It's the power supply...or is it? >:)

I had this computer that won't turn on. No POST. No fans. Everything
was plugged in. I pushed the power button and its totally dead. The
only sign of life was the green light on the motherboard still lit. I
took a multimeter and did a continuity check on the wire going from
the on switch to the motherboard, and everything looked fine. So I
narrowed it down to a problem with the motherboard or a problem with
the power supply.

Luckily I had another power supply hanging around. I did a quick swap,
and to my amazement, the computer turned on fine. I thought I had
fixed it.

Well an hour later I come back to the computer room. The computer had
shut itself off while I was away. Now I have the exact same problem as
I did before. I push the power button and nothing happens. No POST. No
fans. The green light on the motherboard is still lit.

I'm just a little unclear on what this means. If a new power supply
didn't fix the problem, I'd assume the mother board was just bad. But
how come replacing the power supply got the computer to come back to
life (if only temporary). What is your opinion? Is this a power supply
problem, or a motherboard problem?

Also, does anyone really know what that little green light on the
motherboard is for? I've always wondered.

Thank you in advance to anyone who answers.

Previously Norm Loman <> wrote:
> I had this computer that won't turn on. No POST. No fans. Everything
> was plugged in. I pushed the power button and its totally dead. The
> only sign of life was the green light on the motherboard still lit. I
> took a multimeter and did a continuity check on the wire going from
> the on switch to the motherboard, and everything looked fine. So I
> narrowed it down to a problem with the motherboard or a problem with
> the power supply.

> Luckily I had another power supply hanging around. I did a quick swap,
> and to my amazement, the computer turned on fine. I thought I had
> fixed it.

> Well an hour later I come back to the computer room. The computer had
> shut itself off while I was away. Now I have the exact same problem as
> I did before. I push the power button and nothing happens. No POST. No
> fans. The green light on the motherboard is still lit.

> I'm just a little unclear on what this means. If a new power supply
> didn't fix the problem, I'd assume the mother board was just bad. But
> how come replacing the power supply got the computer to come back to
> life (if only temporary). What is your opinion? Is this a power supply
> problem, or a motherboard problem?

Well, it may still be both. If you know the replacement PSU is good,
then it is the mainboard. However if the replacement is an ElCheapo
thing (that you should not use anyways), a DOA or near-DOA PSU is
a real possibility.

> Also, does anyone really know what that little green light on the
> motherboard is for? I've always wondered.

Typically showing that +5Vsb (that standby voltage) is present to
keep you from changiong components and the like, because you
may kill something that way. It basically says "power present, keep
away".

Arno

Norm Loman schrieb:

> I'm just a little unclear on what this means. If a new power supply
> didn't fix the problem, I'd assume the mother board was just bad. But
> how come replacing the power supply got the computer to come back to
> life (if only temporary). What is your opinion? Is this a power supply
> problem, or a motherboard problem?

I had a similar problem myself, where the reason was the insufficient
contact between the CPU and the heat sink, due to the broken heat sink
holder on the CPU socket. Some (newer) CPUs shut down themselves, as
soon as their internal temparature reaches some limit. This can happen
before even the BIOS has a chance to output anything on the screen, or
at a later time.

DoDi

Previously Hans-Peter Diettrich <> wrote:
> Norm Loman schrieb:

>> I'm just a little unclear on what this means. If a new power supply
>> didn't fix the problem, I'd assume the mother board was just bad. But
>> how come replacing the power supply got the computer to come back to
>> life (if only temporary). What is your opinion? Is this a power supply
>> problem, or a motherboard problem?

> I had a similar problem myself, where the reason was the insufficient
> contact between the CPU and the heat sink, due to the broken heat sink
> holder on the CPU socket. Some (newer) CPUs shut down themselves, as
> soon as their internal temparature reaches some limit. This can happen
> before even the BIOS has a chance to output anything on the screen, or
> at a later time.

This is a possibility and I agree that it can happen fast enough
to cause the symptoms described. However for the successfull
start observed, adequate cooling would be necessary. Basically
the only way to check this is to remove the mainboard and have a
very close look at the cooler. Maybe even remove it and remount
with new thermal compound.

Arno

On Feb 17, 12:31*am, Norm Loman <normlo...@gmail.com> wrote:
> Well an hour later I come back to the computer room. The computer had
> shut itself off while I was away. Now I have the exact same problem as
> I did before.

Nobody has answered because, well, in less than two minutes, numbers
from a 3.5 digit multimeter would answer everything the first time.

First, a power supply is only one component of the power supply
system. Is some part of that system defective? Again, disconnect
nothing. Take numbers. Have an answer in minutes.

