Troubleshooting MB and CPU

Hi,

Is there a way to troubleshoot a motherboard or cpu without using
another motherboard or cpu? (Concern about frying something)

I have another P4800-E Deluxe MB with a P43.0 chip. After about 10
minutes it keeps rebooting like it's overheating. I changed out the
power supply but that didn't fix the problem. Had someone test the
RAM and it's good. The BIOS shows that the cpu fan is running as
expected so the only thing left I can think of is either the board
or chip. Is there something I can do to find out which one? I'm
concerned about putting the chip in another board if that's the
problem or taking a chance on frying another chip if the board is
the problem. Is there a way to test one of them?

Thanks,
Samantha

On Aug 16, 5:06 pm, "Samantha" none wrote:
Is there a way to troubleshoot a motherboard or cpu without using
another motherboard or cpu? (Concern about frying something)

I have another P4800-E Deluxe MB with a P43.0 chip. After about 10
minutes it keeps rebooting like it's overheating. I changed out the
power supply but that didn't fix the problem. Had someone test the
RAM and it's good. The BIOS shows that the cpu fan is running as
expected so the only thing left I can think of is either the board
or chip. Is there something I can do to find out which one? I'm
concerned about putting the chip in another board if that's the
problem or taking a chance on frying another chip if the board is
the problem. Is there a way to test one of them?

Step back a minute. Due to shotgunning, then integrity of the power
supply 'system' is still unkonwn. For example, a swapped in power
supply in one sysetm can appear to be defective in another - because
it was actually marginal in both systems. Another problem. Assumed
is an entire power supply 'system' is only a power supply. Also
wrong. A power supply is only one component of a power supply
'system'.

Without numbers and other data, then better informed posters cannot
answer with a useful reply. That means using a 3.5 digit mulitmeter
(and two minutes) to obtain numbers and establish integrity of the
entire power supply 'system'. In your case, the relevant part of this
procedure is to have computer access all peripherals simultaneously
(multitask) and then take voltage measurements on any one of orange,
purple, red, and yellow wires. Entire procedure (including part
important to you) is "When your computer dies without warning....."
starting 6 Feb 2007 in the newsgroup alt.windows-xp at:
http://tinyurl.com/yvf9vh

If the power 'system' is not confirmed good, then it may cause
everything else to appear defective. That only creates even more
confusion. A computer can work just fine for ten minutes when the
power supply 'system' always has numbers that are defective. Those
numbers are most relevant when load is maximum. Therefore watch a
movie, while reading (or writing to) a CD-Rom, while downloading from
the internet, while reading the disk drive, etc ... all
simultaneously. When doing all this, only then will some defective
power supplies be detected by numbers. Notice all those things are
running, but the power supply is defective - and would cause crashes
later. Post those numbers here (even if good) because numbers may
also contain further useful information.

Ram is tested good. Then the same test is performed again when
computer is in a room of 100 degree F - a perfectly good computer
temperature. Or RAM is heated with a hairdryer on highest heat. Any
Ram that fails a diagnostic test in the computer (must be in your
computer - not elsewhere) is defective or motherboard semiconductors
talking to that memory is defective. If it also passes at elevated
temperatures, then move on to other suspects.

Very few items can shutdown a computer. That short list is memory,
power supply, CPU, video card, sound card, and some motherboard
functions. Everything else including disk drive, keyboard, monitor,
and mouse are not on a usual suspect list. Anything can appear
defective if numbers from a loaded power supply are not obtained using
that multimeter. Not having a meter is not even an option since
meters are sold even to K-mart shoppers.

Swapping parts tells us nothing useful and sometimes makes the
problem exponentially complex. Temperatures from the motherboard
temperature monitor may be useful. But again, to solve a problem, get
numbers and other diagnostic facts. Do not speculate. First identify
a suspect long before trying to fix anything.

