Possibly a rather long post here, but I'm hoping to capture everything I've
investigated so far.
Basically, my PC has recently begun switching off with no warning
whatsoever. I.e. fine...fine...GONE! To my logic, this spells an issue in
perhaps just one of two places...motherboard or PSU, because I have come to
the conclusion that I do not have any overheating issues, as I am running
Speedfan, and my logs show nothing out of the ordinary just before a sudden
power off. Typical temps are:
Hardware is MSI 975X Platinum Powerup Edition with Core2Duo E6600 @3GHz.
Been running overclocked for well over a year with no issues.
Tonight, I've dismantled everything in my PC, and vacuumed out everywhere I
could see, including the inside of the PSU. I've also re-seated the CPU, and
reapplied the HSF with new Arctic Silver Ceramique. It hasn't powered off
yet, but from my experience over the last few weeks, it's just a matter of
time. I'd like to think it won't happen...but I'm sure it will.
Anyway, when it *does* suddenly power off, I've noticed that the LED on the
Ethernet port (on motherboard) stays lit. So there must be *something* still
trickling into the board. Yet when I measure the +5V and +12V lines on some
of the mobo connectors (PCI-X and HDD), I see that +5V and +12V are sitting
at 0.18V.
I cannot switch the computer back on again after sudden power off using the
power button. Instead, I have to disconnect the PSU, until I hear a quiet
noise which sounds like a capacitor discharging. When that happens, the
Ethernet LED dies, and I am then able to switch the PC back on again.
OK...sorry if this has rambled a bit...does this mean anything to anyone?
Does it offer any insight as to what could be wrong?
"ShadowTek" <psistormyamato@cs.com> wrote in message
news:eac9a152-b511-4d1a-9ede-a2cd5c88dfbf@a1g2000hsb.googlegroups.com...
> Can you disable the onboard ethernet in BIOs?
>
> If so, then you could test for a while without it.
I can disable it, but not really sure what this will prove.
"Paul" <nospam@needed.com> wrote in message news:frcf5d$2sg$1@aioe.org...
> John Whitworth wrote:
>
> Just for fun, what brand and model of supply is this ? I'm willing
> to bet, a Google search with the brand and model, will likely
> turn up more experiences similar to your own. These things don't
> happen out of the blue, and you'll likely find more cases of
> the same problem.
>
Hi Paul,
Thanks for your reply. Extremely interesting and informative, though got to
go to bed in a minute, so will have to study it more tomorrow. The PSU is a
Hiper PSU 580W - http://www.hipergroup.com/English/pr...hpu-4k580.html.
I've put in a 8800GTX recently, though these issues didn't start happening
until a couple of weeks after that went in, and there have been no other
hardware modifications.
Anyway, I'll study your reply somewhat more tomorrow. Thanks again!
John Whitworth wrote:
> Hi All,
>
> Possibly a rather long post here, but I'm hoping to capture everything
> I've investigated so far.
>
> Basically, my PC has recently begun switching off with no warning
> whatsoever. I.e. fine...fine...GONE! To my logic, this spells an issue
> in perhaps just one of two places...motherboard or PSU, because I have
> come to the conclusion that I do not have any overheating issues, as I
> am running Speedfan, and my logs show nothing out of the ordinary just
> before a sudden power off. Typical temps are:
>
> - System - 43 'C
> - Aux - 36 'C
> - Core 0 - 27 'C
> - Core 1 - 29 'C
>
> Hardware is MSI 975X Platinum Powerup Edition with Core2Duo E6600 @3GHz.
> Been running overclocked for well over a year with no issues.
>
> Tonight, I've dismantled everything in my PC, and vacuumed out
> everywhere I could see, including the inside of the PSU. I've also
> re-seated the CPU, and reapplied the HSF with new Arctic Silver
> Ceramique. It hasn't powered off yet, but from my experience over the
> last few weeks, it's just a matter of time. I'd like to think it won't
> happen...but I'm sure it will.
>
> Anyway, when it *does* suddenly power off, I've noticed that the LED on
> the Ethernet port (on motherboard) stays lit. So there must be
> *something* still trickling into the board. Yet when I measure the +5V
> and +12V lines on some of the mobo connectors (PCI-X and HDD), I see
> that +5V and +12V are sitting at 0.18V.
>
> I cannot switch the computer back on again after sudden power off using
> the power button. Instead, I have to disconnect the PSU, until I hear a
> quiet noise which sounds like a capacitor discharging. When that
> happens, the Ethernet LED dies, and I am then able to switch the PC back
> on again.
>
> OK...sorry if this has rambled a bit...does this mean anything to
> anyone? Does it offer any insight as to what could be wrong?
>
> Cheers
>
> JW
Those are interesting symptoms. Especially the non-zero value on +5V and
+12V, when the thing turns off. If the main switching part stops, the
result should be zero volts on 3.3/5V/12V etc. The 0.18V implies the
main switcher in the PSU is still running, a little bit.
There are two possible places for latch-off behavior in a current
computer. The Vcore converter next to the CPU, can latch off, if it
detects a fault. In that case, your fans would still be spinning.
You could remove power to the +12V, by pushing the front power burron,
or use the rear switch. A circuit that latches off, clears the event,
when power is removed to the circuit.
The power supply may contain a similar latch-off feature. The supply actually
consists of two pieces. The main supply, and the +5VSB supply. When
you see the Ethernet LEDs lit, in fact those are powered by +5VSB.
That is why, even though your measured +5V and +12V were 0.18V
(not enough voltage to light a LED), the LEDs could still be lit.
