whodunnit: motherboard dead

[[email protected]]

Hey folks,

My new motherboard died after only a couple of
weeks on duty. The vendor says he owes me nothing,
the motherborad was ok, it died of a power surge.

Let's see:

- the wall outlet for my PC is properly grounded
- the monitor on the same outlet was not affected
- a serial modem on the same outlet was not affected
- the power supply unit in the PC was not affected
- a SW receiver on the same outlet was not affected

Is it still possible that some power disturbance killed
the motherboard (and the motherboard only) and how?

Thanks.


john

 

[John Popelish]

Hey folks,

My new motherboard died after only a couple of
weeks on duty. The vendor says he owes me nothing,
the motherborad was ok, it died of a power surge.

Let's see:

- the wall outlet for my PC is properly grounded
- the monitor on the same outlet was not affected
- a serial modem on the same outlet was not affected
- the power supply unit in the PC was not affected
- a SW receiver on the same outlet was not affected

Is it still possible that some power disturbance killed
the motherboard (and the motherboard only) and how?

Thanks.

john

Extremely unlikely. Most power surges damage the power supply, not
the mother board, unless they come in through some peripheral device.

 

[Bill Stock]

Hey folks,

My new motherboard died after only a couple of
weeks on duty. The vendor says he owes me nothing,
the motherborad was ok, it died of a power surge.

Let's see:

- the wall outlet for my PC is properly grounded
- the monitor on the same outlet was not affected
- a serial modem on the same outlet was not affected
- the power supply unit in the PC was not affected
- a SW receiver on the same outlet was not affected

Is it still possible that some power disturbance killed
the motherboard (and the motherboard only) and how?

Thanks.


john

I gather this is a FrankenPuter (Homebuilt)? How do you know it's the MB and
not the RAM or CPU? If you installed the CPU yourself, it's quite likely a
heat problem. If the unit was prebuilt, your vendor needs to get stuffed.

 

[w_tom]

The power disturbances include static electricity when you
assembled it weeks ago, transients that enter into mother
board from safety ground, and a power supply this is missing
essential safety functions (very common with power supplies
sold on price and without numerical specifications).

Without specific facts, such as which component on
motherboard that is damaged, then no one can answer your
question.

Power surges are rare events that occur typically once every
eight years. Four adjacent devices can all be confronted by
same transient. But if only one device provides both an
incoming and outgoing path for that transient, then only that
device destructively conducts the transient.

Furthermore since so many don't even know if a power surge
existed, what is it, and how it damages electronics, then the
power surge is that classic salesman excuse. Did a power
surge ever occur? How would you know? How would the salesman
know? Power surges (like heat) are the poor excuse on which
blame is cast. After all, who is going to prove that
accusation wrong in a world where most don't have electrical
knowledge.

Far more likely, especially this time of year, is damage
from static electricity.

And then so many other usual suspects exist including other
defective components, motherboard conductive standoff breaking
through solder mask and now touching a printed circuit trace,
and infant mortality.

Without facts, no one can answer your question. You have
not even taken voltage readings from the power supply - the
first fact one obtains.

 

[Jim Douglas]

Typically most decent places will replace a board this new, and typically
boards that are bad do so within the first 3 months of usage.

 

[Rodney Kelp]

You should get a good UPS with filtering and spike protection. Much safer
than a wall outlet. Static (6000v) could fry something especially now with
so many USB external components. It could have just failed on it's own also.

 

[[email protected]]

Without facts, no one can answer your question. You have
not even taken voltage readings from the power supply - the
first fact one obtains.

I'll follow up on your hint that the power supply
unit may be a cheap one lacking essential features.
[Any models recommended?]

Otherwise, I think I was careful with static
electricity and in any event its effect would not
have taken a 4 week delay. Also the processor
survived.

The problem was at first a wobbly boot: the
machine would start and stop within a few seconds,
fans still turning. To complete the boot, I pushed
reset. Then one day nothing, just the peace of
death.

Regards


john

 

[Rheilly Phoull]

Hey folks,

My new motherboard died after only a couple of
weeks on duty. The vendor says he owes me nothing,
the motherborad was ok, it died of a power surge.

