Random shutdowns



Jason
07-10-2005, 12:49 AM
Ok this is confusing me

Intel 550
D915PBL
1Gb Geil DDr2 Pc4300
Ati X700 Pro 256

basically what is happening is when im doing something pretty hefty on the
pc it will shutdown at random

some times it last for 5 seconds sometime for 10 hours, event viewer was
giving me a Ide controller error so i replaced the ide cable and no error but
still shutting down. I have doubts that it is a heat issue since i have the
thertake tower112 with dual 92mm Vantec tornados, MBM and IDCC both say under
full load im still under 45 celcius , idle around 36 even with prime 95
running it doesnt jump over 47c ever, so i dont think its overheating unless
i have faulty sensors. ive installed about 10 of these heatsinks so i know i
have it right, and im using Artic silver 5 im confused. basically it shuts
completely off like i pulled the plug from the monster surge, i know all 4
fans are running because the sensors all say they are turning at the speeds
they should be. please help

Chuck
07-10-2005, 12:49 AM
This might be almost anything. Loose connections, bad power supply, and so
on. Have you cleaned and reseated all the connectors, cards, memory, etc.
How about the wiggle test for any internal or external cables that might
cause a problem.

"Jason" <Jason@discussions.microsoft.com> wrote in message
news:7D4E72F8-EDF1-461F-B12D-75172554BC68@microsoft.com...
> Ok this is confusing me
>
> Intel 550
> D915PBL
> 1Gb Geil DDr2 Pc4300
> Ati X700 Pro 256
>
> basically what is happening is when im doing something pretty hefty on the
> pc it will shutdown at random
>
> some times it last for 5 seconds sometime for 10 hours, event viewer was
> giving me a Ide controller error so i replaced the ide cable and no error
> but
> still shutting down. I have doubts that it is a heat issue since i have
> the
> thertake tower112 with dual 92mm Vantec tornados, MBM and IDCC both say
> under
> full load im still under 45 celcius , idle around 36 even with prime 95
> running it doesnt jump over 47c ever, so i dont think its overheating
> unless
> i have faulty sensors. ive installed about 10 of these heatsinks so i know
> i
> have it right, and im using Artic silver 5 im confused. basically it
> shuts
> completely off like i pulled the plug from the monster surge, i know all 4
> fans are running because the sensors all say they are turning at the
> speeds
> they should be. please help
>

Jason
07-10-2005, 12:49 AM
yup, just torw the whole thing apart and lapped the heatsink a little
better..go 800 grit! reapplied AS5 and put everything back together, manually
checkd both fans were working outside of case (test bench PSU) all
connections are secure, all soldered wires are insulated with liquid tape and
dried and shrink wrapped...i agree it could be anything...mihgt just hafta
buy a new DFI and then watercool cuz im out of ideas

Michael W. Ryder
07-10-2005, 12:49 AM
Jason wrote:
> yup, just torw the whole thing apart and lapped the heatsink a little
> better..go 800 grit! reapplied AS5 and put everything back together, manually
> checkd both fans were working outside of case (test bench PSU) all
> connections are secure, all soldered wires are insulated with liquid tape and
> dried and shrink wrapped...i agree it could be anything...mihgt just hafta
> buy a new DFI and then watercool cuz im out of ideas


Have you used a monitoring program to check the power supply's voltages?
You might also try a different power supply to see if the current one
is failing. Your mention of the top getting hot sounds like a power
supply problem or the heat getting trapped in the top of the case and
affecting the power supply.

w_tom
07-10-2005, 12:49 AM
The onboard voltage monitor is sufficient for monitoring a
change but is too inaccurate to establish voltages. However
Michael is right on the money. Verifying power supply
voltages both at power up and when everything is accessed
simultaneously is a very first suspect investigated in cases
like this.

Important information from the system (event) log and from
Device Manager do suggest where to start looking. But an IDE
problem alone does not crash XP. Only that subsystem would
shutdown.

All that heat sink and Arctic Silver stuff is nonsense.
Same applies to cleaning contacts or blowing out dust. Those
are actions recommended by the technically naive. Those CPU
temperatures are so well below maximum temperatures as to even
make use of Arctic Silver ridiculous.

First one collects numbers as Michael has recommended.
Provided above are some of the first and most informative.
System's comprehensive diagnostics verify hardware is
functional. In this case, most likely suspects include video
card, sound card, and memory. And of course the disk drive
diagnostic to verify that IDE port. If the system
manufacturer is not responsible (did not provide comprehensive
diagnostics), then those diagnostics must be downloaded from
the component manufacturers or from third parties (ie Memtst86
or Docmem).

Once diagnostics have been run cool, then again execute at
another temperature that semiconductors find lovely - tropical
- ideal. Put a hairdryer on high and heat the selective
components as diagnostics tests those components. If you
touch a seminconductor and it does not burn skin, then it is
not too hot. But if a semiconductor fails diagnostics even
under a hairdryer on high - still extremely cool to
semiconductors - then you have IDed the failed part.

