EICR Code for metal light switches with earth to backbox

[Dartlec]

Posting this more as a discussion than seeking advice, as I've already fixed the issue.

EICR on a house which at some point had been 'done up' with metal face plates for all the switches. When checked with a wander lead R2 was within spec on the ones I tested, but when taking a random one off found the earth to the back box only - so then checked them all and not a single one had been brought forward to the nice obvious earth terminal on the switch :facepalm:



I remade all the earths to the front plates anyway in this case, but was wondering what code people might give this if they found it on a job where they could not do remedials? It's sort of the opposite of the old argument about needing to earth the backbox or add a flylead.

Would people consider an adequate reading of R2 to the front of the switch when correctly installed as not needing a code? Or automatic C2 once you saw the lack of cable to the plate terminal?

In this case, one of the back boxes was metal with insulated lugs anyway, so that would have been an immediate C2. But I remade them all anyway as I was also working in the place.

Does show the limitation of R2 checks on everything, where you aren't removing every plate. Didn't help in this case that the bodger who'd done it had half buried all the plates into the wall to discourage ever removing them!

 

[dodger421]

That’s a good one!

I’d have done the same as you and just remade them there and then, but if I absolutely couldn’t for some reason I’d lean towards a C2 I think. Simply because you can touch the faceplate in the event of a fault and you’re relying on the fixed lug of the back box and the screws to provide continuity to the accessible exposed conductive part.

I know you can rely on the faceplate screws and fixed lugs and do away with a flylead from switch to backbox, and I think that’s because if that continuity was interrupted you still can’t access the now live (in the event of a fault) backbox whereas you can always access the faceplates.

So yeah, good R2 readings aside I think the seriousness of a failure in that screw/lug continuity and a fault developing warrants a C2.

 

[Ian1981]

If the switch was earthed through the back box and the R2 readings where low then I wouldn’t code them at all.
I don’t code metal 3 phase DB covers that are earthed through the fixing screws after all.

 

[westward10]

I have seen this many times in the domestic sector and I just reconnect the cpc from the back box to the front plate whether I am doing remedials or not, saves the paperwork. Not sure I would be using a crimp to a back box fixing screw.
Same with reversed polarity at the odd socket I just correct it.

 

[dodger421]

If the switch was earthed through the back box and the R2 readings where low then I wouldn’t code them at all.
I don’t code metal 3 phase DB covers that are earthed through the fixing screws after all.

That’s a good point, I hadn’t thought of DB’s and CU’s. I think I’ve seen a couple of them with flyleads between case and cover but they might have been purpose built panels for machines and plant equipment.

 

[buzzlightyear]

No code, earthing from screws to the back box . Has for the crimps adlest is earthed to the back box.

 

[Baddegg]

As above surely assuming the back box is fitted properly and doest have nylon lugs it’s all one anyway?...

 

[telectrix]

C3 in my book (if the switch were to be loosened off,say for decorating, then the faceplate would have lost it's earth) .except for the plastic lug box a C2.

 

[Ian1981]

C3 in my book (if the switch were to be loosened off,say for decorating, then the faceplate would have lost it's earth) .except for the plastic lug box a C2.

So you would also C3 sockets that don’t have a fly lead earthing the back box?
If switches are loosened off then exposed conductive parts are the main issue, you can’t cover stupidity, in its normal service the switches are fixed back and earthed, no code

 

[westward10]

Point is the metal front plate requires a cpc not visa versa so whether it is earthed by way of the back box it is a requirement to fit a cpc to the front plate, it needs Coding.

 

[Ian1981]

Class 1 exposed conductive parts require earthing, if it’s earthed In situ then it’s no code.
best practice is to earth the plate directly however i dont see the issue if it’s earthed full stop by one way or another.
The regulations will not tell you how to earth it other than it requires earthing.

 

[westward10]

The instructions with the accessory will demonstrate a cpc to the front plate.

 

[Dartlec]

That’s a good point, I hadn’t thought of DB’s and CU’s. I think I’ve seen a couple of them with flyleads between case and cover but they might have been purpose built panels for machines and plant equipment.

Yes I hadn't considered that - Fusebox CUs do come with a flylead to the front case - but I've not seen it on other ones I've installed recently - and some of them don't have the best screw fixings *cough* BG and MK *cough* in my experience.

Though most CUs are coated steel, so I've never actually checked whether they are normally conductive when in good condition

 

[telectrix]

the reasons i would give it a C3 is that it does not comply with MI ( as westie pointed out), and if the screws were removed for decorating or whatever, a fault on the switch would not trip MCB. therefore it "is improvement recommended. ", because i say so (smiley).

 

[Dartlec]

Class 1 exposed conductive parts require earthing, if it’s earthed In situ then it’s no code.
best practice is to earth the plate directly however i dont see the issue if it’s earthed full stop by one way or another.
The regulations will not tell you how to earth it other than it requires earthing.

The instructions with the accessory will demonstrate a cpc to the front plate.

Very interesting discussion, and I can probably see the validity of every view. Thats the 'joy' of doing EICRs to a consistent standard.

My concern would be the worry that screws might come loose or a lug break with constant use of the switch (something that doesn't happen with CUs so much).

but if an R2 to the front plate is satisfactory, then it is probably acceptable to consider that effectively earthed without necessarily needing to remove it to check the earthing connection....

However, if I remove a front plate and find the earth not to the plate, personally I'd still feel duty bound to correct it there and then to avoid any further thought.

 

[westward10]

If I am in a domestic situation and there are metal light switches I always check them, it will be 50/50 if they have a cpc to them.

 

[Ian1981]

All earth wires must be terminated to the back box.
An example that not all MF’s instructions are the be all and end all.
Pick shows the cpc from the cable not directly to the switch.
This does however highlight that a fly lead is required

 

[Ian1981]

the reasons i would give it a C3 is that it does not comply with MI ( as westie pointed out), and if the screws were removed for decorating or whatever, a fault on the switch would not trip MCB. therefore it "is improvement recommended. ", because i say so (smiley).

I’d be more concerned with the exposed conductive parts exposed by removing the switch personally.

 

[Pete999]

All earth wires must be terminated to the back box.
An example that not all MF’s instructions are the be all and end all.
Pick shows the cpc from the cable not directly to the switch.
This does however highlight that a fly lead is required

It;s a M/C switch, terminate the cpcs into the back box and run a fly lead to the plate switch, how much would that cost you? not much I imagine then no need to code

 

[Midwest]

My two pennyworth; I've always terminated cpc into back box, then a fly lead to faceplate. My thinking is if cpc become disconnected when plate is removed and/or replaced, is just that particular plate, not the rest of that circuit. I've even taken it further with multi gang switches, and put all cpc's in wago, then flylead to back box and plate, for the same reason, and also you can't successfully terminate multiple cpc's into one back box terminal.