Let's see:

- the wall outlet for my PC is properly grounded
- the monitor on the same outlet was not affected
- a serial modem on the same outlet was not affected
- the power supply unit in the PC was not affected
- a SW receiver on the same outlet was not affected

Is it still possible that some power disturbance killed
the motherboard (and the motherboard only) and how?

Thanks.


john

I would be asking the dealer how he could determine the 'surge' ocurred and
perhaps show you the damaged components.

 

[w_tom]

You are correct that static electricity probably is not
reason for failure. But static electricity creates
overstress. Overstress results in parts slowly getting worse
days or months later. Yes, static electricity does remain a
possibility. Just not a more likely reason for failure.

More likely is infant mortality. IEEE Spectrum in a recent
issue featured a cover story on why the human body fails.
Charts there demonstrate how infant mortality causes so many
failures in so short a period - and a general trend on when
failures happen.

Now for another lesser reason for failure. Does the power
supply manufacturer provide a long list of numerical specs.
How to suspect the worst (in any technical product). The
manufacturer so fears you might learn facts as to deny you
those facts. In reality, maybe one percent of consumers
understands specs. But without numerical specs in the hands
of all, then the one percent is disempowered. IOW no long
list of numerical specs - then suspect a scam. The one
percent cannot sound the alarm.

An abridged list of specs that any responsible power supply
manufacturer provides. If no numbers, then assume they are
hiding something - missing essential functions:
Specification compliance: ATX 2.03 & ATX12V v1.1
Acoustics noise 25.8dBA typical at 70w, 30cm
Short circuit protection on all outputs
Over voltage protection
Over power protection
100% hi-pot test
100% burn in, high temperature cycled on/off
PFC harmonics compliance: EN61000-3-2 + A1 + A2
EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
Hold up time, full load: 16ms. typical
Efficiency; 100-120VAC and full range: >65%
Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
Dielectric withstand, input to output: 1800VAC, 1sec.
Ripple/noise: 1%
MTBF, full load @ 25°C amb.: >100k hrs

How top dump power supplies into N America at higher
profit? Forget to include essential functions. Then sell the
power supply at $25 or $40 retail. A minimally acceptable
power supply has a full retail price of $65. Just another
number that should create suspicion.

Concept does not stop with power supplies. Concept even
works for cars. Pontiac is hyping performance. So why do
they not provide both horsepower and liters on the sticker
sheet? Because Pontiac engines remain some of the lowest
performance in the industry. Even the greater noise suggests
their low performance. Just another example why the
manufacturer would hide the numbers. Do the horsepower per
liter arithmetic yourself. Numerical facts that the
manufacturer hopes you will not learn.

Your computer symptoms only say, "Time to collect facts".
In your case, I would start with a 'usual' suspect - the power
supply 'system'. Not just the power supply. System includes
controller on motherboard. To understand the 'system', you
need a tool as essential as a screwdriver - a 3.5 digit
multimeter. Concepts to better understand why boot fails are
delineated in: "Computer doesnt start at all" in
alt.comp.hardware on 10 Jan 2004 at
http://tinyurl.com/2t69q
"I think my power supply is dead" in alt.comp.hardware on 5
Feb 2004 at
http://www.tinyurl.com/2musa .

Another poster here demonstrates how myths are promoted. He
recommends a UPS to stop spikes and surges. Even the
manufacturer does not make that claim. Will the relay inside
a UPS that takes tens or milliseconds to respond stop a
destructive spike that is completed in microseconds? Again,
how numbers expose a myth - such as a plug-in UPS for hardware
protection.

Everything needs a number, or at least a numerical
relationship. Without numbers, then only speculation
remains. Your motherboard supplier demonstrated classic
speculation. Typically destructive power surge occurs about
once every eight years - another number that varies regionally
and locally. Many who don't know anything about electricity
hype myths such as daily surges. They just 'feel' these surges
must exist. If daily surges exist, then we were trooping to
the hardware store everyday to replace destroyed electronics -
even 30 years ago.

Notice the need for numerical information. Compare that to
what you provided AND what your supplier concluded. A more
likely reason for your failure is infant mortality or a
failing power supply. But without numbers, we can only
speculate. Provided are how to get numbers. Even if you
don't understand those numbers, still, those numbers make it
possible for others to provide useful responses. No numbers
means wild speculation - or maybe a scam. No numbers is why
another poster recommended the plug-in UPS.

