sudden restart A7V8X

  • Thread starter Thread starter Ewout
  • Start date Start date
E

Ewout

Hello,
I've got some problems with my PC that I can't locate. Once in a while
(sometimes 3 to 5 times a week), my computer decides to reboot without a
notice. I went to investigate but I can't find where the problem lies!!!
I'm not sure if this is normal, but i've got a small current on my case.
I don't have a ground socket (sorry my English is not very good, i mean
where you get the power from the wall, it doesn't have ground/earth), so
i lead the current off to my radiator.

I have run several tests with a program called Memtest-86, and I had no
problems. I ran it for about an hour, should that be enough?
Then I tried Prime95, which gave me an error almost instantly in test 1:

FATAL ERROR: Resulting sum was 3.219331625009081e+016, expected:
3.30754608703769e+016

OR

FATAL ERROR: Rounding was 0.5, expected less than 0.4

with the torture test the CPU temperature reached 77C, and around that
time it gave another rounding error. These were the values Asus Probe
reported at that moment (not the exact moment, but just before):

CPU Temperature 77 C/ 170 F
MB Temperature 35 C/ 95 F
CPU Fan 4470
+12V 12.768
+5V 4.945
+3.3V 3.344
VCore 1.824

I don't have any reference as to what these values should be like ideally...

The computer usually reboots during heavy operations, but I haven't been
able to get any values from these moments yet, any ideas on this? I have
not yet been able to "create" such a reboot on purpose...

Ok, here comes my setup:

Motherboard: Asus A7V8X
Bios version: 1009
CPU: AMD Athlon XP 2100+
Family 6
Model 6
Stepping 2
Clock Frequencies (I did NOT overclock!!! Didn't even try):
Internal 1,733
External 133.31
System bus 266.6 (DDR)
Multiplier 13
Videocard: NVIDIA GeForce4 Ti 4200 with AGP8X

(is this enough information?)

Does anyone have any idea on how I can find the thing that causes the
system to be unstable?!?

thanks a lot in advance! I will post updates as soon as i have more
information

Ewout
 
Could be a number of things - but sounds like CPU overheating. Youve tested
ram, but Im no expert on prime95/memtest - others can help with that. As for
the overheating thing, try really taxing your system by running a number of
games/applications together or use 3dmark or something. If it does it again,
on reboot go into your bios and check the cpu temperature straight away. 77c
sounds kinda high though.
Then again it could just be a voltage thing.
 
Blaedmon said:
Could be a number of things - but sounds like CPU overheating. Youve tested
ram, but Im no expert on prime95/memtest - others can help with that. As for
the overheating thing, try really taxing your system by running a number of
games/applications together or use 3dmark or something. If it does it again,
on reboot go into your bios and check the cpu temperature straight away. 77c
sounds kinda high though.
Then again it could just be a voltage thing.
Click to expand...

I did some more tests and i reached 80 C, where could I find information
on "normal" temperatures for my processor?

Ewout
 
Ewout said:
Hello,
I've got some problems with my PC that I can't locate. Once in a while
(sometimes 3 to 5 times a week), my computer decides to reboot without a
notice. I went to investigate but I can't find where the problem lies!!!
I'm not sure if this is normal, but i've got a small current on my case.
I don't have a ground socket (sorry my English is not very good, i mean
where you get the power from the wall, it doesn't have ground/earth), so
i lead the current off to my radiator.
Click to expand...

I've used a copper cold water pipe, in a house that didn't have
three prong plugs, so that should be OK. The only risk you run,
is getting a shock from the computer case if it is ungrounded.
The common mode filter in the power supply, if there is one,
dumps a tiny current into the ground wire, and that is what
causes a shock if there is no earth ground.
I have run several tests with a program called Memtest-86, and I had no
problems. I ran it for about an hour, should that be enough?
Then I tried Prime95, which gave me an error almost instantly in test 1:

FATAL ERROR: Resulting sum was 3.219331625009081e+016, expected:
3.30754608703769e+016

OR

FATAL ERROR: Rounding was 0.5, expected less than 0.4

with the torture test the CPU temperature reached 77C, and around that
time it gave another rounding error. These were the values Asus Probe
reported at that moment (not the exact moment, but just before):

CPU Temperature 77 C/ 170 F
MB Temperature 35 C/ 95 F
CPU Fan 4470
+12V 12.768
+5V 4.945
+3.3V 3.344
VCore 1.824

I don't have any reference as to what these values should be like ideally...

The computer usually reboots during heavy operations, but I haven't been
able to get any values from these moments yet, any ideas on this? I have
not yet been able to "create" such a reboot on purpose...
Click to expand...

Get a copy of either Motherboard Monitor (MBM5) or Asus Probe. Install
only one of these programs at a time, as they don't work properly if
two of these programs run at the same time. These programs record
voltages, temperatures, and fan speeds, to a text file. They can record
values every 10 seconds. You can run Prime95 and watch the values
on the dashboard of MBM5, or if you want to run 3Dmark as a test,
you don't have to use the dashboard, as the text file will contain
all the readings.

http://mbm.livewiredev.com/download.html
http://mbm.livewiredev.com/mobolist.html

77C sounds kind of high. 65C might be a more reasonable maximum
value.
Ok, here comes my setup:

Motherboard: Asus A7V8X
Bios version: 1009
CPU: AMD Athlon XP 2100+
Family 6
Model 6
Stepping 2
Clock Frequencies (I did NOT overclock!!! Didn't even try):
Internal 1,733
External 133.31
System bus 266.6 (DDR)
Multiplier 13
Videocard: NVIDIA GeForce4 Ti 4200 with AGP8X

(is this enough information?)

