RCD tripping when E7 switches on

[problems]

Firstly i'll point out that i'm the property owner and the installer of the majority of the electrical fittings in the building, including Main & E7 DB's.
now the history lesson... sorry, but it is relevant
property is 230 years old, I have owned it for 42 years. it was semi derelict at purchase, and i completed most of the rebuild including plumbing & electrics. (i'm not formally qualified in either trade)
building was wired to the existing standards (1978) at the time. and i opted to include E7 as the means of heating.
The system was checked and signed off by a qualified engineer, and the 'electricity board' engineer when meter was fitted.
The E7 DB connected to 6 individual points and 6 heaters of varying capacity were fitted. and the system worked well and reliably.
10 years after installation, i opted for a different heating system and the ESH's were removed, and that side of the system became redundant, though DB and points remained.
approximately 3 years ago the meter was changed, and, as far as i was aware, everything worked ok.
had the odd rcd trip over the years, don't you just love rogue tripping!, on 2 occasions traced to slugs in sockets in the conservatory would you believe... also damp in a JB .... apart from that no probs.

a month ago She Who Must Be Obeyed, decreed that we should have an ESH in the living room and another in the dining room (we won't go into the why's and wherefores right now, i'm just the labourer round here )
so.... ordered 2 nice dual input ESH's (Newlec, but Dimplex by another name), and duly fitted them....
i noted that they required a second (peak) supply into a fused DP point, so put them in as well.

connected the whole shebang up, all seemed great, until the next morning when i noted the rcd had tripped overnight., as E7 was still live, i isolated both ESH's and reset the trip, fine , stayed set ok.
switch on either ESH and the rcd trips.

well, seems i've cocked something up, so, check and recheck all ESH connections... all done by the book, and no obvious errors.

next i make up a cable to connect NSH direct to 24hr supply ..... works fine, no tripping, left for 2 hours for heater to warm up... all ok.

ok, the E7 DB is forty years old and running on 20a cartridge fuses and 60a mcb, (none of which are tripping by the way), so i fit a nice new DB with 100a mcb and 6 20a mcb's ... it would have had to be changed sooner or later anyway.

and we still have the rcd tripping! ....

nothing else, just the main rcd, and if the mcb's for each ESH are isolated then all is ok... and that is just removing live from the circuit, so what can be causing the problem? .... the only thing that has changed from when the old ESH's were removed to when the new ones were installed is the METER ...

going to need some help here guys....

 

[marconi]

https://www.dimplex.co.uk/sites/default/files/assets//Quantum Series F Installation Instructions Issue 1.pdf

If the Newlec/Dimplex ESh with day boost is like the one in the reference, the notable thing is that all functional switching of the elements is done on the neutral side only. See page 16/16. And as we know RCDs 'prefer' neutral switching as make first or break last with respect to the line make/break switching. This could be a factor in the problem.

 

[Deleted account]

having looked at Marconi s wiring diagram above.

long shot.

could you have wired the 2 standard and night supply neutrals together instead of in there individual terminals.

this would explain your system working when transferred into the standard 24 hour board.

if you had done the above then the return night current would have an alternative path through the daytime neutrals Cu. Thus the imbalance tripping the RCD On the main Day CU.

Is the RCD a double pole isolator or single pole isolator. If it’s a single pole I would guess the night storage heater would continue to work and be warm in the morning. Unless the night tarrif CU is protected by another RCD.

 

[Lucien Nunes]

as soon as i tried to make the mcb for either heater the rcd tripped.

When referring to RCD trip events, please could you be explicit as to which RCD trips, as I believe there is now one in each CU. It was confusing at first because I think you typed MCB a couple of times where you meant RCD and I have lost track of which device was tripping in the early stages when the old E7 CU was in use. I read it as the RCD in the peak CU tripped when the supply from the off-peak side was energised but perhaps that was not what you meant.

If in fact the RCD on one side is tripping when the other side is energised, and there is no cross connection, one possibility is that there is an existing neutral-earth fault on the side that trips, that only causes enough leakage to trip when the heater's control unit switches on load in response to the other supply coming on.

 

[problems]

thanks for your enquiries ..
the property is earthed by a rod
there is only 1 rcd (marconi.... one of the things i am COMPLETELY ok with is computers, having worked in IT for 50 years, so i won't be needing a teenager from next door to help me out )
here's a pic of the rcd (was changed from voltage to current 15 years ago)
luciens... only the one rcd ... as per pic, (apologies for the 'cut out' legend.... thats for SWMBO in case i'm not at home and it trips for some reason, like when she dropped the steam iron on the floor a month ago )
the untidyness in the wiring is due to ongoing 'mucking about, trying to fix it' work.
today i'm going to remove the E7 (live) tail from the meter (pic of meter to follow) and stuff it into the main JBox (i.e. the E7 CU will now be connected to the 24hr supply, and in all respects the same as the main CU.)
havn't had so much fun since i broke my leg ....