Second, normal is a defective supply powering a computer. Too many
swap supplies, see a computer work, then assume the problem is
solved. You have simply demonstrated why part swapping is
problematic. A defective supply that boots a computer can only be
identified using the meter.

Critical are voltage numbers from the green, gray, and purple wire
(from power supply to motherboard) both before and when a power switch
is pressed. Those numbers obtained in 30 seconds would have answered
what others suggested in 'check this or check that'. 30 seconds
answers that and much more as long as parts are not swapped.

Purple wire (also called +5VSB) must be greater than 4.87 volts.
Green wire must be above 2.0 volts before switch press. Then fall
immediately to less than 0.7 volts when switch is pressed. Gray wire
must rise above 2.4 volts within two seconds of pressing the power
switch - and stay there. What do you have – an answer that must list
each number.

Those numbers say far more than you can imagine. Without numbers,
those who actually know this stuff will stay silent.

Some sometimes fear. It is only a 3.5 digit multimeter. That
sounds complex! Well, if you cannot use a multimeter, then you have
no business using a cell phone or Ipod. Both are far more complex and
expensive. A meter is sold even where hammers are sold for about the
same price. Even sold in Kmart and Wal-mart. Get the meter. Post
numbers. Have a useful answer (no more 'it could be this or check
that') in the very next reply.

Previously w_tom <> wrote:
> On Feb 17, 12:31?am, Norm Loman <normlo...@gmail.com> wrote:
>> Well an hour later I come back to the computer room. The computer had
>> shut itself off while I was away. Now I have the exact same problem as
>> I did before.

> Nobody has answered because, well, in less than two minutes, numbers
> from a 3.5 digit multimeter would answer everything the first time.

> First, a power supply is only one component of the power supply
> system. Is some part of that system defective? Again, disconnect
> nothing. Take numbers. Have an answer in minutes.

> Second, normal is a defective supply powering a computer. Too many
> swap supplies, see a computer work, then assume the problem is
> solved. You have simply demonstrated why part swapping is
> problematic. A defective supply that boots a computer can only be
> identified using the meter.

> Critical are voltage numbers from the green, gray, and purple wire
> (from power supply to motherboard) both before and when a power switch
> is pressed. Those numbers obtained in 30 seconds would have answered
> what others suggested in 'check this or check that'. 30 seconds
> answers that and much more as long as parts are not swapped.

> Purple wire (also called +5VSB) must be greater than 4.87 volts.
> Green wire must be above 2.0 volts before switch press. Then fall
> immediately to less than 0.7 volts when switch is pressed. Gray wire
> must rise above 2.4 volts within two seconds of pressing the power
> switch - and stay there. What do you have ? an answer that must list
> each number.

> Those numbers say far more than you can imagine. Without numbers,
> those who actually know this stuff will stay silent.

> Some sometimes fear. It is only a 3.5 digit multimeter. That
> sounds complex! Well, if you cannot use a multimeter, then you have
> no business using a cell phone or Ipod. Both are far more complex and
> expensive. A meter is sold even where hammers are sold for about the
> same price. Even sold in Kmart and Wal-mart. Get the meter. Post
> numbers. Have a useful answer (no more 'it could be this or check
> that') in the very next reply.


My Advice: Ignore this incompetent.

Arno

On Feb 17, 12:31*am, Norm Loman <normlo...@gmail.com> wrote:
> I had this computer that won't turn on. No POST. No fans. Everything
> was plugged in. I pushed the power button and its totally dead. The
> only sign of life was the green light on the motherboard still lit. I
> took a multimeter and did a continuity check on the wire going from
> the on switch to the motherboard, and everything looked fine. So I
> narrowed it down to a problem with the motherboard or a problem with
> the power supply.
>
> Luckily I had another power supply hanging around. I did a quick swap,
> and to my amazement, the computer turned on fine. I thought I had
> fixed it.
>
> Well an hour later I come back to the computer room. The computer had
> shut itself off while I was away. Now I have the exact same problem as
> I did before. I push the power button and nothing happens. No POST. No
> fans. The green light on the motherboard is still lit.
>
> I'm just a little unclear on what this means. If a new power supply
> didn't fix the problem, I'd assume the mother board was just bad. But
> how come replacing the power supply got the computer to come back to
> life (if only temporary). What is your opinion? Is this a power supply
> problem, or a motherboard problem?
>
> Also, does anyone really know what that little green light on the
> motherboard is for? I've always wondered.
>
> Thank you in advance to anyone who answers.

THank you everyone. I did check my replacement powersupply with a
multimeter, and everything seems to be fine. I even tried another
power supply. but I cant seem to get it to turn on. I'm going to
narrow it down a bit and say it's the motherboard, and see if a
replacement works.

Thank you all so much!