On Sat, 3 Feb 2018 14:13:22 -0000 (UTC), Cursitor Doom
wrote:

On Sat, 03 Feb 2018 12:12:56 +0000, Archibald Tarquin Blenkinsopp wrote:

The switching device would be short in a number of cases and mosfet or
bipolar the cause was all too often the RC snubber network across the
primary. The purpose of it was to remove the high energy spikes which
are a natural result of switching an inductive load at speed.

There's also an LC mains filter prior to the rectifier and that's highly
prone to failure as well. When an SMPS fires up from cold, the initial
surge current can be huge, 'cos it looks like a short circuit at switch-
on.

Rubbish!

The NTC limits the switch on surge. The filter would be more to stop
hash getting out from the PSU than anything coming the other way.

It isn't prone to failiure either. The coil has very few turns of
comparitively thick wire, so in the extremely unlikely event of a
shorted turn or two, you would never know the difference.

Assuming your neigbours don't listen to long wave on their wireless
that is.

The input capacitors are x or y rated and are not prone to the
failiures that the old polyester devices were subject to.

I personally have never ever replaced a coil or cap on the mains input
of an SMPS,

Now a transorb is a different matter, I have replace a good number,
but the similarity to a capacitor tends to end after body shape.

AB