The on site guide says...

[cjard]

I'm curious about a bit in the on site guide where it is describing earthing of socket finals in high leakage current situations.. It states that ring finals must use sockets with two earth connectors, and that the loop in must be in one connector and the loop out must be in the other

I've wondered about this ever since I watched my spark wiring my house sockets in this fashion(I doubt it's a high current situation as I don't have large numbers of switched mode devices, but he's more often doing non-domestic work so maybe he just does it by nature of repetitive behaviour)

What advantage does it confer to introduce an increased number of joints/potential failure points into a single CPC run? I sort of understand the requirement to have them connected to separate terminals at the DB, increasing resilience in case one connection is poor/screw stripped etc(but you'd think that would be obvious) but I can't fathom why it's better to loop the incoming CPC into one socket terminal and the outgoing into the other terminal?

Reading around I found references to systems of two separate CPCs per circuit in high leakage situations (and I can see how using different socket connectors for that would be sensible), but the OSG diagram doesn't seem to portray that

 

[PrenticeBoyofDerry]

By terminating both CPCs in the same terminal port, if that terminal becomes faulty or cable loose, you will lose continuity at that socket. By terminating them singularly in separate terminals, that will reduce the risk of complete loss of continuity at that socket.
That is my understanding.

 

[ericmark]

As an industrial electrician I was taught one wire one terminal if you wanted two wires then have two terminals, that's how I would build panels I would never dream of building a panel where you had to put many wires into one terminal. But with domestic it seems that having up to three wires in one terminal is accepted.

With a radial circuit you would have two earth wires in each terminal using high earth current, only with a ring final do you have two earth wires with radials you would have 4 earth wires except for last socket. Since with domestic you have a RCD on the sockets it is very unlikely that you have any sockets designed for high earth currents, so using the two terminals is simply preference, with a ring final it really does not matter, my son and I fell out over twisting wires, I would always twist the wires together so they would have good contact even if the terminal was due to vibration to work loose over time, however he said how hard it was to test when the wires are twisted together and over time untwisting them to test could cause the wires to break. At the end of the day both methods have good and bad points it is down to personal preference.

As an industrial electrician I would want to see all terminals fixed, in the main I would use DIN rail with links as required on the tops of the terminals, no way would I simply put a connector strip in a back box with a blanking lid on it, yet it seems that was normal for domestic wiring of central heating? Since 2004 things have changed with domestic installation, but prior to then industrial electricians called domestic electricians "house bashers" and looked at them as being same as an electricians mate, in fact it was common to get semi-skilled guys to wire houses with an electrician employed to inspect and test. Fault finding would need an electrician but the wiring of new houses was pre-2004 often done by semi-skilled labour.

 

[cjard]

So relying on the ring pattern to maintain continuity to the other side? Radials suffer considerably (unless the CPC is in a ring, but then if you were going to the trouble of returning to the DB with a single core of earth, why not just return to the DB with 2c+e and wire a ring rather a radial. Not that I want this to turn into a ring vs radial debate)

Is the most commonly found pattern of connection failure in sockets, one where the screw has come loose and hence a pair of wires inside the connector no longer connect to each other or to the connector?
Is it common for a circuit to have more than one of such failures (thereby introducing sections of unprotected sockets) or is the chief concern a single outlet losing its CPC?

I can't see it being likely that the connector would come away from the earthing bar it's riveted to, so it seems that connecting CPCs individually addresses a situation (loose screw causes break in CPC ring) that has only a small chance of occurring, but doubling the number of connections would double the chance of it occurring..

Hence, it still seems a bit bizarre to me

 

[cjard]

Funny, my spark has used the "house bashing" phrase more than once..

And the twisting thing, that was prominent in my mind when pondering how to ensure a good earth circuit.. I reasoned that twisted wires in a single connector were going to be better than a single wire in each of 2 connectors, as a loose screw should then only cause loss of CPC at one outlet. Of course, crimping would probably be better than twisting but it'd be a hardcore house basher that did so..

 

[cjard]

Surely with the testing thing you can unlink one end at the DB and test the twisted wires without untwisting them? Or is your worry then that you're never really sure whether both DB connections are secure and providing earth when connected back up? But then wouldn't tat apply to any connected, twisted or no?

 

[JohnW2]

Is the most commonly found pattern of connection failure in sockets, one where the screw has come loose and hence a pair of wires inside the connector no longer connect to each other or to the connector?

One would think so - what other type of connection failure at a socket do you envisage?

Is it common for a circuit to have more than one of such failures (thereby introducing sections of unprotected sockets) or is the chief concern a single outlet losing its CPC?

One such failure is rare enough, so I would imagine that two on the same circuit would be very rare. The primary concern therefore presumably is, as you say, the loss of an earth connection at a single outlet. Having a ring with both wires in one terminal reduces that risk. Having a ring with wires in separate terminals reduces that risk further - although the price paid is, probably, a slightly increased risk of a socket losing one of its earth connections (although that would not matter unless there were a second fault).

Kind Regards, John

 

[Iggifer]

People who twist CPCs in sockets (or anywhere else for that matter) should be burned alive

 

[JohnW2]

People who twist CPCs in sockets (or anywhere else for that matter) should be burned alive

I'm old enough to remember the day when the same would probably have been said of people who did not twist conductors (whether CPCs or otherwise). I certainly did that for many years, believing (rightly or wrongly) that it was 'accepted good practice'. I suppose it's fair to say that it was also a time when any sort of routine testing of an electric installation (certainly domestic ones) was probably extremely rare.

Kind Regards, John

 

[Echo the husky]

Twisting several solid single cores together results in loose connections after a short time, as the copper creeps a bit. I replaced 4 sockets today (I did other stuff as well!) 3 had twisted wires and had very loose connections on all 3 terminals, one wasn't twisted and the connections were all secure. all 4 sockets were 1970s MK singles. My old boss gave me a hand one day and he twisted the CPCs before connecting them, they were all loose when I went to alter the wiring to them a week later.

 

[JohnW2]

Twisting several solid single cores together results in loose connections after a short time, as the copper creeps a bit.

In the days I was talking about, solid single core conductors were just a dream

Kind Regards, John

 

[Echo the husky]

Ah I see, did you ever get a bad back carrying all those reels of lead sheathed twin around?

 

[JohnW2]

Ah I see, did you ever get a bad back carrying all those reels of lead sheathed twin around?

I'm not that old - other than to remove it, I've never handled lead-sheathed cable!

Kind Regards, John

 

[Echo the husky]

 

[PrenticeBoyofDerry]

People who twist CPCs in sockets (or anywhere else for that matter) should be burned alive

A bit harsh! But fair!