Smoke

[Radio Wrangler]

It's remarkable that so much smoke can be packed into a small component1

Well, everyone at work's now aware of Rifa capacitors.

I helpfully took one of my HP sig gens in to cover for having a couple of Marconi TF2024s out for repair. I'd checked it at home before loading it in the car. All fine. Plugged in at the lab and switched on. Still OK. 5 minutes later and there was a quiet pop, then came the smoke!

We had all windows and the fire door open for the day.

It was inside a "Schaffner" metal cased filter, part of a line module and it burned a hole to get that smoke out. Of course, it's upstream of the fuse

David

[electroanorak]

I have had these do the same thing in all my 5 Gould DSO scopes. Nasty smelly things as well.....

Daniel.

[ronbryan]

What is equally annoying is the high price of the replacement Schaffner filters - around £40 from suppliers with stock I seem to recall.

Ron

[turretslug]

This case demonstrates close compliance with the "doing a favour" and "maximum exposure" clauses of a well-known law.

Are current/modern Shaffner filters any better? I think I'd be rather averse to replacing like with like unless I had very definite reassurance that they didn't conceal lurking liability.

A couple more recent failures for the hall of shame,

[lesmw0sec]

The greatest smoke I managed to generate was in my early youth connecting two electro's in series without balancing resistors. Not just smoke, but an explosion which had my mother checking to see if I was alive. Took me ages to scrape the goo off the wall!

[Peter.N.]

Not as smelly as selenium rectifiers.

[mole42uk]

I'm not so sure. I had to leave my Gould DSO out in the open for days and leave the workshop windows open for quite a while....

[jonnybear]

I always remember in my early days as a TV engineer, calling at people's houses, knowing as soon as the door was opened, the smell of the good old BUSH selenium rectifier, in nearly all cases the poor old dog got the blame, and banished outdoors, if they didn't have a dog then it was the old man that came under attack, very little smoke though?.
John

[Refugee]

I had one detonate just a couple of days ago.
It was in the vacuum cleaner while I was sucking the workshop floor with it.

[cmjones01]

I've had two 1980s-vintage Rifa capacitors like that go off in the last couple of weeks - one in a BBC Micro at work that I use for some automated testing, and one in a Tektronix 2465 scope at home. In both cases there were no blown fuses, no drama, just a small pop and a cloud of acrid smoke. Both pieces of equipment continued to operate normally. I replaced the capacitor in the Beeb but haven't done anything to the one in the scope yet. I should replace it before I forget that it went!

Chris

[mole42uk]

I've heard of quite a few filters and capacitors going pop in the last months. I wonder if the increased mains-bourne noise from switch-mode power supplies is upsetting them excessively? I would have thought the 6kV transformers would stop it spreading, but if there's a few SMPS on the local phase, might that be significant?

[Radio Wrangler]

It seems to be age-related crazing of the plastic cases followed by moisture ingress, followed by detonation.

I've hunted up NOS parts in the attic and they've crazed without benefit of any electrical input.

David

[Leon Crampin]

I tested these Rifa capacitors in the '70s when I worked for a UK automotive electronics company. Despite being plastered with "approvals" we designed our own tests. These failed dismally and we never ever specified them.

As I recall, the hard epoxy encapsulent had no resilience and was not matched in its thermal coefficient of expansion to the capacitor element. Thermal cycling followed by my "dishwasher test" allowed moisture to enter the element resulting in inevitable rapid failure.

After all these years, the cracked epoxy we see now is a likely result of some degradation of the material itself, but is almost certainly aggravated by the stresses induced by repeated cycling in equipment. "Henry" vacuum cleaners run darned hot...

As we also found with many military approved components - when subjected to our tests - junk. As a generality, Philips and the Japanese made the best components (the rest, for consumer electronics, is history)....

Leon.

[60 oldjohn]

Originally Posted by mole42uk

going pop in the last monthsI've heard of quite a few filters and capacitors . I wonder if the increased mains-bourne noise from switch-mode power supplies is upsetting them excessively? I would have thought the 6kV transformers would stop it spreading, but if there's a few SMPS on the local phase, might that be significant?

Just a thought. Could the recent increase in the Sun's solar activity have any influence on these capacitor failures?

John.

I seem to have lost part of the Quote.

[Maarten]

Most likely not. RIFA capacitors (and impregnated paper capacitors in general, somewhat depending on their construction) have been failing exactly this way for a great many years, irregardless of mains tension or sun activity. Humidity and usage (as Leon Crampin points out, among other things) may make a difference, but they will end up in clouds of smoke eventually.

[Refugee]

I have a couple of NOS ones that are branded Wyma that use the same resin and they have expanded internally splitting the epoxy outer allowing moisture in.
I will nor use them.

[short wave]

In Farnell "G series" switch modes PSU, had two units go on me.
First in a friends garage - lots of smoke - very embarrassing!
Second on a slow "variacing" failed at 140 volts, done outside at home no probs; now has a "fix it" sticker on it.

All on supply input filter.

regards S-W

[G6Tanuki]

Recently, a friend's late-1980s refrigerator went "Bang!" late one night and let the magic smoke out. Investigation the following day showed that the fault was the metal-cased "Delta" suppressor in the thermostat/compressor-start circuit which had eviscerated itself.

He was all for buying a new refrigerator - I fixed the problem by removing the capacitor, then cutting the mains-lead really short and fitting an IEC320 3-pin plug.
This then plugged into a commercial computer-type IEC320 "kettle lead" that had an inline L-C-Delta RFI filter.

A couple of cable-ties round the IEC connector to keep it firmly together and secure against part of the fridge's case, clamp the new kettle-lead to the case using the original power-lead's strain-relief clip - job's a good'un and I got a six-pack of "Badger" Beers as payment.

[Billy T]

Here's an example of a bad RIFA day: Three metal detector boards, eight leaky RIFA caps (various ages) and one exploded smoke generator.

To be fair, the yellowed examples are over 20 years old and the lighter colour are 10+. Unfortunately the large ones cause collateral damage when they go off, the smaller examples just go leaky and are replaced on sight. The two on the right were manufactured in Australia.

Cheers

Billy

[vinylspinner]

Same happened to me whilst listening to my Revox A77 reel to reel late one evening, woken by a nasty smell and lots of smoke, the machine carried on playing though!

Nigel