Every day is a schoolday. House lighting circuits from the past.

[Trigon.]

Originally Posted by Lucien Nunes

All the fittings are earthed, too.

If that's a clue then perhaps earth is used as the return?

Live is wired straight through both switches, each lamp wire is switched between live and earth, so lamp is off when both are earthed or both are live?

Cheers

[Lucien Nunes]

Correct. That is called the 'Carter system' in which the two 2-way switches switch the two sides of the lamp between Line and Neutral. It works, but there are a number of objections of which the key one is that the lamp is fully live in one of the off permutations.

The second trick employed here is earthed-concentric wiring, where the metallic sheath of the wiring system provides a combined-neutral-earth (CNE) a.k.a protective-earthed-neutral (PEN) conductor at every point. (Note - this is a conductor but it doesn't count as a 'core' which is what I kept referring to above). Earthed-concentric installations were permitted in the UK with certain wiring systems, such as mineral-insulated metallic-sheathed and proprietary metallic sheathed types such as Stannos. Specific precautions had to be taken with regard to ensuring continuity, as an interrupted or high-resistance PEN can result in all exposed wiring, fittings and loads becoming live through the load resistance. I.e. a broken PEN is much more serious than a broken protective earth.

Earthed-concentric is one implementation, arguably the best, of TN-C configuration where the PEN extends all the way to the load. TN-C was permitted in many other countries and has been used as recently as 2001 in Finland. Under some regulations a PEN can be identified by colour, e.g. green/yellow with a blue band. In German wiring, a blue will be a neutral, a green/yellow an earth, but a grey could be either a simple neutral or a PEN. These days TN-C is only normally used in DNO's works up to the service terminal equipment, where the neutral and protective earth split as they enter the consumer's installation. Hence, TN-C-S (earthed neutral, part combined, part separate). Obviously full TN-C is not compatible with residual current protection.

I have a sample of Stannos EC wiring connecting up some light switches and a socket. The cable sheaths are sweated into brass plugs that screw into the conduit spouts of fittings. It was a requirement that only threaded spouts be used to maintain PEN continuity, not bushes and couplers used on plain holes in CI or steel boxes.

[duncanlowe]

Originally Posted by Lucien Nunes

Historically, a very common shortcut for 2-way switching of the upstairs light in a stairwell was to borrow the line from the adjacent lower hall light switch to feed the common of the downstairs switch, send a twin+earth with two strappers (L1 & L2) to the upstairs switch, and then just the switched line from the upstairs switch common to the top light fitting which took its neutral from the upstairs lighting circuit. This avoided any 3C+E but if the upstairs and downstairs circuits are separate then this results in a borrowed neutral, dangerous if one circuit only is isolated because the other can make it live again, and incompatible with RCBOs as both circuits are unbalanced. It is often necessary to correct these cheapskate bodges when installing new consumer units as the borrowed neutral can't be left in place.

And indeed I have seen just this. My parents house. In fact it was a slight variation, as the line for both hall and stairwell (landing) were taken from the back of the kitchen switch. Caught me out as in their installation, twin red had been used from the light down to the switch and back. I changed the switch and managed to get the 'borrowed' feed (a single red) paired with the wrong red in the backbox. So hall and landing only worked if the kitchen light was on! On a positive note, the whole house had just a single lighting circuit (and indeed a single power circuit).

I don't know whether that ever got updated when they had a new CU installed.

[Richardgr]

When I first moved into my house in Sweden it had not been rewired since the 60's. The colours in the fittings were White (L), Red (E) and Black (N). Thanks goodness for standards and harmonisation! And many houses still only have two wire (L + N) for sockets since the buildings are wood and retrofitting is not mandatory.

[Lucien Nunes]

Wiring colours for electrical installations are interesting in the way that different standards evolved without de-facto standardisation of the meanings through association with familiar concepts. In the UK for example, we associated red with danger and made that the standard colour for line (or live, as it was then) and L1 of 3-phase. But as 'obvious' as it might sound in retrospect, it clearly isn't obvious by default. Because while we were using it as L1, it was L2 in Switzerland, L3 in Russia, neutral in Holland and Earth in Germany (amongst others). The fact that it has been chosen for all five conductors of a 3-phase TN-S system, suggests that it doesn't actually convey any meaning by itself. It didn't make it through the selection process for harmonisation, for obvious (or not so obvious) reasons.

