Toroidal transformer catastrophic failure.

[Tractorfan]

Hi,
Last week I was asked to look at an electric swimming pool cover that had stopped working.
On investigation, I found that the 240 to 24 volt toroidal transformer had failed quite spectacularly. The enclosure was blackened within and small globules of copper scattered around.
The transformer, although permanently energised, is only under load when the pool cover is opened or closed.
The enclosure is well away from any water or damp, and is IP44 rated, so I can't imagine humidity being a factor.
As far as I know, the transformer was off-load when it went bang. On replacement, the system now works fine.
All very odd.
Cheers, Pete.

[turretslug]

Originally Posted by Tractorfan

small globules of copper scattered around.

Failed insulation between adjacent secondary turns resulting in a high-current burn-up? Maybe, a winding lay-up problem lurking since new, and a period of thermal expansion/contraction was enough to eventually fret through the enamelling.

[duncanlowe]

I have once, many years ago had a torroidal fail. It was encapsulated so never found out what the failure was. It took out the mains input fuse, but also destroyed one of the STK integrated power amplifier chips and one of the voicecoils on the Wharfedale speakers connected to the dead channel. I did repair it at the time.

[kalee20]

Looks like an acute hot-spot. Not an overall overheating, symptomatic of an overload.

I'd second what turretslug has said - insulation failure between turns due to a latent weakness.

[ajgriff]

Could have been shorted turns caused by inadquate insulation of the mounting plate or even overtightening thereof.

Alan

[Ed_Dinning]

A very slight possibility is exceeding the Curie temp of the material due to being in an enclosure and possibly under compression. It happened a lot with ferrites in the early days of SMPS as well as some TV LOPTs

Ed

[Joe_Lorenz]

Well,

at the first look I thought "What a junk". At least I feel that there is much material of very poor quality marketed nowadays. Just my point of view?

[Diabolical Artificer]

I've seen similar failures mainly down to bad manufacture, EG bad insulation and soldering around taps.end of windings.

Andy.

[arjoll]

I've had two toroidals fail - one in an RVR PJ300M FM broadcast amplifier which blew a hole in the side of the transformer, and another in an early revision of the Mackie SRM450 active speaker. I suspect the RVR failure related to moisture ingres during a few months of disuse, while the Mackie failure is apparently common and related to the quality of a batch of transformers. Current models use switch mode and are more reliable.

When I was but an apprentice the task of measuring the inrush current of a toroidal transformer was given to me. Knowing that turn on at the zero crossing point was the best thing to do to promote surge I used a (large) zero crossing solid state relay to energise said transformer.

Oscilloscope with current probe and the X output to turn on the switch at about 2Hz, lovely trace of the surge current, the repetitive surge current. Lots of the other labs personnel came round to see this, I was proud until... BANG, and the transformer looked like a smoking hedgehog.

[Techman]

Now that transformer's a 'pussy' compared to the one that had a catastrophic failure just below my feet where I used to work around a decade or so ago

Below is a picture of the transformer that went bang in the basement of the building I used to work in - picture I took after it was pulled out. The other pictures are of the generator that kept us going running 24/7 for a week or so while the transformer under the building was replaced - one picture taken from above from one of the windows.

Now that's what I call a proper transformer failure - note the hole blown in the top of the casing! And also, that's what I call a proper portable generator

[Andrew2]

My younger brother made an amplifier for his bass guitar from a pair of ILP modules and a fat toroidal transformer and big rectifiers, all supplied together as a kit. At the very first switch-on, BANG!
Upon investigation, there was a large black hole in the side of the transformer, everything else seemed fine.
So he sent the transformer back for investigation and he got a new one a few days later with the explanation that it was one of a faulty batch that had been 'wound too tight'.
New transformer in and BANG! Same again, new transformer. This time all was well and remained so for years.

[duncanlowe]

I have seen a different kind of failure in a transformer, though not a torroidal. It was a standard former based one, that was 240 to 12, but had an extra 5v manually wound on. It was built into a power supply by another trainee, and when he switched on the fuse popped. So he tried a bigger fuse which also popped. The next size fuse actually went BANG when it vapourised. Yes, just glass vapour inside the fuse holder. He asked me if I could figure out what was wrong, and I suggested that connecting the 5v secondary to the mains was probably not the best idea. I'm guessing the 'output' side was generating a lovely HT supply for the fractions of a second until the fuse ceased to exist.

[McMurdo]

I have to change the odd toroidal tx from time to time, but I see quite a few of them in various bits of pro audio. Mostly they fail with no other visible factor, for example going short or flashing through, mainly from far-eastern equipment.

re post 9, all the chinese mackies suffer toroidal failure (s/c) including the bass cabinets, the switchmode ones just blow up.

[kalee20]

Originally Posted by merlinmaxwell

When I was but an apprentice the task of measuring the inrush current of a toroidal transformer was given to me. Knowing that turn on at the zero crossing point was the best thing to do to promote surge I used a (large) zero crossing solid state relay to energise said transformer.

And not only switching on at zero voltage crossing, but also switching OFF at zero crossing, and switching on again at the start of a half-cycle of the same polarity!

Toroidal transformer cores are pretty good (or pretty bad, however you look at it) at retaining reminant magnetism. Certainly till the next day. Of course, this is how magnetic core memories work.

[trobbins]

Tractorfan, do you recall what mains side over-current protection was at site (either with the transformer, or distribution CB?

Any idea of the likely VA rating of the transformer?

[Radio Wrangler]

A curie temperature problem would have put up the current in the whole primary and so the cooking would have been general and not so focused in one area. The failure here is most probably of the enamel, creating a small group of shorted turns.

Secondaries on toroids don't get terribly neatly wound unless an awful lot of care is used and you get wires crossing wires, which form high pressure points on the enamel. Some fretting, over tightening or just excessive clamping pressure can be enough to trigger this sort of failure.

We have much better enamels nowadays than we had in vintage transformers from the thirties, but we do have to use them. I've come across a few 'budget' toroids which had been wound with solder-through enamel rather than the tougher stuff that's better for power transformers.

If the transformer is stripped, it may be interesting to try soldering through the enamel just to see - and have a look to see how well the winding is organised in terms of turns crossing.

It may just have been the site of a hard knock at some time in its past. Someone may have dropped it!

David

[daveobuchanan]

Scared now, have one in the roof for the power shower. Maybe I should put it in a metal box next time I go up there...

[Techman]

Absolutely! I certainly wouldn't ever want to risk having something like that operating in an exposed state in a loft space - was it a professional installation?

I have to say that I have never had this type of transformer fail, but then I tent to regard them as fairly modern, so have experienced far more of what you might call the conventional type of transformer. I've had a few conventional transformers that have failed, but none of them spectacularly, just smoke and burning. The one that failed in the old substation under the building at work, probably dated from the 1950s and I missed the actual bang as I was working four floors above it - so yet to have the pleasure (or not) of a proper failure!

[daveobuchanan]

Originally Posted by Techman

was it a professional installation?

well of course.... Kind of assume that people know what they are doing and following guidelines etc. Then again it seems only recently discouraged plastic fuse boxes.