P4C800-D Processor Failure (Correction)

[Gert B. Frob]

Reposted to correct MB model number:

A quick power failure (flicker?) ruined my CPU. The power went out for just
a second or two and, even though the BIOS is set to remain off, the machine
went into reboot with the nice lady saying the CPU had failed. Sure enough
it had.

I've never seen this before and don't want to repeat it. Anybody
experienced this problem? Any ideas?

Specifics:

ASUS P4C800-D
Intel Northwood P-4 3.2
Antec Power (450w, if I recall)
Surge protection

BTW, our three other machines went through the same incident without a
problem.

 

[KC Computers]

A quick power failure (flicker?) ruined my CPU. The power went out for

just
a second or two and, even though the BIOS is set to remain off, the
machine
went into reboot with the nice lady saying the CPU had failed. Sure
enough
it had. I've never seen this before and don't want to repeat it. Anybody
experienced this problem? Any ideas?
ASUS P4C800-D
Intel Northwood P-4 3.2
Antec Power (450w, if I recall)
Surge protection

Are you sure that the P4 CPU is indeed bad? We are a dealer
and have found that it is very rare for a CPU (especially an Intel one) to
fail like that. Most times it's the motherboard, etc which gets fried. Did
you try the
CPU in another system?

 

[Paul]

Are you sure that the P4 CPU is indeed bad? We are a dealer
and have found that it is very rare for a CPU (especially an Intel one) to
fail like that. Most times it's the motherboard, etc which gets fried. Did
you try the
CPU in another system?

Many multiple supply silicon chips have a power sequence
requirement. Certain of the rails must never rise above
certain other rails. In some cases, a failure can be instantaneous
and in other cases, failure to meet the power sequence
results in degraded "100K power on hours" reliability.

The Vcore circuit has sufficient current output capabilities,
that no other circuit can argue with it. If the Vcore goes
crazy even for a second, that might be enough time for a
processor to get fried.

While I cannot imagine Asus getting a design detail like that
wrong (after all, there is a publically available reference
schematic for how to design a 875 based board), I suppose
anything is possible. I would place your three systems on
a UPS, so the power can no longer glitch. I use a UPS here,
and it is reasonably cheap insurance. I haven't lost a session
for as long as I've owned my UPS.

Paul

 

[Wayne Youngman]

Hi,

I have an ASUS P4C800-E deluxe and it did the same thing a few times to me
(P4-2.6c). After much head scratching I found out the problem went away if
I unplugged my case fans from the mobo fan-header and instead connected them
directly to my PSU?

Been working fine now for 5 months!

Do you have any case fans attached to the mobo?

Wayne ][

 

[Gert B. Frob]

Are you sure that the P4 CPU is indeed bad? We are a dealer
and have found that it is very rare for a CPU (especially an Intel one) to
fail like that. Most times it's the motherboard, etc which gets fried. Did
you try the
CPU in another system?

I did not try the CPU in another system. However, I did place an older P4
1.8 in the system in question and it ran just fine. After an RMA drill, it
is now, once again, sporting a Northwood 3.2.

I agree that CPU failure is very rare. That's what bothers me about this
situation. The board in question ran just fine with a Northwood 2.8 since
August of 2004. About a month ago I obtained the failed 3.2 and swapped the
2.8 to another machine. The 2.8 is still humming along on a P4P800-D which
also rebooted during the event in question. These power drops occur here
all too frequently. It is windy, sometimes rainy and we enjoy arial power
lines.

When the board was fitted with the 2.8, it had PC2100 memory (2X512). It
now sports PC3200 (2X512) so it now runs full speed (but no overclock). The
P3's in my sons' machines (0.8 and 1.0) have been running now for years.
Same for the Northwood P4's (1.8 and 2.8).

I am wondering if the higher clock speed along with running at top FSB
speeds makes these 3+ units more suseptible to voltage problems.

BTW, in a long business and personal history spanning back to the 8086, I
personally have encountered only this failure. Interestingly, the only
other one I've seen was my nephew's failed Prescot 3.2 running at full bus
speed on a P4P800E-D.

 

[Gert B. Frob]

Many multiple supply silicon chips have a power sequence
requirement. Certain of the rails must never rise above
certain other rails. In some cases, a failure can be instantaneous
and in other cases, failure to meet the power sequence
results in degraded "100K power on hours" reliability.

The Vcore circuit has sufficient current output capabilities,
that no other circuit can argue with it. If the Vcore goes
crazy even for a second, that might be enough time for a
processor to get fried.

While I cannot imagine Asus getting a design detail like that
wrong (after all, there is a publically available reference
schematic for how to design a 875 based board), I suppose
anything is possible. I would place your three systems on
a UPS, so the power can no longer glitch. I use a UPS here,
and it is reasonably cheap insurance. I haven't lost a session
for as long as I've owned my UPS.

Paul

I'm thinking a UPS may be necessary here as well. What UPS are you using?

 

[Gert B. Frob]

Hi,

I have an ASUS P4C800-E deluxe and it did the same thing a few times to me
(P4-2.6c). After much head scratching I found out the problem went away if
I unplugged my case fans from the mobo fan-header and instead connected them
directly to my PSU?

Been working fine now for 5 months!

Do you have any case fans attached to the mobo?

Wayne ][

I think my front fan is connected to a board header. Easy enough to remedy
if it helps. Have you had any power events during your trouble free 5
months? What made you think of this in the first place (just curious)?

 

[Wayne Youngman]

I think my front fan is connected to a board header. Easy enough to
remedy
if it helps. Have you had any power events during your trouble free 5
months? What made you think of this in the first place (just curious)?

Hi,

I bought the ASUS mobo 2nd-Hand Bare-bone stylee, but when I built a system
around it I have a few weird errors like what you described. Luckily I have
a few systems so I was able to swap-out and test components.

It was when I stripped the system down to bare essentials and it worked,
that I realised what was going on.

Wayne ][

 

[Paul]

I'm thinking a UPS may be necessary here as well. What UPS are you using?

APC BackUPS 650 - the biggest thing I could get at my local
computer store. I wouldn't want to order one over the Internet,
due to shipping costs. I typically run two computers and one
monitor on it.

Paul

 

[Gert B. Frob]

Reposted to correct MB model number:

A quick power failure (flicker?) ruined my CPU. The power went out for just
a second or two and, even though the BIOS is set to remain off, the machine
went into reboot with the nice lady saying the CPU had failed. Sure enough
it had.

I've never seen this before and don't want to repeat it. Anybody
experienced this problem? Any ideas?

Specifics:

ASUS P4C800-D
Intel Northwood P-4 3.2
Antec Power (450w, if I recall)
Surge protection

BTW, our three other machines went through the same incident without a
problem.

Well, I decided to go with a Cyberpower 1500AVR UPS. Guess that ought to
fix it.