Referring back to the title of the thread, I do enjoy learning about different standards and methods used around the world, which has been important in the day job. After 30 years of working on and designing electrics to various countries' standards I've become somewhat cosmopolitan and comfortable with a range of different approaches to the same end result. It's remarkable how parochial electrical knowledge tends to be, because the trade's teaching syllabus is often based directly on the local regulations. Talk to a British spark and the chances are they won't know the American term for a 2-way light switch circuit, let alone the name of the cable used to wire it, its current rating, requirements for protection etc. Mention edge-grounded high-leg open delta and you might as well be speaking Martian.

In a way, that makes it pointless to harmonise wiring colours because if you come from a background of different installation practice, you won't know which kind of wire to use or how you are allowed to install it. Once you've mastered all of that detail for the local standards, learning half a dozen colours is trivial.

[bionicmerlin]

Just a few things I thought I would mention. In the old days I’m talking about 60s 70s that I thought might be of interest.
It was quite common if we had a lighting circuit with no Earth and we required a earth for a metal light fittings we would run a single each wire from a socket to provide one .not the best practice but we would never dream of re wiring a lighting circuit with no earth.
Also in the old days it was considered bad practice to loop in at the switches and you would almost never have a neutral at a switch .
One other thing that was bad practice was on 2 way switches if you had a twin cable going to the switch. Ie a live and switched live returning to the light fitting. We were never allowed to connect it across the straps on the triple cable going between the the two switches. The live was put in a connector together with the red of the triple going to the next switch .With the straps yellow and blue having no connection apart from between each switch . The common of the first switch had the switch return feeding back to the light .
In later years we stopped putting the connector on the common and put the live and live return across the straps. How things have changed.
One point worth remembering . In the old days on 2 way switching 60% of electricians used the red for common and 40% used yellow. Approximately.
Or at least in my area. Even to this day including myself a large amount of electricians use brown for common .The others use black for common.

No disrespect to anyone but I refuse to call a Earth wire a CPC and Live Line .
Andy

[Nickthedentist]

Our new (2011) house has neutrals in each light switch, something I'd not encountered before and didn't realise was "a thing". A nice touch for the reasons Lucien mentions in post 36, but the actual implementation is messy: each switch is mounted on a 25mm-deep backbox which invariably has choc block connectors squished into its depths, but so that there's enough room for the guts of the dimmer switches, the front edge of the backbox is sunk about 3/4" below the face of the plaster. There are also a hell of a lot of wires in quite a small space!

This photo was taken after I'd hacked out all the plaster that had run into the box!

[bionicmerlin]

Nick
That’s probably the normal these days . If it was me I would use smaller earth sleeving. Also in your picture any lives if not brown should have brown sleeving on them also any neutrals not blue should have blue sleeving on them .
So we don’t go off topic . If you compare that to the old days there would be no looping in of wires on a domestic property at the switch If we couldn’t get all the wires in a ceiling rose then we would use joint boxes under floors. Infact some electricians would use joint boxes anyway so as to only have a twin cable and possibly a earth at the light fitting . We would sometimes have a large joint box under a landing floor and several rooms would have the centre light and switch wires terminate there . In the old days wall lights were common and they would be fed from a similar joint box . Almost never looped between fittings . Andy

[Nickthedentist]

Originally Posted by bionicmerlin

Nick
That’s probably the normal these days . If it was me I would use smaller earth sleeving. Also in your picture any lives if not brown should have brown sleeving on them also any neutrals not blue should have blue sleeving on them .
So we don’t go off topic . If you compare that to the old days there would be no looping in of wires on a domestic property at the switch If we couldn’t get all the wires in a ceiling rose then we would use joint boxes under floors. Infact some electricians would use joint boxes anyway so as to only have a twin cable and possibly a earth at the light fitting . We would sometimes have a large joint box under a landing floor and several rooms would have the centre light and switch wires terminate there . In the old days wall lights were common and they would be fed from a similar joint box . Almost never looped between fittings . Andy

Thanks Andy, that's interesting.

Agreed about the narrower earth sleeving. But it can be surprisingly hard to get hold of. I ordered a 100m hank of it from CPC a few years ago and they sent me the wide stuff which would confortably accept several conductors from 2.5mm T+E. When I queried it, they said the narrow gauge was NLA because it was no longer manufactured, but they hadn't updated their catalogue or website. Having said that, I spotted some in one of the DIY sheds last year, and grabbed it.

Hi,

I am sorry it seems someone has put the cover in the wrong place, We don't have any more stock of 2mm and unable to locate a supplier at the moment. We can issue a refund for you.

