Strange Neutral Earth Fault

[dunny123]

Its a TT supply with 100mA time delay incomer RCD, and ordinary 30mA RCD. The lights and cooker are on on the 100mA RCD side of the board, and the sockets on the 30mA side. I installed the board to replace a rewirable months back. All insulation tests >500M at the time. I got called back as they started to get a lot of nuisance tripping (of the time delay RCD) when using cooker, kettle, toaster etc. I subsequently tested and found a N-E fault (40ohms) on the upstairs lighting circuit, and with this disconnected everything is fine. I was not able to find any obvious faults in any roses/switches or fittings, but they have had rats in the loft which is overboarded so could even be a gnawed cable. What is strange about the fault is that the 100mA RCD only seems to trip (with lighting circuit connected) when you power on the cooker and items on the sockets side, which as I said on on a different RCD. There are no crossed neutrals in the board either.

I've got to go back and try and localise the fault, by breaking down the circuit, unless anyone can suggest a better method, but can't fathom why there seems to be a relationship between the two sides of the board ?

 

[holmslaw]

..

 

[Spark123]

When a high current appliance is in use the resistance of the cables etc will mean the voltage of the neutral will rise slightly above the potential of true earth causing enough current flowing through the N-E fault to earth electrode causing the 100mA device to trip. The 30mA device won't trip as it is still in ballance, the fault being downstream of the 100mA device but upstream of the 30mA device.

 

[TicklyT]

That sounds like everything is working correctly to me.

I wouldn't expect to find a fault on the 30mA RCD side of the board.
The phase and neutral currents will be balanced through the 30mA RCD, so that will not trip. The Neutral - Earth fault is 'upstream' of the RCD, where it cannot monitor currents.

However, it is 'downstream' of the 100mA RCD, which is protecting the whole board. As the load on any part of the board increases the neutral potential may increase enough for the N-E fault current to exceed 100mA.