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.