Vented caps.

[tomizett]

...aren't supposed to do this!

Couldn't resist sharing these images - one of the most impressive failures I've seen for a while. Possibly the most impressive aspect was the dent that it had put in the neighbouring capacitor.

 

[atferrari]

Could you give more details. Very old cap?

 

[JimB]

Vented capacitor?

Apart from the gaping hole in the side, where is the (intentional) vent supposed to be?

JimB

 

[Nigel Goodwin]

It's not vented, it's exploded

As already suggested, it's a very old cap.

 

[tomizett]

Apart from the gaping hole in the side, where is the (intentional) vent supposed to be?

If I'm getting the terminology right, the "vent" is the cross-shaped indentation in the top of all modern caps which is supposed to rupture in a controlled manner and prevent dangerous explosions in the event of a build up of pressure. This does have one, but of course it's out of sight in those pics.

It's not very old, actually, only a few years. The heat has caused the plastic wrap to come off, which gives it the old-fashoned metal can look. Root cause isn't established yet - we did see some early life failures of these, which where put down to a bad batch, but it's possible that something else went wrong here which caused a significant over voltage.
Can't go into too much detail, as I don't want to be accused of bad-mouthing equipment manufacturers.

 

[GromTag]

An possibility.

Internal contamination. There once was an 220uF 16V cap here locked to a 5V supply. The type of cap is the same as the bottom right corner on the board, the name of the caps were not disclosed to me. Serviced 5 month 27,2017.

A small contaminant can expand casing an eventual cross short in an isolated pocket leaving no room to make it to the vent. Or traverse the remainder of the material.

BT series, Nichicon or Seika, leaning towards Seika as the design look is more like on this board.

 

[JimB]

If I'm getting the terminology right, the "vent" is the cross-shaped indentation in the top of all modern caps which is supposed to rupture in a controlled manner and prevent dangerous explosions in the event of a build up of pressure. This does have one, but of course it's out of sight in those pics.

OK, got it!

With a cap that size I was expecting to see a "bursting disc" type vent in the end with the terminals, usually a thing rubber membrane about 3 or 4mm diameter.

JimB

 

[JimB]

What happened to the terminals on the cap?

It looks as if they have had some kind of high current event and vaporised!
I have never seen that before.

JimB

 

[GromTag]

Tricky... Some leads are outer plated with a copper element core, or as such. It would take some current to vaporize a lead as large as the side burst cap. Only thing I can find familiar is the leads to newer components that have that tin or other outer coatings, the copper core if any would burn up whilst the plating will vaporize leaving some residue on the board.

Those leads do not appear to have a core material resulting in better strength and current handling. Looks a main filter cap leaded capacitor.

:Edit 2 :

 

[tomizett]

Yes - it seems the leads have been blown out by a large current. I'm guessing that this cap went short and then the others in parallel discharged through it. I've never seen it before either!

I've just chopped a lead off one of the surviving caps and it's quite hard - aluminium I think? Doesn't look like copper.

(edit)
For reference, at there's no scale, the component pictured is 70x35mm.

 

[Nigel Goodwin]

It's not very old, actually, only a few years. The heat has caused the plastic wrap to come off, which gives it the old-fashoned metal can look. Root cause isn't established yet - we did see some early life failures of these, which where put down to a bad batch, but it's possible that something else went wrong here which caused a significant over voltage.
Can't go into too much detail, as I don't want to be accused of bad-mouthing equipment manufacturers.

As it's not very old, then it's almost certainly a crap quality component, from a crap manufacturer - all too common when equipment manufacturers are chasing the cheapest possible parts

Samsung fitted CapXon electrolytics in the PSU's of the original Sky HD boxes - a friend of mine consulted the CapXon datasheets, and measured the temperature inside the receivers - coming up with an expected life span of around only ten months!!. Needless to say, they failed in HUGE numbers, many while still under guarantee.

 

[tomizett]

It's actually a Nichicon, but the root cause of the failure hasn't been established yet - I suspect it was subject to a massive over-voltage, so that cap itself may not be to blame!

That's some pretty poor engineering by Samsung - sadly not very surprising though.

 

[Nigel Goodwin]

That's some pretty poor engineering by Samsung - sadly not very surprising though.

To make matters even worse, it was the exact same power supply they provided to Grundig (and later Thomson) for their satellite receivers and SD PVR's - and they failed regularly there as well, but not as fast due to their lower operating temperatures. I salvaged a number of PSU's from the old boxes, replaced the capacitors, and then had a decent number of exchange boards to fit in the HD PVR's

 

[GromTag]

Thought process, most thought of, a spike resulting in a capacitor fault resulting in a load draw, heating, ripple added resulting the draw, then an eventual burst that freed the cap whilst the leads were molten?

Would the damaged cap be closer to the power source or furthest? It might have been caught in the middle of a stored effort (other caps) and the driven supply. If it were closer then a power spike could have started that if no other components show to be the resulting cause or damaged themselves.