Spontaneous reboots :-(

  • Thread starter Thread starter S.Boardman
  • Start date Start date
S

S.Boardman

Hello. As found in many posts below, I recently upgraded an Athlon XP2000+
on a MSI KT3 Ultra VIA KT333 chipset to the same board with a XP2500+, 512Mb
Corsair PC3200 RAM, plus new heatsink. I have Win 98SE.

I'd changed the FSB to 166, and RAM settings sorted, CAS 2.5, I thought it
was all stable and OK. Memtest was fine, Windows processor load was 100%
(using SETI) for three or four days running 24/7. The BIOS reported temp has
never been above 39 degrees, with system temp about 33 degrees (side off
case).

Last night, playing a 2D game (Northland demo), with SETI running in the
background as low priority, Windows rebooted with no errors or anything. (I
have played the game like this before with no probs.) Then it did it again
during Scandisk on the reboot. I had just got into the BIOS to check things
and it did it again. I have dropped back down to FSB133, and changed the
memory, too. I don't know if this will cure the problem.

Why should it do this after four days of running fine? The m/b, CPU and
memory should have no problems at this speed (FSB166). The PSU is 300W, it
has a two disk RAID to supply plus a DVD-ROM drive. The graphics card is a
nVidia GeForce 4 MX440.

Please help.
 
Hello. As found in many posts below, I recently upgraded an Athlon XP2000+
on a MSI KT3 Ultra VIA KT333 chipset to the same board with a XP2500+, 512Mb
Corsair PC3200 RAM, plus new heatsink. I have Win 98SE.

I'd changed the FSB to 166, and RAM settings sorted, CAS 2.5, I thought it
was all stable and OK. Memtest was fine, Windows processor load was 100%
(using SETI) for three or four days running 24/7. The BIOS reported temp has
never been above 39 degrees, with system temp about 33 degrees (side off
case).

Last night, playing a 2D game (Northland demo), with SETI running in the
background as low priority, Windows rebooted with no errors or anything. (I
have played the game like this before with no probs.) Then it did it again
during Scandisk on the reboot. I had just got into the BIOS to check things
and it did it again. I have dropped back down to FSB133, and changed the
memory, too. I don't know if this will cure the problem.

Why should it do this after four days of running fine? The m/b, CPU and
memory should have no problems at this speed (FSB166). The PSU is 300W, it
has a two disk RAID to supply plus a DVD-ROM drive. The graphics card is a
nVidia GeForce 4 MX440.

Please help.
Click to expand...

You may have mentioned it in the past, but we've covered a lot of
ground.... what make/model/etc power supply?

Did you get the northbridge heatsink/fan issue resolved, are sure it's
running cool enough?

Does the BIOS or window hardware monitor report good voltage levels?
Is it possible to check the power supply voltages with a voltage
meter?

Does it only become instable after running for a while, or can you
duplicate this when the system has been off, cooled down, hasn't ran
for very long? Are you or were you running with the case cover off?
If you're now using a passive northbridge cooler, in some situations
it can help to leave the cover on.


Dave
 
kony said:
You may have mentioned it in the past, but we've covered a lot of
ground.... what make/model/etc power supply?
Click to expand...

Qtec Model ADT300 300W It has a Pentium 4 sticker on it.
Did you get the northbridge heatsink/fan issue resolved, are sure it's
running cool enough?
Click to expand...

I added the oil and it seems to be fine. The heatsink is on the 'to do'
list. HOw can I tell how cool it is? Unless it is reported as 'system temp'
(which is about 33 deg) I can't see any other temperature given.
Does the BIOS or window hardware monitor report good voltage levels?
Click to expand...

The Windows monitor says VCore 1.65 V, but it's not brilliant, I'll come
back to you with the BIOS ones.
Is it possible to check the power supply voltages with a voltage
meter?
No.

Does it only become instable after running for a while, or can you
duplicate this when the system has been off, cooled down, hasn't ran
for very long? Are you or were you running with the case cover off?
If you're now using a passive northbridge cooler, in some situations
it can help to leave the cover on.
Click to expand...
It had been running for about 4 days, CPU load 100%, with the cover off.
This because firstly I'm keeping an eye on the northbridge fan, although it
seems fine so far. Also the RAID cables are the flat ribbon sort and they
obstruct the air flow. I could close it up now I have the new heatsink/fan,
I suppose. I turned it off last night, after changed the FSB to 133. It has
been running OK so far this morning at 133. I don't want to risk my data too
much. I have the new HDs, so I could attach one and clone the drive(s), but
I don't want it to reboot right in the middle of it!
 
S.Boardman said:
snip

Qtec Model ADT300 300W It has a Pentium 4 sticker on it.

