View Single Post
  #1   Report Post  
RubbishRat
 
Posts: n/a
Default ATX Power supply - defective fuse?


"Ixnei" wrote in message
news
I've had a 350W power supply running a P4-2.5GHz/512MB/80GB/GF4ti system
for quite some time now (probably over 1,000 hours on-time). I was using
it the other day, when out of nowhere I hear this loud 'click' and the
system is powered down. Opening up the power supply reveiled that the
fuse whas physically damaged - there were two shards of glass broken off
of the fuse, which were recovered from inside the supply. Everything else
looked very clean.

This is very odd; I've seen a lot of fuses blow in the past - you simply
see the filament inside the fuse is no longer connected, and sometimes
there is a faint grey/black tinge on the inside of the fuse glass. In my
case, I believe that the fuse had a glass stress defect in its'
manufacture, which eventually lead to a physical defect fail and the fuse
imploding...

Does this sound about right? Are there any other explanations? I haven't
gotten around to getting a new fuse and soldering it to the power supply
board yet (it always amazes me how they can be so cheap as to forego
purchasing the 5-cent fuse clips for easy removal/replacement)...


NO it isn't right at all, in my experience fuses go open circuit in three
main ways, 1/ they die of old age due to thermal cycling in which case there
are no outward signs at all or 2/ they open quietly due to a relatively
minor event such as a part drawing a higher current than it should or high
voltage spike on the line, maybe caused by lightning or next doors faulty
fridge motor ( examination usually reveals a tiny blob of melted metal on
each of the two ends of the fuse wire). or 3/ they blow open violently,
always due to a massive power overload (this latter is the case with your
fuse) caused by a catastrophic failure of an internal component, in this
case there may be external damage to the glass or ceramic fuse case but not
always, and the only trace of the fuse wire is a coating of condensed metal
vapour on the inside of the fuse case. I find it can sometimes help in
diagnosis to carefully break open the fuse case and determine how it failed.
the glass case is under no presure or stress at all and even if you break
the glass the fuse won't necessarily fail. If you replace this fuse with a
new one without removing the internal dead short then the new fuse will last
around a millisecond or so before it demontrates the fundemental truth of
this post.
Pete