asus P4C800-E board, they aren't even close to doing power the
>same way. Intel has plenty of "backfeed cut" circuits, and
>Asus by comparison, has nothing that I could see. So looking
>at the Intel schematic, and trying to find something similar
>on an Asus board, isn't going to work. I can't even begin
>to guess what things Asus might have tied into their power_good
>logic. The digital connections (like Northbridge to processor
>or Northbridge to RAM) would be much more predictable.
>
> Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Pau,
l
How about the P4P800E- Deluxe?
 >> Stay informed about: P4C800 Deluxe Woes 

Bill Schmidt wrote:
> Pau,
> l
> How about the P4P800E- Deluxe?
>

If you aren't attempting a huge overclock, that would be
a good substitute.

Some 865PE boards get AGP (video) artifacts, if you raise
the FSB high enough. Having a high FSB capability via my
P4C800-E didn't make any difference in my case, because
my CPU wouldn't let me overclock much at all

So sure, if they aren't charging an arm and leg for it,
that might make a good substitute.

Remember that all of our boards, have the issue with
"ICH5/ICH5R latchup". Some Southbridge chips have been
destroyed by what I suspect is ESD. It is always a
nagging concern for these motherboards, and one of the
reasons I don't (regularly) use the USB ports on my
motherboard. To protect the ICH5, if I wanted to do a
lot of USB work, I'd probably go for a separate USB2
PCI card. That is the best protection I know of,
against "latchup" failure.

Latchup failure manifests two ways. In a "minor" failure,
the USB ports lose power at the PHY level. Device Manager
still shows all the normal entries, but you can no longer
detect any USB devices when plugged into motherboard
USB ports.

In a major failure, the Southbridge gets so hot, there is
a burn mark in the center of the chip. The motherboard
will no longer POST after that. (In the "minor" failure
case, you can still use the motherboard, sans USB ports.)
It is unclear to me, whether Asus would consider warranty
repair after the normal 3 year warranty, for what is
obviously a design defect. (Lots of other Southbridges
don't blow up when you use the USB ports...)

With regard to the CMOS, the inability to remember
settings, isn't going to make startup easy for the
board. Have you ever tried clearing the CMOS, while
the computer was still plugged in ? On some motherboards,
there is an ORing diode, that joins the CMOS battery
source with a +5VSB based source. If you clear CMOS
while +5VSB is still running, the little dual diode
can get damaged. That is one reason why Asus boards
have the green LED - if the LED is lit, you should
not use the clear CMOS jumper. I'm not aware of
any other good mechanisms for the CMOS RAM to fail.
The ORing diode is the best one for that.

If you go to page 18 here, you can see the diode scheme
in the lower left corner. Now, some Asus boards, use
a three pin (looks like a transistor) dual-diode device.
D2 and D3 would be in the same tiny black three pinner,
and the CMOS jumper isn't wired the same way as shown
in the schematic either. What typically would fail, is
D2 would get burned. Since D3 is sitting right next to
it, sometimes it gets ruined too, and then there is no
source of power for the CMOS. I've helped at least one
poster here, to repair one of those - he soldered two
single diodes in place of the burned three pinner. This
is one of the reasons, that every time someone wants
to know how to clear CMOS, I tell them to unplug
No matter how the circuit actually works on their
board... Safety first.

http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF

Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Thanks again for the info. You're either up early or late. In my case
it's early, since I'm in Philly.

I've been playing with the system, but still can't seem to get it
fully operational. Sometimes it will give me the following messages:

CMOS settings wrong!
CMOS Date.Time Not Set!
Overclocking Failed!
No Hard Drive Detected!

However, when I restart it, it will scan my IDE drives and identify
the mirror as FUNCTIONAL. It will then start to boot into XP, but will
hang at some point in the process.

I tried resetting the clock in the BIOS, but the settings won't hold.
I also noticed a high pictched squeal coming from what appears to be
the power supply, though my booting problem existed long before the
squeal started. So, could it be a PS problem after all?

I looked on EBAY, and found that new P4P800-E Deluxe's are going for a
little less than the P4C's, but used one's can be had for under 100.
Howver, considering that I've owned a number of problem-free boards
from Abit, ECS, and Gigbyte in my life, it's kiind of hard to justify
buying another ASUS, especially given the USB problems you mentioned.
(In fact, I can't understand how ASUS can even stay in business after
putting out so many problematic boards.)

Bill

BTW, since you said "our boards," I have to assume that you work for
ASUS?