Regards,

[duncanlowe]

Our previous house, built 1993 had Neutral behind the switches. First time I'd seen it.

This house built 1999 doesn't have that. I wish it did as I'd be able to use one of an array of 'smart' switches or relays that need N to get power. As it is the lack of Neutral at the switch means the smart dimmer needs to 'steal' power from bewteen its input and output.

[Roger Ramjet]

Originally Posted by duncanlowe

Our previous house, built 1993 had Neutral behind the switches. First time I'd seen it.

This house built 1999 doesn't have that. I wish it did as I'd be able to use one of an array of 'smart' switches or relays that need N to get power. As it is the lack of Neutral at the switch means the smart dimmer needs to 'steal' power from bewteen its input and output.

I was given some old security random timer's that replaced a standard light switch which also stole power from between the feed & return. They only worked with incandescent bulbs although you could get an accessory to convert for CFC lamps. Never used them & never would because I suspect they don't offer true isolation.

Rog

[Nickthedentist]

I had some of those, branded Smiths Timeguard, about 25 years ago. They worked very well for switching on domestic lights automatically at dusk for a preset period, though of course the long "on" periods meant that the lamps wore out quickly and a lot of electricity was used. They were a simple 2-wire connection, replacing a standard lightswitch.

Still got them in the garage if anyone wants to play with one...

[G6Tanuki]

Originally Posted by Roger Ramjet

I was given some old security random timer's that replaced a standard light switch which also stole power from between the feed & return. They only worked with incandescent bulbs although you could get an accessory to convert for CFC lamps. Never used them & never would because I suspect they don't offer true isolation.

Rog

A lot of PIR outdoor security-lights seem to manage with a 2-wire-and-earth supply; the last one I dissected was quite cunning in that the SMPS for the LED was initially triggered by a relay-and-capacitive-dropper but once the relay had closed and applied power to the SMPS there was a separate winding on the SMPS transformer that kept the relay 'in' so long as the PIR told it to do so.

[duncanlowe]

Originally Posted by G6Tanuki

A lot of PIR outdoor security-lights seem to manage with a 2-wire-and-earth supply; the last one I dissected was quite cunning in that the SMPS for the LED was initially triggered by a relay-and-capacitive-dropper but once the relay had closed and applied power to the SMPS there was a separate winding on the SMPS transformer that kept the relay 'in' so long as the PIR told it to do so.

Yes, but in that case the two wires are Line and Neutral. In a light switch, you have Line in and Line out. When the switch is on, they are at the same potential (bar any volt drop across the contacts).

[Roger Ramjet]

Originally Posted by duncanlowe

Yes, but in that case the two wires are Line and Neutral. In a light switch, you have Line in and Line out. When the switch is on, they are at the same potential (bar any volt drop across the contacts).

I reckon the Smiths Timeguard units must have an internal resistor that connects across the live feed & switched live return. The voltage dropped across is sufficient to power the electronics. The guy who gave it to me said he removed it because it always felt warm. I would be nervous about fitting one although doubt it would work on CFC or LED lamps.

Rog

[usradcoll1]

Originally Posted by Nickthedentist

Our new (2011) house has neutrals in each light switch, something I'd not encountered before and didn't realise was "a thing". A nice touch for the reasons Lucien mentions in post 36, but the actual implementation is messy: each switch is mounted on a 25mm-deep backbox which invariably has choc block connectors squished into its depths, but so that there's enough room for the guts of the dimmer switches, the front edge of the backbox is sunk about 3/4" below the face of the plaster. There are also a hell of a lot of wires in quite a small space!

This photo was taken after I'd hacked out all the plaster that had run into the box!

That photo shows excessive box fill! On new construction, it would never pass rough-in inspection in the US. There has to be sufficient box depth to accommodate dimmer switches and newer devices, such as GFCI's.
Dave, USradcoll1

[duncanlowe]

Originally Posted by Roger Ramjet

I reckon the Smiths Timeguard units must have an internal resistor that connects across the live feed & switched live return. The voltage dropped across is sufficient to power the electronics. The guy who gave it to me said he removed it because it always felt warm. I would be nervous about fitting one although doubt it would work on CFC or LED lamps.

Rog

Not sure how that would work. If across the contacts, it would have 240v across the resistor when the switch was open. When the switch closes there would be essentially no volts accross the resistor. Thinking the other way, if the resistor were in series, the voltage would depend on the current being drawn by the load. But then when open, the current would be zero so zero volts.