Certain chips in the computer, are powered by +5VSB, otherwise
known as the "standby voltage". Any chip wishing to be used for
waking the computer, needs to be powered by +5VSB, when the computer
is sleeping. Since many Ethernet chips support "Wake On LAN", they
need +5VSB to do the waking function.
So your symptoms are, the standby rail is still operating (thus the
Ethernet LEDs). But the main part of the supply has shut off. And
the need to switch off at the back, means the supply has "latched" a
fault indication.
I'd start by trying a different supply in it. If the supply you
bought, is one known to suffer from that problem, then certainly
buying a different brand would be a good idea.
Just for fun, what brand and model of supply is this ? I'm willing
to bet, a Google search with the brand and model, will likely
turn up more experiences similar to your own. These things don't
happen out of the blue, and you'll likely find more cases of
the same problem.
'~misfit~' wrote:
> I would love to know what a capacitor discharging sounds like.
_____
It's a high-pitched whimper of despair.
The original poster may be thinking of the sound that can be made by the
fly-back high voltage
circuit 15 KHz frequency Dopplering down as the horizontal drive dies when a
large CRT device is shut off.
That sound plus the bit of information the original poster added in his
second post (just installed a 8800 GTX) is a good indication that the power
supply is whimpering as it shuts down due to overload (probably from over
heating.)
Phil Weldon
"~misfit~" <misfit61nz@yahoot.com.au> wrote in message
news:47da632f$1@news2.actrix.gen.nz...
> Somewhere on teh intarweb "John Whitworth" typed:
>
>
> Sorry, I am unable to offer much advice with the data set given. However:
>
>> I hear a quiet noise which sounds like a capacitor discharging.
>
> I would love to know what a capacitor discharging sounds like.
>
> Rgds,
> --
> Shaun.
>
"Phil Weldon" <notdisclosed@example.com> wrote in message
news:13tldim83v92u88@corp.supernews.com...
> '~misfit~' wrote:
>> I would love to know what a capacitor discharging sounds like.
> _____
>
> It's a high-pitched whimper of despair.
>
> The original poster may be thinking of the sound that can be made by the
> fly-back high voltage
> circuit 15 KHz frequency Dopplering down as the horizontal drive dies when
> a large CRT device is shut off.
Yes...that'll be the one!
> That sound plus the bit of information the original poster added in his
> second post (just installed a 8800 GTX) is a good indication that the
> power
> supply is whimpering as it shuts down due to overload (probably from over
> heating.)
When I've used calculators to work out how much power is really required, it
always falls way short of my 580W though. http://extreme.outervision.com/psucalculatorlite.jsp. I guess it is looking
like the PSU is the culprit though, so a replacement is on the cards.
'John Whitworth' wrote:
> When I've used calculators to work out how much power is really required,
> it always falls way short of my 580W though.
> http://extreme.outervision.com/psucalculatorlite.jsp. I guess it is
> looking like the PSU is the culprit though, so a replacement is on the
> cards.
After connectors, connections, and fans I think PC power supplies are the
most likely hardware failure points.
I my experience, the failure point rankings have been more or less the same
for the last 40 years or so
1. operator error
2. application program error
3. operating system error
4. connectors and connections and fans
5. power supply
6. hard drives
6b. (new entry) graphics adapters
7. motherboard/pc boards/modules
8. CPU
9. memory
Phil Weldon
"John Whitworth" <sexyjw@g_EEEEEEEEEEEEEEEEE_mail.com> wrote in message
news:47daf1bf$0$32045$da0feed9@news.zen.co.uk...
>
> "Phil Weldon" <notdisclosed@example.com> wrote in message
> news:13tldim83v92u88@corp.supernews.com...
>> '~misfit~' wrote:
>>> I would love to know what a capacitor discharging sounds like.
>> _____
>>
>> It's a high-pitched whimper of despair.
>>
>> The original poster may be thinking of the sound that can be made by the
>> fly-back high voltage
>> circuit 15 KHz frequency Dopplering down as the horizontal drive dies
>> when a large CRT device is shut off.
>
> Yes...that'll be the one!
>
>> That sound plus the bit of information the original poster added in his
>> second post (just installed a 8800 GTX) is a good indication that the
>> power
>> supply is whimpering as it shuts down due to overload (probably from over
>> heating.)
>
> When I've used calculators to work out how much power is really required,
> it always falls way short of my 580W though.
> http://extreme.outervision.com/psucalculatorlite.jsp. I guess it is
> looking like the PSU is the culprit though, so a replacement is on the
> cards.
>
> Cheers
>
> JW
Somewhere on teh intarweb "Phil Weldon" typed:
> '~misfit~' wrote:
>> I would love to know what a capacitor discharging sounds like.
> _____
>
> It's a high-pitched whimper of despair.
Heh! I should have guessed. <g>
> The original poster may be thinking of the sound that can be made by
> the fly-back high voltage
> circuit 15 KHz frequency Dopplering down as the horizontal drive dies
> when a large CRT device is shut off.
>
> That sound plus the bit of information the original poster added in
> his second post (just installed a 8800 GTX) is a good indication that
> the power supply is whimpering as it shuts down due to overload
> (probably from over heating.)
Good sluething there Phil. With both bits of info on hand I concur.
--
Shaun.
> Phil Weldon
>
>
> "~misfit~" <misfit61nz@yahoot.com.au> wrote in message
> news:47da632f$1@news2.actrix.gen.nz...
>> Somewhere on teh intarweb "John Whitworth" typed:
>>
>>
>> Sorry, I am unable to offer much advice with the data set given.
>> However:
>>> I hear a quiet noise which sounds like a capacitor discharging.
>>
>> I would love to know what a capacitor discharging sounds like.
>>
>> Rgds,
>> --
>> Shaun.
Sudden Power Off 
Linear Mode