On Fri, 18 Jan 2008 03:43:01 -0500, Paul <nospam DeleteThis @needed.com> wrote:

>Bill Schmidt wrote:
>> Pau,
>> l
>> How about the P4P800E- Deluxe?
>>
>
>If you aren't attempting a huge overclock, that would be
>a good substitute.
>
>Some 865PE boards get AGP (video) artifacts, if you raise
>the FSB high enough. Having a high FSB capability via my
>P4C800-E didn't make any difference in my case, because
>my CPU wouldn't let me overclock much at all
>
>So sure, if they aren't charging an arm and leg for it,
>that might make a good substitute.
>
>Remember that all of our boards, have the issue with
>"ICH5/ICH5R latchup". Some Southbridge chips have been
>destroyed by what I suspect is ESD. It is always a
>nagging concern for these motherboards, and one of the
>reasons I don't (regularly) use the USB ports on my
>motherboard. To protect the ICH5, if I wanted to do a
>lot of USB work, I'd probably go for a separate USB2
>PCI card. That is the best protection I know of,
>against "latchup" failure.
>
>Latchup failure manifests two ways. In a "minor" failure,
>the USB ports lose power at the PHY level. Device Manager
>still shows all the normal entries, but you can no longer
>detect any USB devices when plugged into motherboard
>USB ports.
>
>In a major failure, the Southbridge gets so hot, there is
>a burn mark in the center of the chip. The motherboard
>will no longer POST after that. (In the "minor" failure
>case, you can still use the motherboard, sans USB ports.)
>It is unclear to me, whether Asus would consider warranty
>repair after the normal 3 year warranty, for what is
>obviously a design defect. (Lots of other Southbridges
>don't blow up when you use the USB ports...)
>
>With regard to the CMOS, the inability to remember
>settings, isn't going to make startup easy for the
>board. Have you ever tried clearing the CMOS, while
>the computer was still plugged in ? On some motherboards,
>there is an ORing diode, that joins the CMOS battery
>source with a +5VSB based source. If you clear CMOS
>while +5VSB is still running, the little dual diode
>can get damaged. That is one reason why Asus boards
>have the green LED - if the LED is lit, you should
>not use the clear CMOS jumper. I'm not aware of
>any other good mechanisms for the CMOS RAM to fail.
>The ORing diode is the best one for that.
>
>If you go to page 18 here, you can see the diode scheme
>in the lower left corner. Now, some Asus boards, use
>a three pin (looks like a transistor) dual-diode device.
>D2 and D3 would be in the same tiny black three pinner,
>and the CMOS jumper isn't wired the same way as shown
>in the schematic either. What typically would fail, is
>D2 would get burned. Since D3 is sitting right next to
>it, sometimes it gets ruined too, and then there is no
>source of power for the CMOS. I've helped at least one
>poster here, to repair one of those - he soldered two
>single diodes in place of the burned three pinner. This
>is one of the reasons, that every time someone wants
>to know how to clear CMOS, I tell them to unplug
>No matter how the circuit actually works on their
>board... Safety first.
>
>http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF
>
> Paul
 >> Stay informed about: P4C800 Deluxe Woes 

Bill Schmidt wrote:
> Thanks again for the info. You're either up early or late. In my case
> it's early, since I'm in Philly.
>
> I've been playing with the system, but still can't seem to get it
> fully operational. Sometimes it will give me the following messages:
>
> CMOS settings wrong!
> CMOS Date.Time Not Set!
> Overclocking Failed!
> No Hard Drive Detected!
>
> However, when I restart it, it will scan my IDE drives and identify
> the mirror as FUNCTIONAL. It will then start to boot into XP, but will
> hang at some point in the process.
>
> I tried resetting the clock in the BIOS, but the settings won't hold.
> I also noticed a high pictched squeal coming from what appears to be
> the power supply, though my booting problem existed long before the
> squeal started. So, could it be a PS problem after all?
>

One of our computers here at work had a similar problem where it could
take over a half hour to finally boot after being turned off. It worked
fine once it booted but it was a real nuisance after a power failure. I
finally replaced the power supply with a high end one and have not had a
problem with that computer in over 5 years.