I doubt this is an IDE problem because Windows would not
crash. Up top is a most likely reason for IDE problems which
is why Michael's suggestion is so important. Measure those
voltages with everything being accessed simultaneously - disk
drives, CD-ROM, sound card, network interface, and modem.
Voltages should remain upper 3/4 of limits.

The only way to obtain valid voltage measurements is with a
3.5 digit multimeter - a tool so ubiquitous as to be sold even
in Sears, Home Depot, Lowes, and Radio Shack. A tool also
necessary to calibrate that onboard voltage monitor. In your
case, measurements of the orange, yellow, red, and purple
wires (from power supply) are most important.


"Michael W. Ryder" wrote:
> Have you used a monitoring program to check the power supply's voltages?
> You might also try a different power supply to see if the current one
> is failing. Your mention of the top getting hot sounds like a power
> supply problem or the heat getting trapped in the top of the case and
> affecting the power supply.

Joshua Smith [MSFT]
07-10-2005, 12:50 AM
Aside from all the hardware suggestions it could be a driver issue. To
determine if a driver is causing the problem I need you to enable driver
verifier.
Steps:
1) Windows Key + R
2) Type in 'verifier' and hit enter
3) Make sure 'Create Standard Setting' is selected and hit next
4) Click on 'Select all drivers installed on this computer' and hit Finish
5) Reboot

There is a possibility that your computer will crash on reboot. If this
occurs hit F8 when rebooting just before the windows logo screen and select
the safe mode boot option. Follow the same steps above but on step 4 choose
'Select driver names from a list'; hit next; check the box next to any
driver where the provider is not Microsoft; hit Finish; reboot.

This will slow the performance of you computer a little while enabled but
will hopefully catch the driver causing corruption. Next time you crash
the blue screen will hopefully say something like
"DRIVER_VERIFIER_DETECTED_VIOLATION". If this occurs please send the
corresponding minidump (by default it is at c:\windows\Minidump ) my way.
If you have any questions or I didn't explain something well enough don't
hesitate to e-mail me (remove "online") back. Good Luck,


Joshua Smith
OpenGL Test Lab
Microsoft
-----

Get Secure! www.microsoft.com/security

This posting is provided "AS IS" with no warranties, and confers no rights


"w_tom" <w_tom1@hotmail.com> wrote in message
news:428624EC.28645A79@hotmail.com...
> The onboard voltage monitor is sufficient for monitoring a
> change but is too inaccurate to establish voltages. However
> Michael is right on the money. Verifying power supply
> voltages both at power up and when everything is accessed
> simultaneously is a very first suspect investigated in cases
> like this.
>
> Important information from the system (event) log and from
> Device Manager do suggest where to start looking. But an IDE
> problem alone does not crash XP. Only that subsystem would
> shutdown.
>
> All that heat sink and Arctic Silver stuff is nonsense.
> Same applies to cleaning contacts or blowing out dust. Those
> are actions recommended by the technically naive. Those CPU
> temperatures are so well below maximum temperatures as to even
> make use of Arctic Silver ridiculous.
>
> First one collects numbers as Michael has recommended.
> Provided above are some of the first and most informative.
> System's comprehensive diagnostics verify hardware is
> functional. In this case, most likely suspects include video
> card, sound card, and memory. And of course the disk drive
> diagnostic to verify that IDE port. If the system
> manufacturer is not responsible (did not provide comprehensive
> diagnostics), then those diagnostics must be downloaded from
> the component manufacturers or from third parties (ie Memtst86
> or Docmem).
>
> Once diagnostics have been run cool, then again execute at
> another temperature that semiconductors find lovely - tropical
> - ideal. Put a hairdryer on high and heat the selective
> components as diagnostics tests those components. If you
> touch a seminconductor and it does not burn skin, then it is
> not too hot. But if a semiconductor fails diagnostics even
> under a hairdryer on high - still extremely cool to
> semiconductors - then you have IDed the failed part.
>
> I doubt this is an IDE problem because Windows would not
> crash. Up top is a most likely reason for IDE problems which
> is why Michael's suggestion is so important. Measure those
> voltages with everything being accessed simultaneously - disk
> drives, CD-ROM, sound card, network interface, and modem.
> Voltages should remain upper 3/4 of limits.
>
> The only way to obtain valid voltage measurements is with a
> 3.5 digit multimeter - a tool so ubiquitous as to be sold even
> in Sears, Home Depot, Lowes, and Radio Shack. A tool also
> necessary to calibrate that onboard voltage monitor. In your
> case, measurements of the orange, yellow, red, and purple
> wires (from power supply) are most important.
>
>
> "Michael W. Ryder" wrote:
>> Have you used a monitoring program to check the power supply's voltages?
>> You might also try a different power supply to see if the current one
>> is failing. Your mention of the top getting hot sounds like a power
>> supply problem or the heat getting trapped in the top of the case and
>> affecting the power supply.


Random shutdowns