SMPS ('wall-wart') failure, how common and what's the risk?

[awallin]

12VDC 1A SMPS decided to die yesterday. Blew a fuse also.
http://www.anderswallin.net/2015/04/smps-fail/

Any idea what might be wrong? I did already get a replacement for 8 euros... The backside of the board was quite black but on the front there isn't any single component that clearly is to blame?

What are the risks with these wall-warts? I probably have about a dozen and it doesn't seem very safe if they all look like this one on the inside!?!

[T3sl4co1l]

As in, it blew a fuse elsewhere, not just the internal fuse(s)?

Lovely bit of arc flash... tempted to say it died of lightning strike or other transient.  I'm not too familiar with 240V, honestly; 480V arc flash is an awe to behold (awesome or awful, depending on proximity..).  Our weak 120V over here otherwise doesn't do very much when things die.  So I don't really know if 240 is enough to sustain a good kickin' arc flash, or if you need a big transient on top of it to achieve the results shown here.

Looks badly designed, in any case: the top side shows little or no apparent damage, but the board is absolutely cooked around the switching chip (whatever it is, some sort of integrated switch controller).

Tim

[awallin]

The fuse is an automatic one "hager mcn110e c10" so I guess 10A.

TV, PVR with harddisk, TeraStation RAID-disk, D-link Wifi-router, etc. are all under the same fuse and they seem fine - so luckily there wasn't a spike that would have killed all my electronics
The Cisco cable modem powered by the blown SMPS is also fine after powering up with a generic 12VDC 1A wall-wart.

[SeanB]

Looks like the board turned to conductive char from the power device running so hot in there, along with the hot resistor adding extra heat to help it along. Then the power mosfet ( or transistor) went short circuit, which is why the copper traces blew off from the mains side, which would also have tripped the house breaker as well.

Actually a reasonable SMPS, they put in the fuses, input filtering and proper isolation on the transformer, along with correct creepage distance between primary and secondary. Only killer is that the components are run so hot, a bigger case, a better heatsink and a cooler running design will have a much longer lifetime, just because it does not run so hot the board itself cooks.

[Ian.M]

It has what's likely to be a VDR to clamp over-voltage transients (blue component inside heatshrink on right of second photo - the heatshrink wrapper is a common way of limiting the physical damage when a VDR fails catastrophically) that shows no visual signs of distress so it probably died of old age rather than a severe transient. 

Some part probably failed in the circuit round the chopper chip (or as SeanB just pointed out the board cooked and became leaky) resulting it failing to turn off, core saturation and massive over-current.  A PCB track or current sensing resistor then fused and arced. 

C4 is probably the decoupling on the chip's low voltage supply, and if its gone high ESR, might be the original cause.

[rollatorwieltje]

I used to have the same cable modem, after a few years it started to randomly reboot. My cable provider replaced it with a newer modem. The newer type came with a slightly bigger power supply (15V 1A instead of 12V 1A if I remember correctly). Considering the epc3000 modem ran stupidly hot, I suspect the 12V adapter was a little underspecced for the job and destroyed itself.