PC power on - start, stop, start, stop, etc - eventually powers up

[spodosaurus]

Hi all,

I have a problem that has only recently started (to my knowledge) and I
can't seem to solve it. I've got a P4 3.0GHz PC with an Asis P4P800S-X
board, 512mb RAM, Thermaltake 430W PSU and geforece fx5200 graphics. The
problem is that once the system is warm (and sometimes even from cold) a
shutdown then start gets the computer spinning its fans, powering off,
spinning fans, powering off, and repeating this cycle from 1-10+ times
before actually POSTing. Once it POSTs, it's all fine. Sometimes the
brief power ons result in the drives starting to power up as well, but
usually not. I've flashed the motherboard, replaced the PSU, cleaned and
reattached the HSF, and tried it with only CPU, GPU, RAM, and Video
attached to the motherboard without success in solving this issue. I've
come accross other accounts of similar issues, but they've usually been
solved by the steps I've already taken. Any ideas on where I could go
from here?

TIA,

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/

 

[Ed Medlin]

Hi all,

I have a problem that has only recently started (to my knowledge) and I
can't seem to solve it. I've got a P4 3.0GHz PC with an Asis P4P800S-X
board, 512mb RAM, Thermaltake 430W PSU and geforece fx5200 graphics. The
problem is that once the system is warm (and sometimes even from cold) a
shutdown then start gets the computer spinning its fans, powering off,
spinning fans, powering off, and repeating this cycle from 1-10+ times
before actually POSTing. Once it POSTs, it's all fine. Sometimes the brief
power ons result in the drives starting to power up as well, but usually
not. I've flashed the motherboard, replaced the PSU, cleaned and
reattached the HSF, and tried it with only CPU, GPU, RAM, and Video
attached to the motherboard without success in solving this issue. I've
come accross other accounts of similar issues, but they've usually been
solved by the steps I've already taken. Any ideas on where I could go from
here?

TIA,

Ari

--

Fans spinning up and system not posting is a classic symtom of a failed or
failing power supply. If you are sure your replacement PSU is good, it could
be a number of things including CPU and MB.


Ed

 

[Bryce]

Hi all,

I have a problem that has only recently started (to my knowledge) and I
can't seem to solve it. I've got a P4 3.0GHz PC with an Asis P4P800S-X
board, 512mb RAM, Thermaltake 430W PSU and geforece fx5200 graphics. The
problem is that once the system is warm (and sometimes even from cold) a
shutdown then start gets the computer spinning its fans, powering off,
spinning fans, powering off, and repeating this cycle from 1-10+ times
before actually POSTing. Once it POSTs, it's all fine. Sometimes the
brief power ons result in the drives starting to power up as well, but
usually not. I've flashed the motherboard, replaced the PSU, cleaned and
reattached the HSF, and tried it with only CPU, GPU, RAM, and Video
attached to the motherboard without success in solving this issue. I've
come accross other accounts of similar issues, but they've usually been
solved by the steps I've already taken. Any ideas on where I could go
from here?

TIA,

Ari

Possibly marginal available power at boot? You didn't say how many hard
drives are connected: try with only one. Also try without the video card
installed. Watching the power supply voltages during the boot might help.

 

[spodosaurus]

Possibly marginal available power at boot?

Doesn't even make it through post, much less all the way to boot.
However, sometimes, like right now, it went through POST and boot and is
working fine - if I restart it, though...

You didn't say how many hard
drives are connected: try with only one.

I said above that I've tried it with the bare minimum: motherboard with
only RAM, CPU, and video card. There's only one HDD attached when in
use, however.

Also try without the video card
installed. Watching the power supply voltages during the boot might help.

I guess I can always give that a try. Without the graphics card it'll
just beep.


--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/

 

[spodosaurus]

Fans spinning up and system not posting is a classic symtom of a failed
or failing power supply. If you are sure your replacement PSU is good,
it could be a number of things including CPU and MB.


Ed

At this point I'm guessing it has to be motherboard. The install and
updating of ubuntu went perfectly. it still runs perfectly - when it
actually gets to POST.

