Diagnosing cause of RCD trip

[terryplumb]

Yes ,the CPC ( earth conductor) that would normally be bare copper within the grey sheath isn't connected or visible. Which may not be a problem( if the fan is double insulated its not required ). Beneath the fan is a cable going into white accessory box ( switch ,or junction box maybe ) I would be checking that internally for water ingress. I think the disconnected conductors showing in the fan are unlikely to be the problem ,unless they were showing any signs of water being in contact with them.

 

[conny]

In the first picture there is a switch/pattress box under the fan with a cable going to the bottom left corner of the picture. Is this cable feeding the fan?
Is this cable showing a dark scorch mark or is it a shadow from something? It is right at the left corner just before it disappears from view.

Just had another look and it may simply be a kink in the cable causing a shadow but is that 'insulation' tape inside the box over the cable gland? If so, why is it there? Does it cover a nick in the insulation and why do you have the same tape over what appears to be a joint in the red cable?

 

[SimonH2]

That unsheathed red wire should not be there - it should either be one core in a sheathed cable, of 3C+E should have been used. Ideally 4 core flex should have been used to connect the fan. The grommet cannot seal against a combination of cables like that, and it's questionable whether it could seal against a flat cable (eg 3C+E) - if properly pierced, it will seal against a round flex.
it also looks like white T&E comes out of the backbox, but then turns into grey T&E - suggesting a bodged joint.

The white accessory (is that the isolator switch ?) below the fan is also suspect - if water dripped onto the fan terminal box, then it could have dripped into the accessory.

 

[oldbutnotdead]

If there's moisture in the wiring space then there'll be moisture under that PVC tape- and who knows what other nasties. Hopefully everything terminates in that white surface box below- half a metre of 4 core flex and I don't think you'll get moisture in the fan innards again. Wouldn't show that pic to the suppliers- they might rescind their offer of a replacement fan

 

[SimonH2]

Also, what do the instructions say about mounting orientation ? It is generally considered good practice to have cable entries at the bottom so that any water on the cable runs away from the device rather than relying on the gland to stop it getting in. And the label suggest that the unit is upside-down.
But I understand that this may well put the terminal box on the wrong side for access - but just putting it at the top would help.

 

[u33446z]

SimonH2 - I know the wiring isn't very pretty. The fan I installed is a straight (but different) replacement for a fan previously installed by a 'professional'. I needed to extend the wiring to reach. I take your point about the cable through the gland at the top. I shall replace that with 2C+E round cable (the fan is Class 2 insulated so doesn't need an earth). However, there is no evidence of water ingress through that top gland - the loft space is quite dry.

Mounting orientation is permitted to be any angle within the top 180 deg.

 

[u33446z]

oldbutnotdead - the LED lights have got an earth. If one has failed N-E should it be identifiable by a zero ohm reading across the N+E terminals?

But what bugs me is why such a failure would occur in a lamp < 12 months old?

 

[SimonH2]

I shall replace that with 2C+E round cable (the fan is Class 2 insulated so doesn't need an earth).

Are you aware of the rules regarding how you may use the green/yellow core for anything other than an earth or bonding connection ? It would be better to use 3C+E (ie 4 core) cable and tuck the earth wire out of the way - cutting the earth wire short is "rather annoying" when you come to fit something different in the future and find that it's been snipped off. That would avoid needing to know what the rules for that are.

But what bugs me is why such a failure would occur in a lamp < 12 months old?

As the saying goes, "s**t happens". While rare, it's not unknown for stuff to fail out of the box or in the first few minutes/hours of use. Generally failure rates follow a "bathtub curve". Starting at the tap end, you get initial failures as any latent weaknesses (or "Friday afternoon" manufacturing) show up, then a sharp drop once these have gone. After that, there's a long period of few failures, before age catches up and at the non-tap end the rate starts to rise. The shape of the curve for any particular type of device varies, but they all generally follow that sort of trend.
There's also the matter of what the environment is like for the light fitting. Above it is a cold loft, so it will probably get cold overnight. Then someone takes a shower and the cold fitting is subjected to warm damp air - and if it's not well sealed then it will get damp inside. Damp plus electrickery is not a good mix

 

[u33446z]

Are you aware of the rules regarding how you may use the green/yellow core for anything other than an earth or bonding connection ?

I wasn't aware, but I am now, thanks. 3C+E it is then!