RCD trip/Earth leakage

[333rocky333]

You can get an external isolater, or an external mainswitch in addition, but you cant usually remove the internal mainswitch, I think a CU also has to have one inside to conform with Type Testing.
Distribution boards like in commercial use are allowed to not have a main switch in the board

 

[ericmark]

I was also noting that the LED GU10 has no earth so power in should always equal power out. I did wonder if the electronics could cause an imbalance without a third connection but can't think of any way they could. But so easy to test, just remove the bulbs, if no change then nothing to do with bulbs it must be something within the fitting.

There are a number of ways power can leak, one is of course a item which is faulty and allowing current to flow to earth, the others are capacitive and inductive coupling run a live wire next to an earth wire with an AC circuit and some current must transfer to the earth wire through the capacitive and inductive coupling however this is normally below the 15 mA at which a RCD should not trip. The idea of splitting into circuits to reduce the capacitive and inductive coupling current to below 15 mA is valid. However to split in order to allow fault current to flow is not really the way to solve the problem.

When we look at leaks the main route is carbon tracks. At some point something has got damp then the current starts to flow and builds up a carbon track and once this has happened it's hard to stop. It is very unusual for current to get through insulation so we are looking at the ends of cables not along its length.

So likely causes of carbon also includes over heating normally due to bad connections. Although now using LED it is possible before the LED lamp was fitted the tungsten lamp had a poor connection and produced carbon which has in turn caused some tracking and hence the earth problem. In the main these problems are visible so you have worked out what is causing the fault now it's a case of both testing by removing lamps and inspection looking for any black marks which are likely carbon tracks.

 

[RMIM]

Gu10 lamps are only 2 pin hence there is no earth connection to the lamp, are the fittings earthed as most I have come across are double insulated?
Trying to move the circuit to another RCD doesnt fix the problem, it only masks it.
It is far better to get the issue fixed than to try and work around it. It could be a damaged cable, water ingress etc causing it

Thanks for the reply Spark123.
I see your point on the 2 pin lamps - that saves me throwing them away, I kind of like them.

Yes I agree masking the problem is not the way to go - so the next step is? to remove the bulbs are do an insulation test? [I want to see my meter read less than 1M ohm, so far it has not done much]

 

[RMIM]

You can get an external isolater, or an external mainswitch in addition,

Thanks for the reply rocky

 

[Echo the husky]

Less than 1MΩ is terrible! You want as high a reading as possible not a low reading!

 

[RMIM]

I was also noting that the LED GU10 has no earth so power in should always equal power out. .....

Thanks for the reply ericmark - I guess the next step is to pull the lights out and inspect the wires.

 

[RMIM]

Less than 1MΩ is terrible! You want as high a reading as possible not a low reading!

No I was saying I want my meter to read a fault. So far everything I have tested checks out ok. I spent a lot of money on that meter and if it's only going to tell me everything ok, well that's kind of boring.

Hopefully when I test the light circuit it will read under 1M indicating that's the problem.

 

[Echo the husky]

oh ok!