I added the oil and it seems to be fine. The heatsink is on the 'to do'
list. HOw can I tell how cool it is? Unless it is reported as 'system temp'
(which is about 33 deg) I can't see any other temperature given.

The Windows monitor says VCore 1.65 V, but it's not brilliant, I'll come
back to you with the BIOS ones.

It had been running for about 4 days, CPU load 100%, with the cover off.
This because firstly I'm keeping an eye on the northbridge fan, although it
seems fine so far. Also the RAID cables are the flat ribbon sort and they
obstruct the air flow. I could close it up now I have the new heatsink/fan,
I suppose. I turned it off last night, after changed the FSB to 133. It has
been running OK so far this morning at 133. I don't want to risk my data too
much. I have the new HDs, so I could attach one and clone the drive(s), but
I don't want it to reboot right in the middle of it!
Click to expand...
From the BIOS:
Vcore 1.648V
+5.0V 5.085
+12.0V 10.889 flicks to 11.011
-12.0V -11.989
-5V -5.177
Battery 3.280V
+5 SB 4.872V

DDR voltage is Auto, with options of 2.6, 2.7, 2.8
AGP voltage is Auto, with options of 1.6, 1.7, 1.8

I would also point out that just now when I rebooted, apart from ZoneAlarm
crashing, I got vertical green stripes about a quarter of an inch deep
across the top of the screen and everything froze. I've had this
*occasionally* before. Is that indicative of anything?
 
+12.0V 10.889 flicks to 11.011
!!!

I would also point out that just now when I rebooted, apart from ZoneAlarm
crashing, I got vertical green stripes about a quarter of an inch deep
across the top of the screen and everything froze. I've had this
*occasionally* before. Is that indicative of anything?
Click to expand...

ZoneAlarm could crash due to (other stability) problems or just be in
the mood to crash itself, certain versions seem buggy to me but
unfortunately I don't remember WHICH versions. There's no point in
worrying too much about that or the green stripes yet though, at this
point it's looking like you need a new power supply. Even though
technically there's supposed to be a 10% tolerance on a PSU 12V rail,
in practice when it dips below about 11.7V it's a sign of a power
supply just barely adequate, let alone 10.9V

Best value power supply might be a Sparkle/Fortron 300-350W, about $35
USD, or their 400W for $60 USD (is significantly beefier than the
350W, is worth the price difference IF you needed 400W but right now
you don't) though for more $ there's a lot more choices like Antec or
PC Power & Cooling, Delta...
..


Dave
 
kony said:
ZoneAlarm could crash due to (other stability) problems or just be in
the mood to crash itself, certain versions seem buggy to me but
unfortunately I don't remember WHICH versions. There's no point in
worrying too much about that or the green stripes yet though, at this
point it's looking like you need a new power supply. Even though
technically there's supposed to be a 10% tolerance on a PSU 12V rail,
in practice when it dips below about 11.7V it's a sign of a power
supply just barely adequate, let alone 10.9V

Best value power supply might be a Sparkle/Fortron 300-350W, about $35
USD, or their 400W for $60 USD (is significantly beefier than the
350W, is worth the price difference IF you needed 400W but right now
you don't) though for more $ there's a lot more choices like Antec or
PC Power & Cooling, Delta...
.
Click to expand...
Just rebooted again :-((
You know what's coming next, then, don't you? Is it just a case of checking
the dimensions and screwing another one in its place? I'll go for the 400W.
I'll see watt (geddit?) I can find on the UK sites, Sparkle or Fortron, you
say...
 
snip!
Best value power supply might be a Sparkle/Fortron 300-350W, about $35
USD, or their 400W for $60 USD (is significantly beefier than the
350W, is worth the price difference IF you needed 400W but right now
you don't) though for more $ there's a lot more choices like Antec or
PC Power & Cooling, Delta...
.
Click to expand...
Just bought for £60 inc delivery, FSP 400W ATX PSU (FSP400-60PFN).
Dabs wanted £77, not including delivery!
Please tell me this is the one you meant :-)
 
Just rebooted again :-((
You know what's coming next, then, don't you? Is it just a case of checking
the dimensions and screwing another one in its place? I'll go for the 400W.
I'll see watt (geddit?) I can find on the UK sites, Sparkle or Fortron, you
say...
Click to expand...

Yes, a standard ATX has the following dimensions in millimeters:

rear (side that screws to case) 150W x 86H
Length (front to back of case) 140


Sparkle/Fortron are among the best values in the US, but over there...
don't know.

