Creative light switch wiring

[timhoward]

I was working with a builder I know well today and he'd 2nd fixed some light switches.
All of them were spot on until I came to the switch either end of the kitchen to control the downlights.
My first reaction was "wrong switch - that's an intermediate - it won't work."
He said "Oh - I got it working!".

So I drew it out and had to concede this does actually function:

Then it hit me that all he'd done was essentially abbreviate this (as if there were a 3rd switch in one position):

I had to admit I'd never seen this one before and he in turn admitted it was totally trial and error on his part!
I thought I'd share this so anyone else who's led a completely sheltered life and not experienced this one wouldn't look as silly as I did.
(I changed the switch anyway, it requires too much thought left like that!)

 

[Pretty Mouth]

But what about my single-core cables? How do both line and neutral get to all the switches?

Singles? Could it be that the circuit is using earthed conduit as a neutral?

 

[littlespark]

is the lines marked "supply" and "loop" just singles? How can you have a line and neutral if theres not one supplying? Unless Prettymouth is right and the conduit is being used as return path.

 

[Lucien Nunes]

Conduit possibly, but I actually had in mind a metallic-sheathed cable such as Stannos, which was used and recommended for earthed-concentric wiring. This was true TN-C where there is no separate earth and neutral anywhere, just a combined PEN via the sheath. It was a requirement that all fittings had threaded spouts, into which special bushes were screwed with the sheath soldered in. So every fitting was earthed and neutral connections were made to flyleads from the bushes, much like an MI earth-tail seal. It was very compact, a light could be fed from a cable around 4mm diameter overall.

 

[littlespark]

i kinda knew it wasnt going to be something straightforward.

 

[freddo]

I came across this a year or two back. T&E feed to switch, then a 3C&E up to light, and down to far switch. I then recall there was loop feed out for another switch as the 3 core was carrying a perm live and neutral.

I've seen similar lighting wiring with supply cables looped at the switch connected as below, with the neutrals joined in the 2 way terminal. This worked fine for a few months but failed after it started flashing over and would sometimes trip the MCB when the light was switched off. The switch soon failed after the tripping started. The customer had replaced a Hager light switch with the built in neutral loop terminal with a normal (cheap) switch.

 

[freddo]

But what about my single-core cables? How do both line and neutral get to all the switches?

Here is my attempt:

I've only seen sheath return wiring once the circuit was mostly gone, just a few feet of 1 core imperial MICC feeding a 5A round pin socket. A neat system, shame it can't be used in most installations.

 

[Lucien Nunes]

Here is my attempt:

That is exactly what I had in mind. I try to encourage people to look at and think about some of these obsolete ideas to better understand why we do what we do, the way we do it. Earthed-concentric is neat and can be safe if the workmanship is sound, but cannot realistically be RCD-protected and is therefore obsolete for the wiring of buildings although standard for DNO work.

Small MI is great for concealment and single-core obviously makes the ideal use of its overall cross-section. There was also a reduced-diameter twin core 250V grade; a short-lived idea to make MI more convenient and economical for domestic work, only available in a few sizes. I'm trying to remember the smallest cable, possibly 0.187 ins. diameter for 2x 0.0015 sq. ins. conductor. If anybody has 'The Electricians Mate' to hand, please could you check that as I'm not near mine. Also what was the smallest overall diameter of single-core 440V cable, such as could have been used for an EC circuit? (there was no single-core 250V).

failed after it started flashing over and would sometimes trip the MCB when the light was switched off. The switch soon failed after the tripping started

Arcing L-N is another hazard with the Carter system and indeed anything that switches between supply poles under load (e.g. motor reversing.) OK with long-break tumbler switches and contactors but more likely with micro-break light switches.

 

[freddo]

Small MI is great for concealment and single-core obviously makes the ideal use of its overall cross-section. There was also a reduced-diameter twin core 250V grade; a short-lived idea to make MI more convenient and economical for domestic work, only available in a few sizes. I'm trying to remember the smallest cable, possibly 0.187 ins. diameter for 2x 0.0015 sq. ins. conductor. If anybody has 'The Electricians Mate' to hand, please could you check that as I'm not near mine.

So close! 0.182 ins. (4.6mm) Tiny considering its 16A rating.

Also what was the smallest overall diameter of single-core 440V cable, such as could have been used for an EC circuit? (there was no single-core 250V).

0.157 ins (just under 4mm) diameter for 0.0015 square inch cable, rated 20A when exposed to touch and used for concentric wiring.

 

[Lucien Nunes]

Thanks @freddo.

For the 'younger reader', the sizes of imperial TRS and VIR singles were given as stranding and diameter, e.g. 1/.044 (one strand of 0.044 ins diameter) which is almost exactly 1.0 sq. mm. But MI is always solid-core and the size was given as CSA, with the equivalent to 1/.044 being 0.0015 sq. ins. Then as now, this was the smallest cable in general wiring use.

OTOH the matching of cable, glands and seals was done by overall diameter, not by number and size of cores as it is today (e.g. 2L1 cable requires RGM2L1 glands). Instead you would order glands with a number indicating the cable diameter in thousandths of an inch. So for the 2x 0.0015 250V cable mentioned above, the diameter was 0.182" requiring a gland size 182. You will find this 3-digit number on the cap of all imperial MI glands.

Going back to the Stannos cable I referred to earlier, which was an early metal-sheathed cable designed with EC applications in mind, I think it was around in the era of cable sizing in SWG, so the nearest core size was 1/18 (one strand of 18SWG) which is slightly larger.