RCD tripping

[electrified]

Rcd fused spur feeding extension from ring final circuit.

Goes to a 2 gang switch with 1 switch operating 3 downlights and another feeding extractor fan.

Zs, R1+R2 and RCD tests completed and are good.

Problem is that when you leave the lights off for a few days without touching them, when you come back to turn the lights on the rcd has already tripped don’t know when but is strange as there is no loads.

The only thing i can think of is that the rcd fused spur is faulty

Any thoughts?

 

[CBW]

What else is on the rcd circuit?

 

[electrified]

What else is on the rcd circuit?

it’s a lighting circuit, so it goes from the rcd fcu down the switch then feeds the 3 downlights with one switch and another switch feed extractor fan. nothing else on circuit, reasonable short too. Note that extractor fan has not been put in yet but downlights are.

 

[Wayners]

Undo and tighten all the terminals you can get to.
Switches and lights ect

 

[electrified]

Undo and tighten all the terminals you can get to.
Switches and lights ect

what do you think terminals are loose or too tight?

 

[CBW]

it’s a lighting circuit, so it goes from the rcd fcu down the switch then feeds the 3 downlights with one switch and another switch feed extractor fan. nothing else on circuit, reasonable short too. Note that extractor fan has not been put in yet but downlights are.

So it’s a single rcd in a single additional consumers unit?

 

[electrified]

So it’s a single rcd in a single additional consumers unit?

no it’s an RCD fused connection unit with a 5 amp fuse, that is fed from a ring final circuit.

 

[mikeyd]

A neutral to eatrth fault can still trip the RCD even with no load on that circuit.
It will likely happen when there is a relatively large load in use in the rest of the property.
The way I think of it is that a fraction of the current "returning" from the large load on the neutral side ends up going to earth through the fault in the rcd protected circuit, rather than taking the nornal route out through the incoming mains.
If that fraction of current ends up being over 30mA then RCD trips.
I'd be looking at the wiring to the unconnected extractor first.
I don't think Zs or R1+R2 would help with a neutral/earth fault. An IR test might tell you something. I'm not a sparky though.

 

[bernardgreen]

 

[electrified]

so you’re saying high loads that can leak current before the rcd can cause the rcd to trip even though it’s before the rcd?

 

[bernardgreen]

The "leak" is from Neutral to Earth via the fault. There is no fault in the load.

The fault provides a route for some of the current from the load to bypass the Neutral conductor in the RCD sensor.

 

[mikeyd]

so you’re saying high loads that can leak current before the rcd can cause the rcd to trip even though it’s before the rcd?

No, not really. We are saying that a fault after (or downstream) of the RCD may only trip the RCD when a high load is in use elsewhere in the property.

 

[electrified]

No, not really. We are saying that a fault after (or downstream) of the RCD may only trip the RCD when a high load is in use elsewhere in the property.

ah yes coming back on the earth through the neutral causing an imbalance. So if i megger N-E from the rcd this may find the fault?

 

[bernardgreen]

So if i megger N-E from the rcd this may find the fault?

Yes but first you need to disconnect the Neutral conductor from the Neutral busbar in the consumer unit,

Ensure main switch and the MCB are both OFF for your safety

 

[mikeyd]

ah yes coming back on the earth through the neutral causing an imbalance. So if i megger N-E from the rcd this may find the fault?

I agree with Bernard.

I would be very surprised from your description if it wasn't a neutral to earth fault after the RCD. Sure, it could be a bad RCD but you said it tested fine so I find that unlikely.
One thing not mentioned yet is that a neutral to earth fault usually means a low resistance or short between CPC and neutral wires, but it could also just be a neutral wire touching anything "grounded".
I would disconnect the lights and extractor wiring from the RCD and do a normal insulation resistance test between neutral and CPC etc. on all the cables downstream of the RCD and take it from there.