It IS important that one pay attention to the specs of WHATEVER he's
buying. All my cases are P4 certified.
That's not significant. Most if not ALL cases/PSU made during P4 era will
be claimed P4 certified. Even Intel's ATX12V spec is just a guideline,
not the actual schematics, component spec for any particular wattage unit
nor that unit's lifespan. 12V spec is +-10%, but would you consider such
a P4 certified supply a good one if it constantly had +10% to -10%, 2.4V
peak to peak ripple?
IMHO, only the DISCUSSIONS seem to be more common.
Scenario...something doesn't work. Groups suggest...'try a new ps'.
Tries it...everything works again. Culprit? PS.
Well given only those details it sure does seem like the PSU was the
problem. Perhaps it was a connector or whatever, but how much detail do
we want to get into, taking each post of problems and suggesting
out-of-round microns for a plug contact? It's true that too often I may
assume a system was "unchanged", it remained static and a fault developed.
I used to ask "what's changed since onset of problem", but after a certain
point I guess I leave it to the poster to mention what they feel is
significant. If they'd just installed Service Pack 85 and windows crashes
but they only mention their doubts about a power supply (motherboard
monitor) 12V reading, certainly the thread would be off to a bad start.
But...new power supply has new connectors. All kinds of components are
r&r'd...cables of all kinds are moved, etc.
So...what REALLY fixed the problem? Moreover, the old PS never gets
checked...spec'd out.
Some PSU faults are beyond the scopre of user's ability to troubleshoot or
equipment is too costly. We're both aware than in usenet it's VERY easy
to spend someone else's money, but IMHO, a good system is build out of
good components... seldom do I suggest replacing what would seem to be an
appropriate amperage name-brand power supply.
But, again...only the ps got replaced. So, of course, the ps solved
the problem.
But often users are instructed to strip system down to only minimal parts,
so that has similar potential. Would it make more sense then to suggest
the user completely uninstall the power supply from chassis then REinstall
it again? Perhaps we need a good FAQ that includes these basics that are
basic enough to be overlooked/assumed/etc.
60%-70% of all hard drives RMA'd are found to be NOT defective.
Yet...these drives do not work for the owner when sent in. And they
DO work for the owner when he gets the new one back.
Well I have no answer for that... if it makes a squealing sound or
generates a smart error or trouble code from manufacturer's diagnostics
I'd recommend backing it up and getting it RMA'd, but if the drive isn't
defective (or failed for *other* reason) then why the problem? It has
to be handled on a case-by-case basis... I'd like to think that your
60-705 figure of non-defect is only in general, not applicable to our
group... remember that many people need to be told what/where a hard drive
is and how to remove it, let aone knowing what usenet is except for that
"other" link on a Google search page where "real people have
conversations"... nevermind mentioning IRC.
I feel blame on the power supply is on the same plateau.
To some degree yes, but on the other hand you might just be in denial of
the situation with many generics.
No. Actually, I've only had a few problems lately...and its been with
the AMD chips...because there's no certification for their high-end
chips that's easy to advertise on the case literature. But I've never
had a problem with the P4-advertised cases.
It IS more likely with a generic because their wattage is overstated.
Some even claim amperages that the regulators alone cannot support and be
in-spec, nevermind the rest of the unit.
The problem with AMD chips is similar, when a unit is labeled to provide
"X" number of amps, but can't. Quite simply put, when a name-brand 300W
unit can power a system that a Generic 350W with higher 5V amps can't even
POST, something is wrong, and that something is most often the misrating
of the generic.
Remember...we're not talkin' about building a server here. There IS a
difference in components...as we both know. But, for the average home
user, a generic ps has always worked fine for me.
Actually I made no distinction whatsoever for a server vs. a PC, beyond
the amperage needed per components. It would be different if considering
redundant PSU, but for a single unit I demand a power supply that can work
24/7 under full load for years. That isn't all that hard to realize
either for someone who has spyware or (SETI/FAH/etc) screensavers, etc.,
until they turn that system off at night, IF they do. Basically, if a
system fails within 5 years with any extra maintenance needed beyond
cleaning the dust out (or possibly relubing those crap video card fans) I
wouldn't bother to build it. Perhaps that's just MY hobby, building
systems that are more than just rock-solid but with a huge margin. I want
a power supply that can support a video card swap or a couple more hard
drives, whatever is appropriate for the room in the given chassis, else
WHY have a big space-consuming chassis to begin with?
Again...yer certainly welcome to your opinion. But I've never heard
of ANYONE bench testing each power supply before they start building.
I know I never do. If its P4 rated, it suits my needs.
Then how could you know if generics are really suitable or just barely
working? When a motherboard fails, will you suspect the generic power
supply or just blame the board? Same goes for video card, hard drive,
etc.
How can you disagree when you already stated that you don't test them?