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/

 

[westom1]

I have a problem that has only recently started (to my knowledge) and I
can't seem to solve it. ... once the system is warm (and sometimes
even from cold) a shutdown then start gets the compute spinning its
fans, powering off, spinning fans, powering off, and repeating this
cycle from 1-10+ times before actually POSTing. Once it POSTs, it's
all fine.

Fans spinning up and system not posting is a classic symtom of a
failed or
failing power supply. Also a failing power supply controller. Also a
video controller problem, etc. Most things replaced and swapped
would not address those symptoms and are classicshotgunning. In some
cases, if it fixed failure, it really only cured symptoms.

First establish what is and is not working. That means numbers and
simpler facts. Havinig swapped a power supply, nobody still knows if
either power supply is good or bad. Swapping determined nothing
useful. A perfectly good supply can be defective in another machine.
And a perfectly defective power supply can boot another computer.
Without numbers from a supply when powered off and as powering on,
then nobody knows if the power supply is good. With VDC numbers from
a 3.5 digit multimter, then a power supply's state is known AND in
seconds, also known is other power supply 'system' components such as
the supply controller and power switching.

By wildly speculating, a most likely reason for failure is a power
supply controller. Without numbers, every answer will only be
speculation.

Measure VDC on the green, gray, and purple wires (power supply to
motherboard) both before and when switch is pressed. Record those
numbers. Once the system does boot, measure VDC numbers from any one
of red, orange, and yellow wires. Only with numbers can we eliminate
power supply, its controller and other power supply system components
from the list of suspects. Only then do we have something on a list of
accomplishments. Only then move on to other potential problems. Only
then has labor provided numbers so that others can reply with useful
assistance.

Most items 'fixed' may have been eliminated immediately in the
minute a multimeter was learning numbers. Get the mulitmeter so that
a next reply is informative.

 

[westom1]

Possibly marginal available power at boot?  You didn't say how many hard
drives are connected: try with only one.

If one or four hard drives made a difference, then the power supply
is grossly undersized even with zero drives. And the multimeter would
have made that obvious even with zero drives connected.

But again, 'it might be hard drives' is another perfect example of
shotgunning - this time without basic knowledge of how computer power
supplies work. Disk drives do not even make a list of responsible
speculation - which would have been obvious had that reply come with
some numbers such as the power consumption by a disk drive. No
numbers; therefore wild and not useful speculation.

 

[Franc Zabkar]

Hi all,

I have a problem that has only recently started (to my knowledge) and I
can't seem to solve it. I've got a P4 3.0GHz PC with an Asis P4P800S-X
board, 512mb RAM, Thermaltake 430W PSU and geforece fx5200 graphics. The
problem is that once the system is warm (and sometimes even from cold) a
shutdown then start gets the computer spinning its fans, powering off,
spinning fans, powering off, and repeating this cycle from 1-10+ times
before actually POSTing. Once it POSTs, it's all fine. Sometimes the
brief power ons result in the drives starting to power up as well, but
usually not. I've flashed the motherboard, replaced the PSU, cleaned and
reattached the HSF, and tried it with only CPU, GPU, RAM, and Video
attached to the motherboard without success in solving this issue. I've
come accross other accounts of similar issues, but they've usually been
solved by the steps I've already taken. Any ideas on where I could go
from here?

TIA,

Ari

It may be instructive if you can determine whether it is the
motherboard that is turning the PSU off in response to a fault
condition, or whether it is the PSU that is turning itself off in
response to a fault in the load. Or it may be that the motherboard is
resetting itself for some reason without turning off the PSU. To this
end it *might* help if you monitored the PS_ON signal (pin 14, green)
with a multimeter. A DMM may not detect rapid transitions on this pin,
though, even with a peak hold function.

See http://pinouts.ru/Power/atxpower_pinout.shtml

Does the PSU fan stop spinning, or is it only the motherboard fans
that do this?

Have you tried replacing your graphics card with a low power PCI card?