Maybe it's a combination? Load resistor across the contacts to draw current through a series resistor when the contacts are open, relying on the load itself to draw current when the contacts are closed? All sounds a bit risky though and I'd hate to do calculations based on guessed loads.

In the case of the devices I'm referencing, they use a semiconductor to switch. I'd guess a triac though I've never opened one to check. When off, you have full mains voltage between L and switched L. When on, if you don't switch on for 100% of the cycle then you have full mains voltage across the L and SL for the short period you switch off. Sneaky but it works. On the other hand it does mean you can never, quite, get 100% brightness, and you MUST use dimmable LEDs. I don't know what the actual duty cycle on '100% brightness' is and what real effect it has on the lamps output.

[Roger Ramjet]

Hi Duncan,

Forgot to mention that the Timeguard unit actually replaces the light switch.

So I would assume that when Timeguard OFF, the internal resistor in the is in series with the filament of the bulb [load] so in essence we have a potential divider circuit with the volt drop across the Timeguard resistor powering the electronics. As you say when the Timeguard switches to ON this is via a Triac so there will be enough residual non duty power via the internal resistor to keep the control electronics happy.

Only a theory ?

Rog

[Lucien Nunes]

Duncan and Roger's analysis of parasitic power for dimmers & smart switches without a neutral is pretty much it. A capacitive and resistive dropper combo is typical. Many dimmers now use IGBTs insead of triacs, to enable trailing-ege control, which is more compatible with LED loads, plus more flexible power leeching at high overall conduction angles. For low power lighting loads it is often necessary to shunt the load with an RC network - a 0.1μF + 120Ω snubber network is popular - to ensure there is sufficient juice available for the dimmer. Others won't work at all with LEDs; compatibility is unpredictable and pretty much any new permutation of dimmer model and light model has to be verified before installation.

That photo shows excessive box fill! On new construction, it would never pass rough-in inspection in the US. There has to be sufficient box depth to accommodate dimmer switches and newer devices, such as GFCI's.

The UK wiring standard BS7671, unlike the NEC, has no regulations about box fill as such, at least not using the numerical methods of wire gauge, count and box volume that the NEC uses. Wiring must be carried out with good workmanship, suitable components and be easy to maintain, but it is left to the electrician to decide what can fit in which size box. UK light switches have traditionally been very shallow, often occupying as little as 1/4 inch depth of the box as the remaining 1/4" was within the bakelite plate. This means that very shallow boxes were historically the norm, but started to become unsuitable when bulkier dimmers and smart switches arrived. We don't do wirenuts in the UK, at least not for the last 60 years; we used screw terminal blocks and these days better still, slimline lever nuts that lie flat in the back if required, not occupying more than 1/4". This is one bit that a lot of UK sparks don't do nicely, unfortunately. Why anyone would use anything other than a Wago 2273 or equivalent to splice a neutral in a switch box is beyond me.

Although not perfect, because a medium-depth box has been recessed too far leaving the masonry exposed, Nick's box actually has enough room for UK-style accessories with that number of cables, if properly arranged and dressed into place. Slightly more wire length would be preferable, so that they enter at the top, pass down the back of the box, turn forward at the bottom and come up to the correct terminal positions on the switch or dimmer. That way, once uncrewed, the switch will 'swing' down easily with minimal strain on the wires. It also allows spare in case any get damaged or need altering later. It should never be necessary to 'scrunch' them in behind the switch, unless they have been left too short.

A factor that makes UK switches and outlets more forgiving of wire fill is the lack of risk of shorts. There are no side-wire screw terminals on our components, all screws and wire entries are recessed into the insulating body. You could rub your fingers all over the back and internal parts without touching live metal, just as you can with modern US / NEMA-format dimmers and smart switches with captive pigtail wiring. Plus, we fit PVC sleeving over all bare grounds, so there is no chance of a stray bare copper wire touching anything live.

All these differences fascinate me. When I was designing and specifiying materials for electrical installations aboard EU-built but US-based cruise ships, that had to meet UK, EU, US and Lloyds regulations all at the same time (erm, what colour is a hot wire, people?) I had reps from Hubbell, Bryant etc. leaving me stuff for evaluation . I ended up with a big stash of NEMA-style accessories, mostly Hubbell's best, on and under my desk. I loved playing with stuff like that, some of the NEMA outlets for example put ours to shame, very nicely made and durable. Swings and roundabouts, no system has all the best bits.

[Lucien Nunes]

Deleted post requires re-writing.