Heat is a diagnostic tool. If heat creates problems, then fix the
defective part. Don't cure symptoms by adding fans. Problem is solved
first by breaking the problem down into parts. Then establish each
part as 'known good' or 'known bad'. Currently you don't know of
anything in either state. Everything is still questionable -
unknown. Start by first confirming entire power supply 'system'
integrity. Don't shotgun - swapping parts as if that proves
anything. Demonstrated - a defective (marginal) power supply can
still make computer appear to be working.

maybe it is overheating.........what temps are you showing in the
BIOS??..keep it in the BIOS for 10-15 minutes and watch the temps
maybe your Temp shut dowm limit is too low??...what is it set at??.....in
the BIOS
peter
"Samantha" none wrote in message
...
Hi,

Is there a way to troubleshoot a motherboard or cpu without using
another motherboard or cpu? (Concern about frying something)

I have another P4800-E Deluxe MB with a P43.0 chip. After about 10
minutes it keeps rebooting like it's overheating. I changed out the
power supply but that didn't fix the problem. Had someone test the
RAM and it's good. The BIOS shows that the cpu fan is running as
expected so the only thing left I can think of is either the board
or chip. Is there something I can do to find out which one? I'm
concerned about putting the chip in another board if that's the
problem or taking a chance on frying another chip if the board is
the problem. Is there a way to test one of them?

Thanks,
Samantha

"w_tom"
| On Aug 16, 5:06 pm, "Samantha" none wrote:
| Is there a way to troubleshoot a motherboard or cpu without
using
| another motherboard or cpu? (Concern about frying something)
|
| I have another P4800-E Deluxe MB with a P43.0 chip. After about
10
| minutes it keeps rebooting like it's overheating. I changed out
the
| power supply but that didn't fix the problem. Had someone test
the
| RAM and it's good. The BIOS shows that the cpu fan is running as
| expected so the only thing left I can think of is either the
board
| or chip. Is there something I can do to find out which one? I'm
| concerned about putting the chip in another board if that's the
| problem or taking a chance on frying another chip if the board
is
| the problem. Is there a way to test one of them?
|
| Step back a minute. Due to shotgunning, then integrity of the
power
| supply 'system' is still unkonwn. For example, a swapped in power
| supply in one sysetm can appear to be defective in another -
because
| it was actually marginal in both systems. Another problem.
Assumed
| is an entire power supply 'system' is only a power supply. Also
| wrong. A power supply is only one component of a power supply
| 'system'.
|
| Without numbers and other data, then better informed posters
cannot
| answer with a useful reply. That means using a 3.5 digit
mulitmeter
| (and two minutes) to obtain numbers and establish integrity of the
| entire power supply 'system'. In your case, the relevant part of
this
| procedure is to have computer access all peripherals
simultaneously
| (multitask) and then take voltage measurements on any one of
orange,
| purple, red, and yellow wires. Entire procedure (including part
| important to you) is "When your computer dies without
warning....."
| starting 6 Feb 2007 in the newsgroup alt.windows-xp at:
| http://tinyurl.com/yvf9vh
|
| If the power 'system' is not confirmed good, then it may cause
| everything else to appear defective. That only creates even more
| confusion. A computer can work just fine for ten minutes when the
| power supply 'system' always has numbers that are defective.
Those
| numbers are most relevant when load is maximum. Therefore watch a
| movie, while reading (or writing to) a CD-Rom, while downloading
from
| the internet, while reading the disk drive, etc ... all
| simultaneously. When doing all this, only then will some
defective
| power supplies be detected by numbers. Notice all those things are
| running, but the power supply is defective - and would cause
crashes
| later. Post those numbers here (even if good) because numbers may
| also contain further useful information.
|
| Ram is tested good. Then the same test is performed again when
| computer is in a room of 100 degree F - a perfectly good computer
| temperature. Or RAM is heated with a hairdryer on highest heat.
Any
| Ram that fails a diagnostic test in the computer (must be in your
| computer - not elsewhere) is defective or motherboard
semiconductors
| talking to that memory is defective. If it also passes at
elevated
| temperatures, then move on to other suspects.
|
| Very few items can shutdown a computer. That short list is
memory,
| power supply, CPU, video card, sound card, and some motherboard
| functions. Everything else including disk drive, keyboard,
monitor,
| and mouse are not on a usual suspect list. Anything can appear
| defective if numbers from a loaded power supply are not obtained
using
| that multimeter. Not having a meter is not even an option since
| meters are sold even to K-mart shoppers.
|
| Swapping parts tells us nothing useful and sometimes makes the
| problem exponentially complex. Temperatures from the motherboard
| temperature monitor may be useful. But again, to solve a problem,
get
| numbers and other diagnostic facts. Do not speculate. First
identify
| a suspect long before trying to fix anything.
|
| Heat is a diagnostic tool. If heat creates problems, then fix
the
| defective part. Don't cure symptoms by adding fans. Problem is
solved
| first by breaking the problem down into parts. Then establish
each
| part as 'known good' or 'known bad'. Currently you don't know of
| anything in either state. Everything is still questionable -
| unknown. Start by first confirming entire power supply 'system'
| integrity. Don't shotgun - swapping parts as if that proves
| anything. Demonstrated - a defective (marginal) power supply can
| still make computer appear to be working.

