On Wed, 28 Feb 2007 14:03:26 -0800, "Bob F"
<bobnospam@gmail.com> wrote:

>
>"CBFalconer" <cbfalconer@yahoo.com> wrote in message
>> When someone offers an XXX watt PSU, load up sufficient bulbs to
>> dissipate XXX watts, distributed between the various lines
>> according to the mfg spec for max load. Connect, turn the PS on,
>> and wait. I predict 90% of PSUs will self destruct in 30 min or
>> less. So do this test at the sellers location, before buying. If
>> it passes, buy it. :-)
>
>How many sellers have you done this test at?
>
>Bob
>

It is good point that one can't do the test at a store,
before purchase. The PSU could still be tested after
acquired, whether it be one bought or that came *free* with
a case, to determine that it meets minimal acceptible
standard for implementation in the intended system (if free)
or the labeled specs (if purchased separately).

One that came free with a case could make the return
unfeasible to the buyer due to the high return shipping cost
of an entire case, but if having the PSU included caused the
case to cost significantly more or if purchased separately
it would not be unreasonable to return it for a refund or
demand replacement.

The grey area here is that returning for a refund may not
be acceptible to the seller if the unit is damaged, it might
instead be replaced with same or have to be RMA'd to the
manufacturer and replaced with same (potentially poor)
quality or mislabeled PSU.

Many PSU that are poor will run for a few hours or longer
before self-destructing which is unfortunate as the longer
it lasted before failing, the greater the odds the testing
was terminated and it was integrated into a system. It can
be reasonable to pay a bit more for the name-brand for some
expectation of quality instead of trying to depend on
published specs or assuming it meets ATX guidelines for safe
shutdown.