I'll follow up on your hint that the power supply
unit may be a cheap one lacking essential features.
[Any models recommended?]

Otherwise, I think I was careful with static
electricity and in any event its effect would not
have taken a 4 week delay. Also the processor
survived.

The problem was at first a wobbly boot: the
machine would start and stop within a few seconds,
fans still turning. To complete the boot, I pushed
reset. Then one day nothing, just the peace of
death.

 

[Bill Stock]

Without facts, no one can answer your question. You have
not even taken voltage readings from the power supply - the
first fact one obtains.

I'll follow up on your hint that the power supply
unit may be a cheap one lacking essential features.
[Any models recommended?]

Otherwise, I think I was careful with static
electricity and in any event its effect would not
have taken a 4 week delay. Also the processor
survived.

The problem was at first a wobbly boot: the
machine would start and stop within a few seconds,
fans still turning. To complete the boot, I pushed
reset. Then one day nothing, just the peace of
death.

Sounds like it might be a bad PS. I had one flake out slowly, where the
system kept freezing up.

If it's an AMD CPU, you can check their website for recommended power
supplies. Intel may have similar info.

See if your dealer will let you 'borrow' a power supply to test your theory
or better still get him to do it for you. I've had several bad boards when
putting together systems, several from one vendor. (Stopped dealing with
them). Others vendors make you jump through the return hoops and will
eventually send you a refurb months later. The latest vendor tests the dead
board on the spot and hands you a new one. But I've only ever had ONE dead
board from them. Some vendors resell dead boards hoping to get rid of their
problems.

 

[w_tom]

Borrowing or swapping power supplies takes longer, puts the
system at greater risk, and my only mask the original
problem. The faster and cheaper procedure to first identify
the problem - and one always wants to identify the problem
before changing any parts - was posted previously in at:
http://tinyurl.com/2t69q and
http://www.tinyurl.com/2musa

Problem sounds like a bad power supply. It also sounds like
a bad power supply controller. It also sounds like a bad
power switch. It also sounds like bad motherboard
capacitors....

First get facts before speculatively replacing anything.
Saves money. Saves time. Solves problem the first time. And
teaches a human the 'whys' about computer failure.

 

[[email protected]]

Thanks a lot, Tom, but you are ways above me.

Even if (big if) I manage to do all those
measurements you insist on, what would be the
practical consequence? Discard motherboard and PSU
until you get products with good test values.
Sorry, I'm not even trying. I bet you don't do it
yourself.

My PSU was indeed cheap chinese junk. So dirty
power and junky PSU may give an explanation of
what happened, some building up of damages over
time.

I'll get a good PSU for my new motherboard and
I'll get an UPS, although you seem sceptical in
this respect. Surely better than nothing?

Regards


john

 

[w_tom]

Those voltage measurements may be performed by junior high
school science students. There is nothing complex, difficult,
or expensive in what was posted. In fact, measurements can be
accomplished in just minutes. Far longer to read and learn
what must be done. IOW the knowledge is that powerful and
useful. As another once said, "the only thing you have to
fear is fear itself". Get the meter. A tool as essential to
computer work as a screw driver. So ubiquitous as to be sold
even in Home Depot, Lowes, Sears, and Radio Shack.

In the meantime, a UPS is for data protection. Hardware
protection that works at the appliance is already inside the
appliance - as required by industry standards. Protection
that may be overwhelmed IF destructive transients are not
earthed before entering the building. The UPS costs what -
maybe $100 - and to only protect one appliance? Ineffective
hardware protection as made obvious even in manufacturer's own
numerical specs. The UPS is only for data protection and not
for hardware protection.

Effective hardware protection costs about $1 per protected
appliance. Installed to protect everything. Superior
protection especially for the computer. With a 'whole house'
protector, then a surge will be earthed before it can
overwhelm internal computer protection. One effective
protector sold in Home Depot is Intermatic IG1240RC. For
hardware protection, spend less money for a protector that
really works - because it makes a short and so essential
connection to earth ground. A UPS for hardware protection is
too often recommended by those who don't even know how
electricity works. Where is the earth ground to that UPS?
Just a concept the UPS manufacturer instead avoids. A
protector is only as effective as its earth ground.