Has anyone had an RCD fail as spectacularly as this one?

[ProDave]

Today I was rewiring a 1 way shower CU (moving from normal supply to total control supply to get the cheap rate). I did not test the CU before disconnecting it.

Re applied power. Turned RCD on. Power comes out of CU. Turn RCD off, power goes off, so it's working as an isolating switch.

Turned it back on again. Pressed test button. Nothing happens. Pressed test button again, nothing happened.

Pressed and held test button for about 3 seconds. Loud buzz, a flash and bad day explode

Sparks and smoke came out of the RCD. Fortunately it was a submain from the total control CU and it took out the 45A cartridge fuse feeding it. If this had been a main CU and the RCD had been acting as the incomer, then it would have been a bigger bad day explode and would have taken out the DNO's fuse.

I'm going to try and dismantle it later on to see if I can see what happened, but this is the first time I've had an RCD fail in such a spectacular way. Something must have gone short circuit inside. The shower was off at the shower isolator so was not drawing any load.

It was a Proteous 63A / 30ma 2 pole RCD acting as the incomer for this 1 way shower CU that's only really there to provide RCD protection for the shower.

 

[NozSpark]

I've ad a similar thing happen with a Proteus RCD...

I found out that they do not like it up 'em the wrong way, where as some other brands take it any way!

 

[ProDave]

It WAS connected the correct polarity.

 

[NozSpark]

Not just correct polarity...

was in connected to in(top) and out to out??

 

[ProDave]

Not just correct polarity...was in connected to in(top) and out to out??

Yes, definitely correct, in to top and L and N as marked on the RCD.

 

[sparkytim]

Had similar experiance with a cheapy Garage Unit RCD a few weeks ago, had been in a couple of weeks, had to go back to job and just tried the test button - and it buzzed then smoked poored out!

bad day explode

I suppose at

 

[PC Electrics]

I tried using a couple of those cheap RCD modules from Toolstation once - CED I think they're called. One simply stayed open circuit, the other burned out in about 10 minutes with a moderate load (quite smokey). Not as spectacular as your event, but I'm sure it would have scared the client quite well (especially after I'd sold them on the safety benefits of RCDs). Lesson learned though - don't touch the cheap stuff now.

 

[NozSpark]

Just a reminder..

Proteus have a 20 or 25 year warranty (parts only)

 

[lilman]

I carried out a PIR last week - Proteus 63A/30mA RCD. Failed 180 degree tests and kept tripping on no-trip loop tests. 2nd time I've had problems with this make/rating of RCD.

 

[baldelectrician]

Had the same thing 4 or so years ago with a CHINT rcd.

Pressed test, nothing happened, pressed and held - bang and took out fuse on sub circuit

 

[ProDave]

Just a reminder..Proteus have a 20 or 25 year warranty (parts only)

Too late. I've dismantled it to have a look.

Not much to see, apart from a lot of charred remains. Even alowing for this being a crappy camera, there's not actually much to be seen even when you see it for real, just a black mess.

 

[ProDave]

Had the same thing 4 or so years ago with a CHINT rcd.Pressed test, nothing happened, pressed and held - bang and took out fuse on sub circuit

Interesting. This is not a unique case then. I wonder what the failure mechanism is then?

The only thing that I can think of is the test button puts a high resistance load from phase to earth to create an imbalance. So if you hold the test button down and it doesn't trip, it looks like this test circuit fails in a rather spectacular way.

Now if the RCD is working and it trips, the test load will never be applied for long, but if the RCD has failed, the test circuit is not designed for activation for more than a few mS.

To my mind that's a major design failing. I wonder how many end users have blown their RCD incomer like this and knocked out the DNO's fuse?

 

[NozSpark]

Interesting. This is not a unique case then. I wonder what the failure mechanism is then?The only thing that I can think of is the test button puts a high resistance load from phase to earth to create an imbalance. So if you hold the test button down and it doesn't trip, it looks like this test circuit fails in a rather spectacular way.

Now if the RCD is working and it trips, the test load will never be applied for long, but if the RCD has failed, the test circuit is not designed for activation for more than a few mS.

To my mind that's a major design failing. I wonder how many end users have blown their RCD incomer like this and knocked out the DNO's fuse?

RCD test button puts a resistor across the outgoing L and incoming N to cause the imbalance..

:coat

 

[binky]

Protues RCD went bang on me about 5 year ago, pressed test button and it blew smoke rings out the cable entries. I've also had an MCB fail but leting 200v through. Glad to say neither item was fitted by me, and I have no intention of ever using this stuff:red card

 

[SLIPSHOD & SLAPDASH]

We fitted some 'Harok' boards for a few months which we filled with Rcbo's (rarely use duel Rcd boards). They are copies of Contactum. After the 6th or 7th Rcbo Blew up on energising we stopped buying them.