You mentioned something interesting that I did not consider - the
video card. It was acting strange for a bit and it's old one Geforce
5600FX Ultra. For some reason the OS switched to a PAL setup when it
should have been NTSC. I first noticed it in Excel. The video card
worked but it didn't with tv-out, obviously. The regional settings
were set at US even though I wasn't getting '$' but Euro in Excel.
After uninstalling and reinstall new and old video drivers I finally
noticed that it was using PAL. I switched it to NTSC and it worked
for 3-4 days and then started acting up by rebooting after 10
minutes. Unfortunatly, I just busted one of the legs on the cpu fan
and I don't have an extra one to replace. Rather than possibly
wasting money on a cpu fan, I ordered a new board and chip,
upgrading to the core duo side. Now I'm concerned about the Geforce
actually might be the problem. Is it safe to use with the new parts?

As far as the power supply, the original one has been working
non-stop in another computer for 3-days without an issue. It's
handling power intensive apps like video cap, re-rendering and
gaming like a charm. I'm persuaded to believe that this is not the
problem. I still would like to salvage the P4 if possible. If (and
hopefully) the video card is not the problem... going back to my
original question, is there a way to test either the board or chip
without using a replacement?

"peter"
| maybe it is overheating.........what temps are you showing in the
| BIOS??..keep it in the BIOS for 10-15 minutes and watch the temps
| maybe your Temp shut dowm limit is too low??...what is it set
at??.....in
| the BIOS
| peter

Hi Peter. At the moment I can't recheck because as I mentioned in
the other post, I busted one of the legs on the cpu fan. It was
running a bit warmer than I liked - around 125 degrees but I think I
read somewhere that this board's shutdown limit is around 145-150.
Even though idealy it should be around 110-115, it appears to be
within the acceptable parameters, unless the info I read was
incorrect. There isn't a setting for the shutdown limit, perhaps
that's one of the functions of the AI. Do these setting seam
reasonable to you?

On Aug 17, 1:45 pm, "Samantha" none wrote:
As far as the power supply, the original one has been working
non-stop in another computer for 3-days without an issue. It's
handling power intensive apps like video cap, re-rendering and
gaming like a charm. I'm persuaded to believe that this is not the
problem. I still would like to salvage the P4 if possible. If (and
hopefully) the video card is not the problem... going back to my
original question, is there a way to test either the board or chip
without using a replacement?

Exactly my point. A defective supply can do exactly as you have
posted for numerous technical reasons that could be learned if using
the meter. For example, do you know about the signaling between a
power supply and the rest of the system? If bad signaling was
compensated for in that second machine, then power supply would work
just fine. You are assuming the entire function in a power supply is
power output. Many other functions are ongoing with is why a
defective supply can appear to be working just fine in a second
system.

BTW, why do we fix things? Not to save time or money. We do it to
learn. Just another reason to get the meter to learn why a supply
that works just fine in a second system may actually be defective.
Swapping parts to fix a problem would teach nothing useful.

Meanwhile, that CPU is Intel? Then CPU will run slower during
higher temperatures - not crash and reboot.

Computer must work just fine even when room (and therefore CPU)
temperatures increase 20 degrees C. If temperature were a problem
(and not just CPU), then problem would occur very quickly when
operated in a 100 degree F room or when first heated as hot as
possible with a hairdryer on highest heat. Heat is a diagnostic tool.

If your running the CPU at 125 in a 100 degree room...temps are not the
problem...at least not CPU temp.
Which brings us to the Video card.................maybe its time for a new
one...
But its all moot now that you have ordered a new mobo and CPU.
When you've put that together take a look at the temps of the Video
Card....NVidea Control Panel...or upgrade it at the same time.
peter
"Samantha" none wrote in message
...
"peter"
| maybe it is overheating.........what temps are you showing in the
| BIOS??..keep it in the BIOS for 10-15 minutes and watch the temps
| maybe your Temp shut dowm limit is too low??...what is it set
at??.....in
| the BIOS
| peter

Hi Peter. At the moment I can't recheck because as I mentioned in
the other post, I busted one of the legs on the cpu fan. It was
running a bit warmer than I liked - around 125 degrees but I think I
read somewhere that this board's shutdown limit is around 145-150.
Even though idealy it should be around 110-115, it appears to be
within the acceptable parameters, unless the info I read was
incorrect. There isn't a setting for the shutdown limit, perhaps
that's one of the functions of the AI. Do these setting seam
reasonable to you?

