I think they stopped because the power supply detected a short and
shut down. ...
Another possibility are any USB cables that go from header pins on the
mobo to the case's front USB sockets. ...
Do you have a floppy disk drive in the system? If so, verify that its
power connector is plugged in right because it's easy for it to be off
by one pin and cause a short.
It could be .... the power supply controller detected a bad voltage
due to excessive ripple voltage. Or it could be power supply
controller has a circuit defect. Or a CPU function caused power
supply controller to shutdown. Or power switch gets stuck. Or power
consumption is too large causing power supply to enter current
foldback limiting - a shutdown state that is temporarily cured by a
complete power off. Or a safety cutoff circuit was activated that can
only be reset by removing the power cord. Or ... maybe another 50
reasons.
Instead suspect is immediately identified if numbers are first
obtained. Numbers that "definitively" point to a suspect. No
speculation. No shotgunning. And numbers that mean replies with
useful information are posted here. That same point is made
repeatedly in CSI. "Follow the evidence." 'Could be this and could
be that' is only wild speculation. How many items should Lara replace
or disassemble? That is called shotgunning. Keep replacing things
until something works? Does that mean the problem was solved? No,
shotgunning only means symptoms were solved - and only maybe the
problem. Worse, shotgunning can also exponentially complicate a
problem - create additional defects.
Unknown? Yes, even a defective power supply can still boot a
computer. Others will assume a good power supply only because the
computer boots. But a defective power supply is found "definitively"
defective if using numbers. Again, the two minute procedure results
in answers that are 'definitive'..
Which component in that power supply 'system' is defective? Lara
shotgunned a power supply how many times? A failure still exists. At
what point do we learn why bad techs (in all industries) may become
unemployed by shotgunning. But again, logic and science says, "Follow
the evidence".
A procedure that teaches good diagnostic techniques goes right
after the problem. First, a suspect is identified. Only then is the
suspect replaced.
What is the most complex part of this problem? Getting one to learn
good diagnostic techniques; to not shotgun. How many power supplies
were shotgunned and that computer still does not work? After much
labor, zero has been accomplished. Unidentified is even one
"definitively" good component. If a list of what is 'definitively
good' and of what is still unknown existed, then something was
accomplished. Exactly what the Japanese meant when they said, "Don't
work harder; work smarter." Only two minutes with that procedure.
"Follow the evidence". We still know nothing because, instead,
shotgunning was performed. After all that labor (shotgunning), zero
has been accomplished. Everything remains in an 'unknown' category.
Nothing is 'definitive'. Therefore nothing has been accomplished.
Definitive means numbers. Definitive means no shotgunning.
Definitive means replacing only a known suspect. Definitive means
knowing the problem is solved; not just its symptoms. But learning
may be too difficult. Instead, they shotgun. They work harder; not
smarter.