One drawback to the Sparkle/Forton 400W is no air intake on the
underside. That is a useful feature and it's unfortunate that they
don't have it. With your CPU socket so near the top of the board it
might be even more beneficial to you as it would more directly remove
heatsink exhaust instead of recirculating it through the 'sink fan
again. Antec is another option, has the fan underneath.


Dave
 
S.Boardman said:
It had been running for about 4 days, CPU load 100%, with the cover
off.
Click to expand...

FWIW (and you may already know this), but running with the cover off
tends to really screw up the airflow that your case attempts to create
with its cooling fans. And running with CPU pegged tends to generate
a lot of heat. It's possible your BIOS could've sensed an overtemp
condition and triggered the reboot, or its possible the higher temps
threw an exception or caused some sort of error that caused the OS to
rebot.

Is this Windows? If so, turn auto reboot on error off--that way you
can see the BSOD to see what exactly happened if the OS triggered the
reboot.

In win2k this setting is here:
start>settings>controlpanel>system>advanced>startupandrecover>
Uncheck AutomaticallyReboot
Check write event to system log

If write event to system log is already checked, perhaps the cause of
your reboot is in your EventViewer log in the system log
section.
Event viewer can be found in win2k under:
Start>settings>controlpanel>administrativetools>eventviewer

Best Regards,
 
snip!
Just bought for £60 inc delivery, FSP 400W ATX PSU (FSP400-60PFN).
Dabs wanted £77, not including delivery!
Please tell me this is the one you meant :-)
Click to expand...


Wow, you really get things done quick!
Yes, it's the one I meant.


Dave
 
kony said:
Wow, you really get things done quick!
Yes, it's the one I meant.
Click to expand...
Perhaps I should have bought the Antec one. Never mind, my PC is nice and
cool now anyway. How can I tell what the temp is under the northbridge fan?
 
Todd H. said:
FWIW (and you may already know this), but running with the cover off
tends to really screw up the airflow that your case attempts to create
with its cooling fans. And running with CPU pegged tends to generate
a lot of heat. It's possible your BIOS could've sensed an overtemp
condition and triggered the reboot, or its possible the higher temps
threw an exception or caused some sort of error that caused the OS to
rebot.

Is this Windows? If so, turn auto reboot on error off--that way you
can see the BSOD to see what exactly happened if the OS triggered the
reboot.

In win2k this setting is here:
start>settings>controlpanel>system>advanced>startupandrecover>
Uncheck AutomaticallyReboot
Check write event to system log

If write event to system log is already checked, perhaps the cause of
your reboot is in your EventViewer log in the system log
section.
Event viewer can be found in win2k under:
Start>settings>controlpanel>administrativetools>eventviewer

Best Regards,
Click to expand...

Hi. I'm running Win 98, and thanks to a good heatsink and fan, my Athlon has
never got above 39 degs, in Windows, full load.
 
Perhaps I should have bought the Antec one. Never mind, my PC is nice and
cool now anyway. How can I tell what the temp is under the northbridge fan?
Click to expand...

it the CPU stays cool enough I won't worry about the airflow, the main
thing is that you now have a decent power supply... the previous power
supply was obviously having problems, let's just hope it's problems
were causing the main problem... even if they weren't the old power
supply was dying, you're lucky to have caught it before you were stuck
without a system for a long time.

So long as the northbridge fan is turning at a fair rate and the
heatsink looks properly attached to the board, using thermal compound
or transfer tape cement (whatever, including the original solution) it
should be adequate.


Dave
 
kony said:
fan?

it the CPU stays cool enough I won't worry about the airflow, the main
thing is that you now have a decent power supply... the previous power
supply was obviously having problems, let's just hope it's problems
were causing the main problem... even if they weren't the old power
supply was dying, you're lucky to have caught it before you were stuck
without a system for a long time.
Click to expand...

Hence the fast call for a new one :-)
So long as the northbridge fan is turning at a fair rate and the
heatsink looks properly attached to the board, using thermal compound
or transfer tape cement (whatever, including the original solution) it
should be adequate.
Click to expand...
I assume it is. I have no way of knowing *how* fast though. It doesn't smell
'hot', so I assume it's OK.
 
I assume it is. I have no way of knowing *how* fast though. It doesn't smell
'hot', so I assume it's OK.
Click to expand...

So long as the fan isn't wobbling around and turning very slow it
should be adequate, though not knowing the quality of that specific
fan or the amount of wear I can't guess how long it will continue to
function.... could work for years or fail at any time, though I'd
expect it to give similar warning if/when it fails again.


Dave
 
Yeah well. Wasn't in stock was it? Got an Antec Truepower one. Didn't have a
400W in stock (lead time 21 days) so got a 430W one. Not going to run out of
Watts, am I ;-)
 
Back
Top