When you get down to the motherboard OR CPU, and if everything else has
been **correctly** eliminated, it is almost always the motherboard. CPU
chips very rarely fail unless they are grossly abused by overheating or
overvoltage.

There is really no way for an end-user to really test these components
other than by substitution.

Are you SURE that the heatsink is mounted properly? I'd watch a CPU
temperature monitor (based on the INTERNAL CPU temperature sensor of the
chip itself) while (e.g. until) the failure occured to rule out overheating.

How did "someone test the RAM"?

Samantha wrote:
Hi,

Is there a way to troubleshoot a motherboard or cpu without using
another motherboard or cpu? (Concern about frying something)

I have another P4800-E Deluxe MB with a P43.0 chip. After about 10
minutes it keeps rebooting like it's overheating. I changed out the
power supply but that didn't fix the problem. Had someone test the
RAM and it's good. The BIOS shows that the cpu fan is running as
expected so the only thing left I can think of is either the board
or chip. Is there something I can do to find out which one? I'm
concerned about putting the chip in another board if that's the
problem or taking a chance on frying another chip if the board is
the problem. Is there a way to test one of them?

Thanks,
Samantha

I tried the video card in another computer and it's working good.
Back to square one, but I'm glad it's not the card. I guess down the
road I'll get a used mb to test the cpu. Thanks for your help,
Peter.

Samantha

"peter"
| If your running the CPU at 125 in a 100 degree room...temps are
not the
| problem...at least not CPU temp.
| Which brings us to the Video card.................maybe its time
for a new
| one...
| But its all moot now that you have ordered a new mobo and CPU.
| When you've put that together take a look at the temps of the
Video
| Card....NVidea Control Panel...or upgrade it at the same time.
| peter
|
| "peter"
| | maybe it is overheating.........what temps are you showing in
the
| | BIOS??..keep it in the BIOS for 10-15 minutes and watch the
temps
| | maybe your Temp shut dowm limit is too low??...what is it set
| at??.....in
| | the BIOS
| | peter
|
| Hi Peter. At the moment I can't recheck because as I mentioned
in
| the other post, I busted one of the legs on the cpu fan. It was
| running a bit warmer than I liked - around 125 degrees but I
think I
| read somewhere that this board's shutdown limit is around
145-150.
| Even though idealy it should be around 110-115, it appears to
be
| within the acceptable parameters, unless the info I read was
| incorrect. There isn't a setting for the shutdown limit, perhaps
| that's one of the functions of the AI. Do these setting seam
| reasonable to you?

Hi Barry. I pretty much resolved to let it go for now until I can
get a used mb. But to answer your questions, yes I did check the
heatsink and dusted out all the fans. As far as the RAM, I tried it
in another computer first and it seamed fine but you never know. So
I took it to someone who had diagnostic software and everything
passed. (didn't check the name of the program) I think at this point
I'm at a dead-end until I can get another mb, unless you have
another suggestion. Thanks for replying.

"Barry Watzman"
| When you get down to the motherboard OR CPU, and if everything
else has
| been **correctly** eliminated, it is almost always the
motherboard. CPU
| chips very rarely fail unless they are grossly abused by
overheating or
| overvoltage.
|
| There is really no way for an end-user to really test these
components
| other than by substitution.
|
| Are you SURE that the heatsink is mounted properly? I'd watch a
CPU
| temperature monitor (based on the INTERNAL CPU temperature sensor
of the
| chip itself) while (e.g. until) the failure occured to rule out
overheating.
|
| How did "someone test the RAM"?
|
| Samantha wrote:
| Hi,
|
| Is there a way to troubleshoot a motherboard or cpu without
using
| another motherboard or cpu? (Concern about frying something)
|
| I have another P4800-E Deluxe MB with a P43.0 chip. After about
10
| minutes it keeps rebooting like it's overheating. I changed out
the
| power supply but that didn't fix the problem. Had someone test
the
| RAM and it's good. The BIOS shows that the cpu fan is running as
| expected so the only thing left I can think of is either the
board
| or chip. Is there something I can do to find out which one? I'm
| concerned about putting the chip in another board if that's the
| problem or taking a chance on frying another chip if the board
is
| the problem. Is there a way to test one of them?
|
| Thanks,
| Samantha
|
|