- Franc Zabkar

 

[spodosaurus]

Fans spinning up and system not posting is a classic symtom of a
failed or
failing power supply. Also a failing power supply controller. Also a
video controller problem, etc. Most things replaced and swapped
would not address those symptoms and are classicshotgunning. In some
cases, if it fixed failure, it really only cured symptoms.

First establish what is and is not working. That means numbers and
simpler facts. Havinig swapped a power supply, nobody still knows if
either power supply is good or bad. Swapping determined nothing
useful. A perfectly good supply can be defective in another machine.
And a perfectly defective power supply can boot another computer.
Without numbers from a supply when powered off and as powering on,
then nobody knows if the power supply is good. With VDC numbers from
a 3.5 digit multimter, then a power supply's state is known AND in
seconds, also known is other power supply 'system' components such as
the supply controller and power switching.

By wildly speculating, a most likely reason for failure is a power
supply controller. Without numbers, every answer will only be
speculation.

Measure VDC on the green, gray, and purple wires (power supply to
motherboard) both before and when switch is pressed. Record those
numbers. Once the system does boot, measure VDC numbers from any one
of red, orange, and yellow wires. Only with numbers can we eliminate
power supply, its controller and other power supply system components
from the list of suspects. Only then do we have something on a list of
accomplishments. Only then move on to other potential problems. Only
then has labor provided numbers so that others can reply with useful
assistance.

Most items 'fixed' may have been eliminated immediately in the
minute a multimeter was learning numbers. Get the mulitmeter so that
a next reply is informative.

Hi,

Thanks for the reply.

My voltages were all within 5% of nominal voltages on both PSUs - I
tested these before I did anything else. I had a PSU test okay with my
digital multimeter before, but there were fluctuations on the +5VSB line
that were causing problems that were only detected by a PSU tester that
was inaccurate and only reported to one decimal place, but it did update
frequently enough that I could see the +5VSB fluctuations. Just in case
there was a +5VSB issue, I switch the rear usb power from that line to
+5V using the motherboard's jumpers without success.

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/

 

[spodosaurus]

It may be instructive if you can determine whether it is the
motherboard that is turning the PSU off in response to a fault
condition, or whether it is the PSU that is turning itself off in
response to a fault in the load. Or it may be that the motherboard is
resetting itself for some reason without turning off the PSU. To this
end it *might* help if you monitored the PS_ON signal (pin 14, green)
with a multimeter. A DMM may not detect rapid transitions on this pin,
though, even with a peak hold function.

See http://pinouts.ru/Power/atxpower_pinout.shtml

Does the PSU fan stop spinning, or is it only the motherboard fans
that do this?

PSU, CPU (connected to motherboard), and case fan (connected to molex
from PSU).

Have you tried replacing your graphics card with a low power PCI card?

- Franc Zabkar

This last suggestion got me thinking. I didn't recall this PC having
these problems before it was given to me. I thought it was using a
different graphics card and that I'd put the FX5200 in it (turns out the
actual card was the original geforce mx4000 after all). I tried putting
in a higher powered Radeon 9600 Pro card in there and turning it on from
cold - no joy, even though it always starts when it's at room
temperature after a few hours idle. So I popped in an old PCI card - and
it started immediately even though it had been running.

I'll update this if it starts having problems again in a few hours when
I restart it after it's been running a while.

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/

 

[westom1]

My voltages were all within 5% of nominal voltages on both PSUs - I
tested these before I did anything else. I had a PSU test okay with my
digital multimeter before, but there were fluctuations on the +5VSB line
that were causing problems that were only detected by a PSU tester that
was inaccurate and only reported to one decimal place, but it did update
frequently enough that I could see the +5VSB fluctuations. Just in case
there was a +5VSB issue, I switch the rear usb power from that line to
+5V using the motherboard's jumpers without success.

Your specs -5% - are an ATX spec but are not correct for what is
measured. Post those numbers here because being inside another spec
is only one part of information in those numbers.

If a power supply test (near zero load) detected a problem, then
removing a near zero load (USB) also accomplishes nothing. When the
+5VSB is connected to a load (motherboard), what does the multimeter
actually read?