System Failure - CPU Task

  • Thread starter Thread starter Keske Saram
  • Start date Start date
K

Keske Saram

I have a P4B motherboard which has run well under win NT for several
years. Suddenly, today it does not start any more at all (no disk
spinning etc) and an ugly voice repeats (see above).
There is another tread about this message, but it is completely
focusing on power issues.
My question: is this the only possible cause for such a failure - (my
power supply seems to be OK, feeding also different devices). Or does
it *have to be* a power issue? otherwise, what else? could you
get such a message if the CPU is (partly) broke (e.g. overheating
etc)?
tx
 
Without numbers, then you have no idea if a power supply is
good. A bad power supply will still illuminate lights and
spin disk drives. Meanwhile the power supply is sending the
computer a signal that it is defective. How do human eyes see
that signal? You don't without a 3.5 digit multimeter.

Start the diagnosis by establishing what is good and what is
unknown. Since everything else in the system can appear to be
bad when a marginal power supply causes problems, then the
power supply 'system' is where one starts to verify what is
good. That means two minutes and a meter.

Once power supply 'system' (which is more than just a power
supply) is confirmed good, then move on to then next 'usual'
suspect - computer hardware. This is where comprehensive
hardware diagnostics - no operating system software - is the
next step. Long before we look at windows, first verify what
critical hardware operates using tools such as computer (or
component manufacturer) comprehensive diagnostics.

Overheating is mostly a myth where one has not first learned
above diagnostic procedures. Did the system work in a 100
degree F room like all systems should? Then overheating at 70
degrees is an unlikely reason for failure. However heat is
another hardware diagnostic tool. Heat finds defective
hardware faster. The naive will instead install more fans
rather than replace the defective - and therefore heat
sensitive - component. One fan in the power supply should be
more than enough ventilation for a properly constructive
system that even works in a 100 degree F room.

Start by first learning what is good. That means specific
facts and numbers from things like the 3.5 digit multimeter
and comprehensive hardware diagnostics. Break the problem
down into parts. The power supply 'system' is confirmed.
Then move on to essential computer hardware. One who tried to
diagnosis all everything by using the OS is simply
complicating the problem - spinning wheels - confusing himself
- and provide few useful facts for others here who have better
knowledge and experience.
 
w_tom said:
Without numbers, then you have no idea if a power supply is
good. A bad power supply will still illuminate lights and
spin disk drives. Meanwhile the power supply is sending the
computer a signal that it is defective. How do human eyes see
that signal? You don't without a 3.5 digit multimeter.

Start the diagnosis by establishing what is good and what is
unknown. Since everything else in the system can appear to be
bad when a marginal power supply causes problems, then the
power supply 'system' is where one starts to verify what is
good. That means two minutes and a meter.

Once power supply 'system' (which is more than just a power
supply) is confirmed good, then move on to then next 'usual'
suspect - computer hardware. This is where comprehensive
hardware diagnostics - no operating system software - is the
next step. Long before we look at windows, first verify what
critical hardware operates using tools such as computer (or
component manufacturer) comprehensive diagnostics.

Overheating is mostly a myth where one has not first learned
above diagnostic procedures. Did the system work in a 100
degree F room like all systems should? Then overheating at 70
degrees is an unlikely reason for failure. However heat is
another hardware diagnostic tool. Heat finds defective
hardware faster. The naive will instead install more fans
rather than replace the defective - and therefore heat
sensitive - component. One fan in the power supply should be
more than enough ventilation for a properly constructive
system that even works in a 100 degree F room.
Click to expand...

Overheating is Not a myth. If the CPU fan fails, or the heat sink is
clogged with debris, or the heat sink clamp fails than the computer Will
overheat regardless of everything else. The OP should first check the
CPU fan and heat sink to make sure that they are installed correctly and
working properly. Then he can worry about trying to determine if there
is some other part failing. The visual check takes only a second and if
the computer was working for years it is very possible that the CPU is
Overheating.
 
The system does not even startup. Even a system suffering
from overheating problems will startup. Furthermore, if the
processor is Intel, the system would still keep working -
albeit slower. But again, this is not the OP's symptoms. The
OP said:
Suddenly, today it does not start any more at all (no disk
spinning etc) and an ugly voice repeats (see above).
Click to expand...

Where does overheating have anything to do with this?

Again, demonstrated is how others are so quick to blame
overheating rather than break down the problem, collect facts,
and step through the problem. Too many will jump at only what
they know rather than learn how to 'first analysis, second
fix' a problem.
 
w_tom said:
The system does not even startup. Even a system suffering
from overheating problems will startup. Furthermore, if the
processor is Intel, the system would still keep working -
albeit slower. But again, this is not the OP's symptoms. The
OP said:
Click to expand...

Since I have personal experience to refute your Claim I know that you
are wrong. My last build used the new LGA775 mount. The first time I
tried to install it in the very cramped space I did not get all of the
posts properly fastened it would start a couple of times and then
refused to start at all. The only indicator was that the power LED on
the motherboard still worked. Installing the heat sink again fixed the
problem. If it is overheating the CPU Will shut down after it gets too
hot, the throttling will only work until the CPU gets past the shut down
temperature.
 
Hi Folks,
I have now "stepped through the problem", first removed the CPU fan
and heat sink, which was clogged with a nice cushion of dust, but then
I tested the fan separately with a 12V source and found it spinned OK,
so it had to be its power supply and indeed, I only measured 2V
instead of 12. Then I measured a few other 12V connectors and found
they also had only 2V on them, so it had to be the power supply itself
and I removed it, unplugging all the connectors and put it again on
the mains just to notice - a bit late, but your hindsight is always
20/20 - that the fan local to the power supply box (an ATX power
supply with an extra 12V line for the CPU) did not spin, I should have
noticed that rightaway. My worry is that maybe there is some short
circuit downstream on the motherboard or one of the peripherals which
may have caused this breakdown. On the other hand, the power supply
contained a nice cushion of dust and since it is an electronic type
(no big transformator), it may have caused overheating of one of the
components, some of which do heat up quite a bit. As a matter of fact,
I don't have a schematic of that power supply and I did not find any
component that was obviously burned - all the power diodes measured OK
too, so, tomorrow, I'll just buy a new power supply and give you some
news. (Of course, if there's a short downstream, that one will die
too....)
Cheers!
KS
 
Back
Top