Exploding X2 capacitor

[Heatercathodeshort]

Originally Posted by dglcomp

I keep meaning to replace the one in my Gran's Henry but never do, at +30 years old (pre cord rewind so supposedly that means that it's pre 1985) it's due to fail, and it certainly looks like it's starting to bulge.

Make it your New Year resolution. The shock of it failing may take your Gran with it. John.

[dglcomp]

It'll shock me more as she doesn't use it as she now finds it a bit too heavy and so has got herself a cheap corded 'stick' vac and so the Henry now spends most of its time in the garage and only comes out if I'm over and decided to do a bit of vacuuming. or my sister had driven over and the car needs cleaning.

It's certainly on the to-do list along with a new post motor filter as that is a crumbling mess and I will probably be staying over there for a while soon so will have the time to do it.

[mark2collection]

Earlier this year the RCD in my workshop would trip on the 'odd' occasion, odd given that I was only using my soldering iron on an unplugged set at the time.

Turns out the PSU in my Vectorscope was always running (power switch being a 'soft' switch) & the cap strapped Live to Neutral & Live to Earth, Neutral to Earth was a Rifa X/Y combi cap.

The noise was a gentle crackle, followed by a little fizzle ... The amount of smoke though! Also, the electronic damage it caused. It blew the SMPS Driver I.C, the opto isolator went for a chop & the PCB itself scorched ... It works great now, turns out the PSU electrolytics were tired too, so while I was in there ... The instrument was made in the early 90's & has seen moderate use.

Shortly after that, my S.E Laboratories SM2001A gave me some idea of the sound of rapid gunfire followed by an almighty 'POP'! This too was a Rifa, took a resistor with it, all good again now. This instrument was made in the late 70's & has seen little use.

Law of averages says these things come in three's, the next failure would have been my Philips Wow & Flutter meter (the cap was found to be electrically leaky) so this was replaced (non-Rifa, an odd cylindrical device with 3 wires coming from it, 2 black wires & a green one, no writing but past experience tells me what it may have been). Little used instrument from the 80's?

Lastly, my Marconi valve millivoltmeter has (& I didn't know this at the time) two waxie's, Live to Earth, Neutral to Earth. Needless to say it failed the PAT & needs addressing, this unit is hardly used now so it'll keep for a rainy day. Instrument dates from the mid-60's according to the NSN sticker

All other instruments in my workshop checked visually & then ran on my PAT, which I really must get in the habit of using on my own gear ...

Have a great Christmas, stay safe out there ...

Mark

[AC/HL]

This is one of our perennial threads. It would seem that connecting anything across raw mains is going to lead to trouble at some time in the future. Rifa, like the Hunts of old, are dominant suppliers in this market, so will naturally feature in the table of honours.
To quote a well known ditty, it all makes work for the working (man) to do!

[Hartley118]

Have often wondered whether capacitors connected across mains or between mains and earth have any beneficial effect on mains-borne interference. I find that they can usually be removed without any problem.

If they're part of a pi filter located in the mains connector and the electronics is in a screened case, then I guess they can be beneficial, but otherwise I suspect they're a waste of time and just a potential source of trouble.

Unless anyone knows different.........?

Martin

[Nuvistor]

They can be useful to help remove spikes etc that can damage silicon diode recitifers. Apart from the cases cracking and causing failure, a noisy mains supply can reduce the life of the capacitor.
Without the capacitor silicon devices in the PSU can fail early.

[vidjoman]

Most Japanese gear was usually fitted with ceramic caps in these X & Y positions. Typically rated about 2KV DC and 600 AC. Never seen one of them explode.

[Heatercathodeshort]

Completely agree. I salvage the ones used in Japanese power supplies. These can be ceramic or a yellow colour, circular in shape.
I have never had one fail. John.

[MrBungle]

Just to note, I had a failed ceramic one in Heathkit equipment about a year ago. It failed short and blew the fuse though. They certainly aren’t self healing like the film X2’s.

Replying to Hartley118, yes they do have a purpose and are not a waste of time. If you think that the reactance decreases significantly above mains frequency, you will see that they act as a low pass filter against spikes, high frequency noise and all sorts of nasties on the mains. This reduces noise on the equipment it is installed in and just as importantly it stops noise reaching the mains as well. This is particularly important on things with switching power supplies and flyback converters and low noise devices like RF front ends, oscilloscopes etc.

If it goes across the line, I’d replace with new plastic metalised film parts. They can handle silly fault currents and don’t burn your house down like the paper ones from RIFA. Salvaged ceramics are quite dangerous as they tend to crack around the leads and burn up if you’ve got a bad one. If it’s on the line supply, safety first, but as always this is all IMHO.

[Radio Wrangler]

If we plugged some rifa X-capacitors in now, do you think we could time them for the 5th of November?

David

[Refugee]

Mains rated ceramic capacitors are very reliable.
They also do VDRs in the same package.

[MrBungle]

Only if not damaged which is difficult. One thing I’ve seen is where they are soldered across IEC inlets across L-N. Every time you insert and remove the mains they crack a bit.

[Refugee]

I have had a shock off one of the ones on the unswitched mains. It is a "nippy" little shock that could make you drop a set if you get nipped unaware.

[TonyDuell]

Originally Posted by Refugee

I have had a shock off one of the ones on the unswitched mains. It is a "nippy" little shock that could make you drop a set if you get nipped unaware.

I have a Solartron instrument (a Data Transfer Unit) where I learnt the hard way to tap the 13A plug pins on the chassis after unplugging it from the mains. The filter capactors are on the input side of the mains switch and are large enough to give quite a spark.

I've seen devices with a suitably-voltage-rated 1M resistor in parallel with the filter capactors which should prevent this little problem.

To get back on-topic, I've had plenty of these capacitors fail in clouds of smoke, but fortunately no other damage. One time I remember I was using a rare-ish 'HP' (made by Samsung...) monitor and clouds of smoke came out. I thought at first it was line output failure. Upon opening the case I found the remains of the capacitor and realised what had happened. And although I don't like replacing parts that are still good, I will admit I replaced all the capacitors (X2 and Y) in that mains filter circuit.

[Refugee]

I would certainly consider replacing them in anything more than 10 years old.
It is a messy clean up if not.

[dglcomp]

Looking on eBay they sell "pre-owned" RIFA capacitors, if there ever was a product that should not be salvaged and sold it's these!

[Refugee]

Perhaps the buyer is supposed to apply power and have a whiff.

[Herald1360]

That is weirder than a weird idea

Better than selenium rectifier, I suppose......

[Refugee]

Selenium rectifiers would never sell very well.
There are plenty of cats with working back ends around for the buyers to sniff.

[Maarten]

Originally Posted by ted_ntsx

I was not aware that X2 types were so prone to failure.

Not in general, it's mostly the impregnated paper ones. Rifa fails fast, Wima a bit slower as do some generic south- and east European types that were sometimes used in domestic appliances.

Polyester and Polypropylene X2s can fail too, but less often and mostly will quietly loose their capacity. Don't forget they have a quite tough job to do, sitting accross the mains.