Does anyone have any idea on how I can find the thing that causes the
system to be unstable?!?

thanks a lot in advance! I will post updates as soon as i have more
information

Ewout
Click to expand...

Things you can try:

1) Set the OVER_VOLT1 jumper to disabled. (Section 2.7 "Jumpers"
in the manual.)
2) Set CPU Vcore voltage in the BIOS to a lower value.
3) Remove the heatsink/fan and use some thermal paste to improve
the contact between the CPU die and the heatsink/fan. Remove
whatever material is currently on there before applying
the paste. Clean the surface of the heatsink thoroughly
before adding new paste. Be careful not to chip the silicon
die of the CPU - it is brittle. If the CPU die is cracked or
chipped, that could cause the CPU to run hot or cause the
CPU to stop running entirely. Examine the CPU to see if there
are any burned spots on it. Take the computer to a repair
shop if you aren't comfortable doing this procedure.
4) Use a copper core heatsink or an all copper heatsink, as copper
has better thermal conductivity than aluminum.
5) Increase the air flow rate of the CPU fan. You can buy fans
that move more air than your current fan. Disable Q-Fan if
the computer has that feature in the BIOS.
6) Your motherboard temperature of 35C isn't bad. You could add
another case fan to your computer case, but that might only
give you a few more degrees C improvement.

Sometimes the fault in a case like this, is the processor drawing
more current than it is supposed to.

HTH,
Paul
 
Using the radiator to ground the computer should be a good way to do it,
if the connections are good. You need zero ohms resistance between the
computer ground (usually case) and the radiator.

Your real problem seems to be overheating. With a MB temp of 35C the
processor should not get much above 50C-60C. The ideal situation would
be for the processor to be at the same temp as the MB. I'd suggest that
a new fan/heatsink be tried or at least make sure the current one is
correctly positioned and has adequate heat sink paste or pad applied.

Check your computer case for obstruction of air flow. Make sure the case
fan(s) create air flow over the processor heatsink. Make sure that the
fan on the processor heatsink is moving. Clean dust off fan blades if
necessary.

Try running the computer with the case open. If the CPU temp stays
lower, then there is definitely an air flow problem.
 
Things you can try:
1) Set the OVER_VOLT1 jumper to disabled. (Section 2.7 "Jumpers"
in the manual.)
2) Set CPU Vcore voltage in the BIOS to a lower value.
3) Remove the heatsink/fan and use some thermal paste to improve
the contact between the CPU die and the heatsink/fan. Remove
whatever material is currently on there before applying
the paste. Clean the surface of the heatsink thoroughly
before adding new paste. Be careful not to chip the silicon
die of the CPU - it is brittle. If the CPU die is cracked or
chipped, that could cause the CPU to run hot or cause the
CPU to stop running entirely. Examine the CPU to see if there
are any burned spots on it. Take the computer to a repair
shop if you aren't comfortable doing this procedure.
4) Use a copper core heatsink or an all copper heatsink, as copper
has better thermal conductivity than aluminum.
5) Increase the air flow rate of the CPU fan. You can buy fans
that move more air than your current fan. Disable Q-Fan if
the computer has that feature in the BIOS.
6) Your motherboard temperature of 35C isn't bad. You could add
another case fan to your computer case, but that might only
give you a few more degrees C improvement.

Sometimes the fault in a case like this, is the processor drawing
more current than it is supposed to.

HTH,
Paul
Click to expand...

Thanks very much for you answer!!! The problem turned out to be the
temperature, definately. I ran the torture test from Prime95 a few
times, and the temperature stayed around 80C, and it stopped almost
instantly evertime!!! Then it finally happened, the computer rebooted!
And again, and again... I just had the chance to see the temperature was
around 80 degrees according to Asus probe. I decided to see what the
BIOS said, which turned out to be worse: 86,5C!!!!!!!!!!! Which is
awfully hot apparently :o)
Then I discovered the heatsink blades (the tiny ones) were all fulled
with dust (only after a year and a half!!!) I will post a picture of
that http://www.extrapuur.nl/stof.jpg
I cleaned it and put the fan back on it. Then I bought some new thermal
paste and applied it (after removing the old stuff with
nailpolish-remover). The temperature was back at 45C again without much
CPU use, and even with Prime95 i didn't get above 57C!!! Which is kind
of a difference! Also, Prime95 reported that my system was 100% stable
after all.

Thanks everyone for your help/advices!

Ewout
 
Robert Gault said:
Using the radiator to ground the computer should be a good way to do it,
if the connections are good. You need zero ohms resistance between the
computer ground (usually case) and the radiator.
Click to expand...

***ASSUMING THAT YOUR GROUND ON THE OUTLET IS 0 Volts***

If there is any voltage difference between the ground on your outlet and the
ground on the radiator, you could be pulling voltage THROUGH your case from
the outlet.

If the radiator is grounded better than the ground on the outlet, and
something elsewhere blows up, the voltage will go through your PC to the
radiator to get to ground instead of through the electrical system.

ONE GROUND PER DEVICE AND KEEP THEM ALL COMMON!
 
Back
Top