> I looked on EBAY, and found that new P4P800-E Deluxe's are going for a
> little less than the P4C's, but used one's can be had for under 100.
> Howver, considering that I've owned a number of problem-free boards
> from Abit, ECS, and Gigbyte in my life, it's kiind of hard to justify
> buying another ASUS, especially given the USB problems you mentioned.
> (In fact, I can't understand how ASUS can even stay in business after
> putting out so many problematic boards.)
>
> Bill
>
> BTW, since you said "our boards," I have to assume that you work for
> ASUS?
>
>
> On Fri, 18 Jan 2008 03:43:01 -0500, Paul <nospam.DeleteThis@needed.com> wrote:
>
>> Bill Schmidt wrote:
>>> Pau,
>>> l
>>> How about the P4P800E- Deluxe?
>>>
>> If you aren't attempting a huge overclock, that would be
>> a good substitute.
>>
>> Some 865PE boards get AGP (video) artifacts, if you raise
>> the FSB high enough. Having a high FSB capability via my
>> P4C800-E didn't make any difference in my case, because
>> my CPU wouldn't let me overclock much at all
>>
>> So sure, if they aren't charging an arm and leg for it,
>> that might make a good substitute.
>>
>> Remember that all of our boards, have the issue with
>> "ICH5/ICH5R latchup". Some Southbridge chips have been
>> destroyed by what I suspect is ESD. It is always a
>> nagging concern for these motherboards, and one of the
>> reasons I don't (regularly) use the USB ports on my
>> motherboard. To protect the ICH5, if I wanted to do a
>> lot of USB work, I'd probably go for a separate USB2
>> PCI card. That is the best protection I know of,
>> against "latchup" failure.
>>
>> Latchup failure manifests two ways. In a "minor" failure,
>> the USB ports lose power at the PHY level. Device Manager
>> still shows all the normal entries, but you can no longer
>> detect any USB devices when plugged into motherboard
>> USB ports.
>>
>> In a major failure, the Southbridge gets so hot, there is
>> a burn mark in the center of the chip. The motherboard
>> will no longer POST after that. (In the "minor" failure
>> case, you can still use the motherboard, sans USB ports.)
>> It is unclear to me, whether Asus would consider warranty
>> repair after the normal 3 year warranty, for what is
>> obviously a design defect. (Lots of other Southbridges
>> don't blow up when you use the USB ports...)
>>
>> With regard to the CMOS, the inability to remember
>> settings, isn't going to make startup easy for the
>> board. Have you ever tried clearing the CMOS, while
>> the computer was still plugged in ? On some motherboards,
>> there is an ORing diode, that joins the CMOS battery
>> source with a +5VSB based source. If you clear CMOS
>> while +5VSB is still running, the little dual diode
>> can get damaged. That is one reason why Asus boards
>> have the green LED - if the LED is lit, you should
>> not use the clear CMOS jumper. I'm not aware of
>> any other good mechanisms for the CMOS RAM to fail.
>> The ORing diode is the best one for that.
>>
>> If you go to page 18 here, you can see the diode scheme
>> in the lower left corner. Now, some Asus boards, use
>> a three pin (looks like a transistor) dual-diode device.
>> D2 and D3 would be in the same tiny black three pinner,
>> and the CMOS jumper isn't wired the same way as shown
>> in the schematic either. What typically would fail, is
>> D2 would get burned. Since D3 is sitting right next to
>> it, sometimes it gets ruined too, and then there is no
>> source of power for the CMOS. I've helped at least one
>> poster here, to repair one of those - he soldered two
>> single diodes in place of the burned three pinner. This
>> is one of the reasons, that every time someone wants
>> to know how to clear CMOS, I tell them to unplug
>> No matter how the circuit actually works on their
>> board... Safety first.
>>
>> http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF
>>
>> Paul
>
 >> Stay informed about: P4C800 Deluxe Woes 

Thanks for mentioning that, Mike. I was hesitant to throw good money
after bad by buying a new PS, but maybe it's worth a shot.

Bill

On Fri, 18 Jan 2008 09:29:41 GMT, "Michael W. Ryder"
<_mwryder DeleteThis @worldnet.att.net> wrote:

>Bill Schmidt wrote:
>> Thanks again for the info. You're either up early or late. In my case
>> it's early, since I'm in Philly.
>>
>> I've been playing with the system, but still can't seem to get it
>> fully operational. Sometimes it will give me the following messages:
>>
>> CMOS settings wrong!
>> CMOS Date.Time Not Set!
>> Overclocking Failed!
>> No Hard Drive Detected!
>>
>> However, when I restart it, it will scan my IDE drives and identify
>> the mirror as FUNCTIONAL. It will then start to boot into XP, but will
>> hang at some point in the process.
>>
>> I tried resetting the clock in the BIOS, but the settings won't hold.
>> I also noticed a high pictched squeal coming from what appears to be
>> the power supply, though my booting problem existed long before the
>> squeal started. So, could it be a PS problem after all?
>>
>
>One of our computers here at work had a similar problem where it could
>take over a half hour to finally boot after being turned off. It worked
>fine once it booted but it was a real nuisance after a power failure. I
>finally replaced the power supply with a high end one and have not had a
>problem with that computer in over 5 years.
>
>
>> I looked on EBAY, and found that new P4P800-E Deluxe's are going for a
>> little less than the P4C's, but used one's can be had for under 100.
>> Howver, considering that I've owned a number of problem-free boards
>> from Abit, ECS, and Gigbyte in my life, it's kiind of hard to justify
>> buying another ASUS, especially given the USB problems you mentioned.
>> (In fact, I can't understand how ASUS can even stay in business after
>> putting out so many problematic boards.)
>>
>> Bill
>>
>> BTW, since you said "our boards," I have to assume that you work for
>> ASUS?
>>
>>
>> On Fri, 18 Jan 2008 03:43:01 -0500, Paul <nospam DeleteThis @needed.com> wrote:
>>
>>> Bill Schmidt wrote:
>>>> Pau,
>>>> l
>>>> How about the P4P800E- Deluxe?
>>>>
>>> If you aren't attempting a huge overclock, that would be
>>> a good substitute.
>>>
>>> Some 865PE boards get AGP (video) artifacts, if you raise
>>> the FSB high enough. Having a high FSB capability via my
>>> P4C800-E didn't make any difference in my case, because
>>> my CPU wouldn't let me overclock much at all
>>>
>>> So sure, if they aren't charging an arm and leg for it,
>>> that might make a good substitute.
>>>
>>> Remember that all of our boards, have the issue with
>>> "ICH5/ICH5R latchup". Some Southbridge chips have been
>>> destroyed by what I suspect is ESD. It is always a
>>> nagging concern for these motherboards, and one of the
>>> reasons I don't (regularly) use the USB ports on my
>>> motherboard. To protect the ICH5, if I wanted to do a
>>> lot of USB work, I'd probably go for a separate USB2
>>> PCI card. That is the best protection I know of,
>>> against "latchup" failure.
>>>
>>> Latchup failure manifests two ways. In a "minor" failure,
>>> the USB ports lose power at the PHY level. Device Manager
>>> still shows all the normal entries, but you can no longer
>>> detect any USB devices when plugged into motherboard
>>> USB ports.
>>>
>>> In a major failure, the Southbridge gets so hot, there is
>>> a burn mark in the center of the chip. The motherboard
>>> will no longer POST after that. (In the "minor" failure
>>> case, you can still use the motherboard, sans USB ports.)
>>> It is unclear to me, whether Asus would consider warranty
>>> repair after the normal 3 year warranty, for what is
>>> obviously a design defect. (Lots of other Southbridges
>>> don't blow up when you use the USB ports...)
>>>
>>> With regard to the CMOS, the inability to remember
>>> settings, isn't going to make startup easy for the
>>> board. Have you ever tried clearing the CMOS, while
>>> the computer was still plugged in ? On some motherboards,
>>> there is an ORing diode, that joins the CMOS battery
>>> source with a +5VSB based source. If you clear CMOS
>>> while +5VSB is still running, the little dual diode
>>> can get damaged. That is one reason why Asus boards
>>> have the green LED - if the LED is lit, you should
>>> not use the clear CMOS jumper. I'm not aware of
>>> any other good mechanisms for the CMOS RAM to fail.
>>> The ORing diode is the best one for that.
>>>
>>> If you go to page 18 here, you can see the diode scheme
>>> in the lower left corner. Now, some Asus boards, use
>>> a three pin (looks like a transistor) dual-diode device.
>>> D2 and D3 would be in the same tiny black three pinner,
>>> and the CMOS jumper isn't wired the same way as shown
>>> in the schematic either. What typically would fail, is
>>> D2 would get burned. Since D3 is sitting right next to
>>> it, sometimes it gets ruined too, and then there is no
>>> source of power for the CMOS. I've helped at least one
>>> poster here, to repair one of those - he soldered two
>>> single diodes in place of the burned three pinner. This
>>> is one of the reasons, that every time someone wants
>>> to know how to clear CMOS, I tell them to unplug
>>> No matter how the circuit actually works on their
>>> board... Safety first.
>>>
>>> http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF
>>>
>>> Paul
>>
 >> Stay informed about: P4C800 Deluxe Woes