Ok....So I'm only coming back to this as Im genuinely stuck with it and for good reason. I am closing my account after this thread but may start up again using another name. As someone mentioned it was a bad move using my company name on a forum where everyone is far superior to you and every post you make it scrutinised in the hope that the OP is a Electrical Trainee or equally unqualified.
Ill ignore all drivel about how "easy" this should be to fix and how embarrassed folk would be visiting multiple times to this fault.
No one can say that without seeing the job. Its all very well solving theoretical faults from a computer but Id welcome anyone to come and subby for me for a day....Find this fault in 1 day Ill pay you 1k.
So...Ive been at it all day and my head hurts! Back on Monday with a fresh head and equipped with 25 RCBOs. Ill post my findings in case anyone sees something obvious Im missing...Im not ashamed to say I can't find it, and also not interested in drivel about how you would find it with ease.
This is a fairly new installation and one would assume all connected properly. The property, or mansion I should call it is very grand. Oak flooring throughout with no access underfloors therefore no way to physically trace the cables.
So, "Narrowed" it down to a "ring" circuit with around 10 sockets. It turns out this is not actually a ring.
Although its belling out (Fluke Test Lamps ) continuity is showing >2000MOhms on all cores (Metrel Test Gear ).
Test lamps are brand new and gear was calibrated last month although with my findings Im questioning whether test gear is working !
Disconnecting every socket in the circuit I Meggered both incoming legs LNE at 1000v, All good. I then Megger each internal leg at 1000v, all good.
I then connect all legs together and megger both sets of cables at the board - 1.1MOhms between all cores.
I measure one leg at a time with full ring connected - all good.
I set the megger up at the board and go through each socket, connecting as I go. I get to the last socket - all good. I connect the last socket forming "ring" and again 1.0Mohms, sometimes 12Mohms.
In 2 radials all good , make it a "ring" and fault occurs.
Now, there is another circuit which is showing an occasional reading of 9Mohms between L>E.
Remember this is only tripping every few days...not banging.
In this newly installed consumer unit, we also have a spare 2.5 floating around which is live and only isolated by switching the 16Amps Immerser breaker on ANOTHER board.
Im thinking that there is a really bizarre fault somewhere or a nasty junction box under the floor, which is not helped by the fact they have confused legs when installing the board. I think a leg of one ring has been mixed up with the immerser and perhaps another leg mixed up with another ring. Obviously not been tested after install. The guy who owned and built the house ran a large building firm. The consumer unit was a mess too, really badly done.
The customer has agreed that RCBOS is a good idea to stop nuisance tripping of all circuits but still, Ill need to find this fault.
A combination of a strange fault and poorly installed and unorganised unit.
Rewiring this circuit is not really an option. New build property, no access under floor, oak flooring everywhere. The house is gigantic. The ring in question is around 20m from the consumer unit.
Next plan of attack is ripping everything out and starting again at the consumer unit installing the RCBOS. Hopefully finding the correct rings will at least make further fault finding easier.
Decorators are in right now firing up wallpaper at likely a few hundred quid per roll. Everything is slow at the job, Carpet socks on and off , working around 5 other trades, while taking great care for the interior of this property most can only dream of.
So anyone up for a bet ? 1k for a days work ?
And please , if you are going to help read my full post before posting "its a N-E fault".
