I ran into the following issue while upgrading an older PC that finally
failed me. I could not find any solutions online but after 2 days of
research, I finally found the solution and thought I would offer it
(and circumstances behind my issue) for those who might encounter the
same issue.
My Upgrade:
New Gigabyte S-Series Mobo (GA-EP35-DS3L)
New Intel Core 2 Duo CPU 3GHz
New 4GB G-Skill RAM
New 650GB WD SATA HDD
New GeForce 8800GT Graphics Card
The only things kept original where the two DVD-Rs, 160GB WD SATA HDD,
CoolerMaster ATX case and 500watt Norwood Micro Power Supply Unit
(PSU).
During the initial install of the mobo, cpu, RAM and graphics card,
there were no issues. No issues after installing Win XP Pro, SP3. I
needed to get data off two other HDDs. At some point, and I'm not sure
exactly where, but when I would shut down, the PC would shut down and
the screen would turn off. But within a second, the power light, mobo
light and fans came on. But the PC did not reboot. The only way to
turn off the PC was with the power switch on the PSU. I was thinking
this could be an issue with my ACPI power mgmt. But I did not make any
changes so that perplexed me. I reset the CMOS to factory settings.
That did not resolve the issue. I noticed 5 lights on my mobo were lit
upon shutdown. According to Gigabyte manual, it says my CPU is
overheating. This makes no sense since the CPU monitor safety is set
to shut down before overheating and the CPU monitor displays a temp of
30c.
I searched for days on "XP won't completely shutdown". I now realize
that I was chasing a symptom and not the problem. Thinking the PC
might be overheating, I shut if off for hours (power switch on the
PSU). When I turned the power back on, the power light, fans, etc. came
back on. That got me thinking: At this point, the OS is not even
running. How can this be the cause of my problem. It must be with the
mobo. Man, I hope my new mobo did not die. Then I read one last volley
of "my system won't shut down" and the very last post said he had a
similar issue and it was the PSU. I decided to pull another PSU out of
a working PC and give it a try. Sure enough, it worked fine. So, long
story short, my PSU crapped out on me. What a PITA. Hopefully, none of
you will have to experience this issue.
--
UCLAN - 28 Jul 2008 19:01 GMT
> I ran into the following issue while upgrading an older PC that finally
> failed me. I could not find any solutions online but after 2 days of
[quoted text clipped - 40 lines]
> story short, my PSU crapped out on me. What a PITA. Hopefully, none of
> you will have to experience this issue.
Chances are that the +5v standby circuit failed, and the voltage could not
maintain a Logic 1 level. A simple check of this with a DVM at the start
might have eliminated your nightmare completely.
Thomas Chatfield - 28 Jul 2008 19:57 GMT
> > I ran into the following issue while upgrading an older PC that
> > finally failed me. I could not find any solutions online but after
[quoted text clipped - 45 lines]
> could not maintain a Logic 1 level. A simple check of this with a DVM
> at the start might have eliminated your nightmare completely.
I'm not sure I would have (or many others) figured that out in the
beginning. With the thousands of messages I poured over, none looked in
that direction. At any rate, how can this be checked? I have a digital
multimeter. Would that be the Soft (on/off) or Standby on the main
power connector?
--
Thomas Chatfield - 28 Jul 2008 20:05 GMT
> > > I ran into the following issue while upgrading an older PC that
> > > finally failed me. I could not find any solutions online but
[quoted text clipped - 52 lines]
> digital multimeter. Would that be the Soft (on/off) or Standby on the
> main power connector?
Sorry, I just saw in your message header - Standby. I'm going to check.
Although, I already have a new PSU on the way.
--
kony - 28 Jul 2008 22:14 GMT
>> Chances are that the +5v standby circuit failed, and the voltage
>> could not maintain a Logic 1 level. A simple check of this with a DVM
[quoted text clipped - 5 lines]
>multimeter. Would that be the Soft (on/off) or Standby on the main
>power connector?
A multimeter won't always find a fault or inadequacy in the
5VSB circuit, sometimes an oscope would be needed.
Basically, the seemingly poor quality of the generic power
supply is the most likely problem, it wore out too fast
because of that or certain internal component tolerances
were exceeded to cause some failure though to an extent that
is the same as wearing out for the intended application.
The PSU ought to be replaced even if it isn't the direct
cause of the immediate problem.
Hopefully you have replaced it with a major branded, known
good quality 400W or higher model. Given the gaming video
card and multiple hard drives, a 500W+ model would be more
appropriate while the original generic was probably worth
closer to 300W, even less sustainably, regardless of it's
optimistic labeling.
Thomas Chatfield - 30 Jul 2008 12:37 GMT
> >> Chances are that the +5v standby circuit failed, and the voltage
> >> could not maintain a Logic 1 level. A simple check of this with a
[quoted text clipped - 24 lines]
> closer to 300W, even less sustainably, regardless of it's
> optimistic labeling.
Thanks! I have an Antec 500w PSU (w/2 12v rails) on the way.
--