 

[westward10]

Need a wider picture of the full set-up. Something doesn't seem correct, assuming the service terminal block is outgoing from the rcd it has five conductors connected which seems wrong. It appears one conductor bypasses the rcd.
Is the meter above left of the rcd, I suspect the line to your E7 is not rcd protected but the neutral is going through the rcd.

 

[problems]

hi westward thanks for your reply
yes, the meter is above left of the rcd.
the 5th (far right) connection on the meter is the E7 live o/p, this goes direct to the E7 CU.
so what you see at the service JB, as you correctly point out is live & neutral from the rcd, live & neutral to the main CU, and lastly neutral to the E7 CU
[automerge]1573126646[/automerge]

Need a wider picture of the full set-up. Something doesn't seem correct, assuming the service terminal block is outgoing from the rcd it has five conductors connected which seems wrong. It appears one conductor bypasses the rcd.
Is the meter above left of the rcd, I suspect the line to your E7 is not rcd protected but the neutral is going through the rcd.

you're quite correct, and this was how it was left after the new meter was fitted.
are you suggesting that i fit a second rcd for the E7 circuit, so that (as far as the E7 circuit is concerned) both live & neutral go via an rcd ?

 

[westward10]

That is your problem. The line and E7 neutral are creating an imbalance as one uses the rcd while the other does not. If you have a TT system and that is the only rcd then the E7 circuits a not likely to have fault protection. The E7 needs a separate rcd with its neutral taken from before the current rcd.
[automerge]1573127713[/automerge]
This is the set up you require.

 

[problems]

That is your problem. The line and E7 neutral are creating an imbalance as one uses the rcd while the other does not. If you have a TT system and that is the only rcd then the E7 circuits a not likely to have fault protection. The E7 needs a separate rcd with its neutral taken from before the current rcd.

ok, that makes sense to me... and easily done, apart from taking the neutral prior to the current rcd .... there is no available (extra) connection point, unless i fit a JB into the incoming neutral line ?

will nip off and get required bits...

 

[westward10]

You need another single pole service terminal block. At this point you need the services of an electrician as you are touching wiring which strictly speaking can only be isolated by an authorised person.

 

[Deleted account]

You need another single pole service terminal block. At this point you need the services of an electrician as you are touching wiring which strictly speaking can only be isolated by an authorised person.

With all the past experience he has and the advice he has received from us I don’t think he will take your advice, as good as it is.

 

[Lucien Nunes]

It's a pity we didn't have that pic to start with. The OP did stress that the only thing that had been changed recently was the meter, but I suppose we all overlooked something so blatant as the tails being connected up wrongly, which could never have worked.

 

[westward10]

The E7 line which bypasses the rcd does seem an alteration, black cable tie fitted. I suspect this has nothing to do with the meter change. I suspect when the rcd was fitted to replace the voltage operated device some time back the E7 was no longer in use so no problem was evident.

 

[problems]

ok guys, thanks for all the advice, and I will be taking it. i have all the gear (and some idea ), so i will make the connections today, and if it all works ok, i will still leave the appointment for local sparks to come monday pm.... and he can check it over for cockups and poor workmanship, and sign it off ..... (hopefully) ...

i'll post results tomorrow ... should i still be alive ....

 

[Lucien Nunes]

It's naughty, whichever way it came about. With any load on the E7 board, in the event of a trip the house blacks out, but every single point is still live, both on the E7 and the 24h and there is more risk of shock with the RCD off than on. It's a good thing it was self-incriminating.

 

[problems]

It's naughty, whichever way it came about. With any load on the E7 board, in the event of a trip the house blacks out, but every single point is still live, both on the E7 and the 24h and there is more risk of shock with the RCD off than on. It's a good thing it was self-incriminating.

it's a good point.... but, as it stands, with the rcd tripped, the only live would be to the E7 cct, both live & neutral to the 24h cct would be isolated.
[automerge]1573134350[/automerge]

That is your problem. The line and E7 neutral are creating an imbalance as one uses the rcd while the other does not. If you have a TT system and that is the only rcd then the E7 circuits a not likely to have fault protection. The E7 needs a separate rcd with its neutral taken from before the current rcd.
[automerge]1573127713[/automerge]
This is the set up you require.View attachment 53677

great!.... got the bits, just consuming last supper (well, lunch actually) and i'll be at it !