On Sat, 03 Feb 2018 06:38:34 -0800, tabbypurr wrote:

On Saturday, 3 February 2018 14:00:49 UTC, ss wrote:

Update:
So fuse arrived today, soldered in to the charger, switched on and
POP! That will be for the bin then as now beyond my understanding of
electronics.
At least it never exploded this time.

Ok now waiting on replacement fuse & holder for the other charger,
fingers crossed.

worth replacing the 4 mains diodes if you know how, those are the
most likely cause.


NT

I have taken out the doide that reads faulty. I assume its a straight
forward order, its a IN5399 any other rating to consider for ordering.

Is it worth replacing any other parts at this stage? I will replace the
fuse again.

1N5399 defines the ratings (not IN), so it's that simple. It just needs
to go in the right way round, as per the original, line at the same end.

Other parts? We don't know is the only definitely correct answer.
Anything else would result in no end of time wasting here. You might or
might not need other parts.

When you power it up if you do so via a 100w light bulb it'll spare
diode & fuse if anything else is faulty.

Unfortunately, that's likely to lead to even more component failure if
you've failed to identify and replace *all* of the faulty components
since the rapid destruction of the safety fuse will have limited the
damage in the original failure event by timely suppression of the
destructive transients before they could do any further damage.

A typical SMPSU is effectively made up from a selection of components
arranged as a fleet of "Accidents waiting to happen". Almost every
component in an SMPSU can lead to a catastrophic cascade of faults with
very few of them resulting in a non-catastrophic failure depending on
whether they'd obligingly failed open or short circuit as required to
allow such a non-destructive failure event.

Considering the workings of a SMPSU gives me the heebee jeebies every
time[1]. It's rather like an electronic form of Jenga or Russian Roulette
where 5 bullets have been loaded into the "Six Shooter Revolver" where
the victim is that obligatory safety fuse whose primary role is to "Put
out fires before they get a chance to take hold and destroy the whole
forest.".

SMPSUs *can* be repaired but unless it's one integrated into the main
PCB of an expensive piece of kit, it's usually far cheaper to simply
replace what has now become a commodity item with a new replacement unit.

One normally wouldn't even bother *attempting* to repair a ten quid
commodity PC ATX PSU these days. The only time I broke this rule of thumb
(don't get sucked into wasting precious time on a folly) was when a
customer returned their PC after a completely successful repair after
blowing the safety fuse in its ATX PSU by accidentally nudging the dual
voltage switch into the 110v setting when they'd been setting it up on
returning home.

This particular failure mode gave me pause to reconsider the benefit of
opening it up to do some basic testing of the HT module with a view to
simply try a replacement fuse. Said tests suggested there was a
possibility that the fuse had blown in time to prevent damage so it was
duly replaced and the offending and redundant voltage selector switch was
disconnected. Contrary to my expectations, the repair proved successful.

Prior to that, I've repaired a custom AT PSU (NEC Powermate II) that had
become so reluctant to restart that I needed to utilise a hair-dryer to
persuade it to spring back into life (it was a Novell Netware 3.11
fileserver box). This involved replacing a small 100mF 16v capacitor that
had dried out from the heat of an adjacent plastic power transistor.

Much earlier than that, I'd managed a successful repair of a 1970s Gould
5v 10A SMPSU which had blown one of its 1N4000 series bridge rectifier
diodes along with its user serviceable glass fuse. I can't remember the
exact number of the diode, only that it was a 700PIV rated 1N4000 series
diode. I remember the PIV rating rather than the precise part number only
because it struck me as rather odd that Gould should have chosen a diode
with a barely sufficient PIV rating for the job.

A bridge rectifier needs to have a PIV rating equal to or greater than
double the peak voltage of the maximum permitted mains supply (2 times
root 2 of 265 = 2 x 375v = 750v) to avoid the need to provide current
spike limiting resistors when used with a smoothing capacitor. The 700v
rating was barely enough to cover mains voltage excursions up to 247v
which aren't uncommon events where I live (urban underground supply).

Gould could so easily have used 1N4000 series diodes with 800 and even
1000 volt PIV ratings which begged the question, "Why not?" which then
led me to conclude that this may have been a deliberate choice based on
the concept of "Better to have a ten cent diode sacrificially fail to
protect a more difficult to replace ten dollar HT switching transistor".

This was reinforced by the fact that the diodes were far more accessible
than the HT switching transistors and the fact that the SMPSU would have
cost far more than a typical Hi-Fi set up of the day, maybe even as much
as a small capacity motorbike!

Today, thanks to consumer demand driven mass production, the once cheap
'n' cheerful 50Hz mains transformer bridge rectifier capacitively
smoothed low voltage DC supply is now both inefficient and heavy and the
most expensive in the cost of its raw materials that the new "Cheap
'n'Cheerful" has become the ubiquitous SMPSU. Only the most canny of
futurists with his wits at full stretch would have foreseen this dramatic
turn around in the time of just a single generation (Thanks to the PC
revolution, commoditised SMPSUs became a fact of life in the late 8os to
early nineties).

[1] The only other example of a PSU circuit diagram (a DC to DC converter
actually) giving me the heebee jeebies was seeing the use of thyristors
in place of switching transistors relying on secondary thyristors to
discharge a capacitor across the primary switching thyristors in order to
turn them back off. Predictably, fuse blowing events were a fairly common
failure mode with this novel use of thyristors on DC supplies. :-)

--
Johnny B Good