Pc acting weird during boot up

  • Thread starter Thread starter spuddy0108
  • Start date Start date
S

spuddy0108

My neighbours pc is acting very weird.

When pressing the powerbutton the machine starts up and so does the
monitor, but after only 5seconds the monitor goes off again and then
on again for a few seconds then off again and so on. The pc doesnt
shut down however as the green led is still on and the cpu fan is
going. Just cant seem to get past post mode and when it finaly does
get past post mode and windows is booting, the monitor goes off again
and im right at the beginning, in post mode again. Very odd. Tried
another monitor but same problems so its not that. Checked the cpu and
cleaned the fan but still no luck. Its a cheap system with onboard
everything lol. Celeron 2.6 cpu and sis chipset. Hope someone could
help
 
My neighbours pc is acting very weird.

When pressing the powerbutton the machine starts up and so does the
monitor, but after only 5seconds the monitor goes off again and then
on again for a few seconds then off again and so on. The pc doesnt
shut down however as the green led is still on and the cpu fan is
going. Just cant seem to get past post mode and when it finaly does
get past post mode and windows is booting, the monitor goes off again
and im right at the beginning, in post mode again. Very odd. Tried
another monitor but same problems so its not that. Checked the cpu and
cleaned the fan but still no luck. Its a cheap system with onboard
everything lol. Celeron 2.6 cpu and sis chipset. Hope someone could
help
Click to expand...

Do you have a spare power supply to test with ?

Some prebuilt machines have smaller power supplies in them, and
that can make it hard to find a repair substitute. If the supply
is a standard ATX one, then finding something to fit in the same
space will be easier.

Also, keep in mind, that some supplies don't adhere to standards,
Dell, over the years, had model years with non-standard supplies.
The Dell web site has pinout information for a number of their
designs, so you can actually check the pinout for non-compliance.

So while, for the most part, power supplies are swappable, there
are cases of non-compliance.

Paul
 
Paul said:
Do you have a spare power supply to test with ?

Some prebuilt machines have smaller power supplies in them, and
that can make it hard to find a repair substitute. If the supply
is a standard ATX one, then finding something to fit in the same
space will be easier.

Also, keep in mind, that some supplies don't adhere to standards,
Dell, over the years, had model years with non-standard supplies.
The Dell web site has pinout information for a number of their
designs, so you can actually check the pinout for non-compliance.

So while, for the most part, power supplies are swappable, there
are cases of non-compliance.
Click to expand...



That definately sounds like a bad power supply and the ones on Emachines are
standard...
so most any ATX supply that has enough power...will work
 
Thank you both for your input. I will remove a psu from an older pc
see what it gives. Ill come back with news once its done. Cheers
 
I bought a new psu yesterday because the ones i had lying arround were
old ones that did not have a 4pin connector for the cpu. Anyway i
placed the new psu but the problem remains.
 
I bought a new psu yesterday because the ones i had lying arround were
old ones that did not have a 4pin connector for the cpu. Anyway i
placed the new psu but the problem remains.
Click to expand...

So the CPU fan doesn't stop at all ?

I thought the description matched an oscillating power supply. But
I missed the part where you said the fan was still spinning.

You've already checked the CPU, and if the CPU was overheating,
that could cause the power to go off. But a good design would
make the machine stay off (at least, until you flipped the power
switch on the back, off and then on again).

The next thing I'd inspect is the reset and power switches.
Are the switches in good shape, or are they binding ? Both the
reset and the power switches are momentary contact types. They
only conduct when the button is pushed. They should stop conducting
when you release them. In an emergency (assuming the header has the same
pin spacing), you can substitute the reset switch in place of the power
switch. (Then you use the reset button to turn the computer on and off.)
I have also used a slot head screwdriver tip, between the power
switch terminals on the PANEL header, to give a momentary contact to
start a computer. But I generally do that while the motherboard is out
of the case, as then it is easier to see what I'm doing, and touch
only the two pins where the power switch would normally go. If the
motherboard is inside the case, it is pretty hard to see what you're
doing, so the screwdriver idea wouldn't be wise then.

Other than that, if the thing seems to be oscillating in reset-POST
cycles, it could be some kind of motherboard problem. But at least
have a look at the switches first.

You say the green LED stays on, and you might also want to observe the
green LED during one of the blank-out cycles. See if the green LED winks
out, even for a moment. If it did, I'd get out my clamp-on DC ammeter, and
check how much current is being drawn from +5VSB. But if +5VSB was being
overloaded, the thing should shut off and stay off. Which really
leaves a power switch or a reset switch problem, and barring that,
a motherboard problem.

The five second number, sort of sounds like the amount of time you have
to hold in a power switch, to make a machine shut off. Like the power
switch was still making contact.

Does any text appear on the monitor, during the first five seconds ?
Does the reset happen just as the text reaches a certain point ?

Also, another test you can try. Power off and unplug the machine.
Remove the RAM and store it in an antistatic bag. Turn on the machine.
The machine should beep a "missing RAM" pattern. With the RAM
removed, will the machine stay running ? You can do a similar test
with the video card (assuming the motherboard is not using integrated
graphics). If the responses of the system change, with components
unplugged, that might also hint at the nature of the problem.

Paul
 
The prob is solved and guess what it was... the bloody powerbutton.
Anyway thanks for your input Paul, i realy appreciate it.
 
Back
Top