Random Reboots ?

  • Thread starter Thread starter andyw
  • Start date Start date
A

andyw

I recently built a new PC around a Koolance water-cooled case.

For a over a month now it has been running fine, however now all of a
sudden it has started to reboot at random for no apparent reason, I can
be using it for 4 hours and it will restart or it could be 10 minutes.

Here is the spec of PC:

Intel P4 3.4
ASUS P5AD2 Premium
ASUS X800 PCIe
PC Power and Cooling Turbo Cool 510
1 x 1 gig stick of crucial PC 4200 DDR2 RAM
2 36gig raptors drive configured on Intel RAID 0
XP Pro

As I mentioned before the system is water cooled, that includes the
CPU, graphics chip and hdd.

At first I though the culprit was something overheating was causing the
system reboot, however the system does always reboot under heavy loads,
just before it rebooted yesterday the CPU was at 28c and the
motherboard was at 45c.

Last night however I though I had cracked it as I removing an external
hdd I recently added. The reason for this is that this hdd required an
external power supply and at the moment I have one wall socket with a
six way extension bar coming out of it all of which I utilised by
various computer products. My reasoning behind this was perhaps the
demand for power was out stripping supply.

I then left it on over night but alas it still rebooted.

The only other problem I've been getting it is that if the USB card
reader is plugged in the system will sometimes not boot fully, not sure
if this is related or not.

I'm really looking for any tests that I can do which will prove that
a component is faulty, as stripping this system down when it's full
of water is no easy task.

Oh yes, all overclocking has been removed, the system is running as
'standard'
 
Not really, I purposley spent about 4 times more than the average price
of a PSU to avoid that actual problem.
I'm abit worried that 45c is too hot for the motherboard ?
 
Yes, 45c is to Hot.
Is you Northbridge under an fan-cooler or is it watercooled?

Perhaps the Northbridge gets to hot...

If u are sure your Northbridge-Cooler is ok, i think it can be the
power-Supply...

Joerg
 
Northbridge ?? Not quite sure what you mean, if your referring to the
mother chip than that has no cooling, except what was already on the
motherboard ?
 
Your Mainboard have a North and a Southbridge.

Sometimes the Northbridge-Chip is activ cooled (with fan). Sometimes there
is only a passiv-Cooler.
 
"andyw" said:
Northbridge ?? Not quite sure what you mean, if your referring to the
mother chip than that has no cooling, except what was already on the
motherboard ?
Click to expand...

The first thing you should do, is set up windows to "blue screen",
rather than automatically reboot (I don't have WinXP, but possibly
in Control Panels, System, Advanced tab, Startup and Recovery).
That way, you'll get a bunch of numbers, but also perhaps the
name of the software module where the crash is occurring. If the
error is in the same module each time, then the name of the module
should give a good idea of what hardware or driver code is
responsible for the crash. If the module is random, that points
more at a hardware error, like bad RAM/motherboard/CPU.

Your CPU and motherboard temperatures appear to be reversed, as it
is unlikely the CPU is at 28C and the motherboard at 45C.

You might try uninstalling all hardware monitor programs. If
you've got Asus Probe and MBM5 loaded, they access the SMBus
and both programs shouldn't be installed at the same time.
Uninstall both and see if the computer is more stable.

If the CPU is at 45C, then that means the water cooler
is working. You won't find many people with air cooled
3.4GHz processors at that temperature at full load.

The standard tests I recommend are memtest86 (memtest.org)
and Prime95 (mersenne.org). Memtest86 is the first test to
run, and it tests all the memory. Prime95 tests memory,
CPU, and Northbridge, and it sometimes uncovers problems that
don't show up with memtest86 (i.e. the memory can be
bad and still pass memtest86). By doing these two tests,
you'll know whether the computing core is solid or not.

For 3D testing, you can use your favorite game as a test,
or I use 3DMark as a stability test (loop the demo mode for
an extended test). It will do a good job of warming up the
hardware and moving a lot of data around. If the machine
won't spontaneously reboot when doing Prime95, but does reboot
when running a 3D game, then it could be a hardware problem
with the graphics card or a software problem with the
graphics card driver.

Also, a question for you. Why only one stick of RAM, on
a dual channel machine ? Are you buying a matching one
soon to go with it ? If you only want 1GB ram, then
2 x 512MB sticks would be the way to go, if you plan
on overclocking. Dual channel gives you more memory
bandwidth.

HTH,
Paul
 
Back
Top