Switch wire not labelled.

[JohnW2]

I must admit my first thought was BANG until John pointed out the bit about disconnection the switch which I too had missed mentally.

Yes, we both initially missed that. Indeed, I confess that I was half way through writing a 'scathing' reply(about BANGs!) before, thankfully, I noticed the bit I had missed!

I think Flameport`s comment is the best and I reckon he would not be too keen on someone attempting to use a multimeter instead of approved testers in approved ways yet he is mindful of perhaps the "least worst" if not the proper way to do, being pragmatic.

Agreed, I was pleased (and a little surprised) to see what he suggested, since 'pragmatism' is not something he very conspicuously practices, his preference seemingly being for 'asserted idealism'.

To the OP. ... as a thought exercise, its merely a perm any 2 from three connections and seeing which works and which two do not work.
If it was done with a simple battery and bulb (lamp - filament lamp) I would be happier than someone using the mains in this way.

Agreed, but it is crucial that everyone reading "perm any 2 from three and see which two do not work" understands that that approach can ONLY be taken if the power is off and one is using something (like a battery and bulb) to test continuity. If one tried to ascertain which of the permutations 'worked' (or didn't work') when the power w as turned on, some of options would result in a BANG!

 

[JohnW2]

Many meters these days, include a buzzer continuity tester, so all that is needed is a bit of wire to extended on of the leads.

The leads wouldn't even need to be extended if (with power off!!) someone operated the switch whilst one applied the leads of the meter to the pair of conductors in each cable, in turn

 

[SUNRAY]

Yep, very true, after all it's only a few wires, how complicated can it be - so they wade in, without a thought....

I prefer the simple, no contact needed, volt stick method, with circuit powered up. First, find the L. With switch open, connect L, in turn to each red, until red at switch shows to be live - the black of that T&E is your switch wire at the ceiling, so mark it. Rest of the reds, must be looped reds, to other lights.

Not being so risk averse, I would do all that, with it live, but using extreme care.

And 'tis the most likely method I'd use, however historically when I even hinted at anything similar I've been shot down in flames, to the extent I asked mods to remove the whole section of the threads... the thought of "working live" or using anything other than a 2 prong tester... heaven forbid.
In more recent times we get the odd "Junk" comment from one member or chuck it in the bin but they are certainly a little more acceptable now.

 

[Harry Bloomfield]

The leads wouldn't even need to be extended if (with power off!!) someone operated the switch whilst one applied the leads of the meter to the pair of conductors in each cable, in turn

True! It fact, with the switch turned on, power off, the red and black for the switch, could be determined by resistance measurement, the lowest value across the pair = the switch, then confirmed by rechecking with the switch turned off.

 

[JohnW2]

True! It fact, with the switch turned on, power off, the red and black for the switch, could be determined by resistance measurement, the lowest value across the pair = the switch, then confirmed by rechecking with the switch turned off.

Yep- even simpler!

 

[SUNRAY]

In fact using my preferred version, which I find more useful, is the LED screwdriver as that also includes a continuity function.

 

[ebee]

Yes, we both initially missed that. Indeed, I confess that I was half way through writing a 'scathing' reply(about BANGs!) before, thankfully, I noticed the bit I had missed!

Agreed, I was pleased (and a little surprised) to see what he suggested, since 'pragmatism' is not something he very conspicuously practices, his preference seemingly being for 'asserted idealism'.

Agreed, but it is crucial that everyone reading "perm any 2 from three and see which two do not work" understands that that approach can ONLY be taken if the power is off and one is using something (like a battery and bulb) to test continuity. If one tried to ascertain which of the permutations 'worked' (or didn't work') when the power w as turned on, some of options would result in a BANG!

Sorry I should have qualified that comment a bit.

When considering 3 x T & E cable to a ceiling rose in its simplest form (many variants are possible) then you have three at the rose, mains in (from the supply or the last rose before the one being tested) , the the second one is mains out to the other lighting roses further down the line, the third one is the switch wire (feed and return).
With this simplest of setups it could be possible to join two twin and earths (A & B) , test the lights further down the circuit to see if they work as should be.
If that happens then C the remaining one is the switchwire.
If not then use A & C to join and switch the others again and if so the B is the switchwire.
If not then it should be B & C to join together meaning that A is the switchwire so do the switches test gain.

This all depends upon there being other switches on roses further down that circuit so you can test switch.
Most important though it depends upon you not connecting mains up to a fault or short, that could result in quite a bang or great danger and it relies upon being successful picking the correct switch to remove the switchwire from in the first place.
Any errors and it could cause great calamity as well as confusion.
So be safe - do not use the "240 Bang Bang Test" in the first place, a multimeter or a continuity tester (bought or home made), is safer but never really totally reliable for safe use.

A "Volt Stick" should NEVER be used as a Test for Dead but might be considered as a Test for Live/
Using it solely as a "back up quick double check" after using proper equipment and methods is not such a bad thing but can never be relied on purely by itself - double ditto for a "Mains Tester Screwdriver" although they seemed to be all the rage in the 60s and early 70s.

 

[pete01]

If not then use A & C to join and switch the others again and if so the B is the switchwire.
If not then it should be B & C to join together meaning that A is the switchwire so do the switches test gain

Not forgetting to disconnect one of the switch connections , just incase

 

[Harry Bloomfield]

Using it solely as a "back up quick double check" after using proper equipment and methods is not such a bad thing but can never be relied on purely by itself - double ditto for a "Mains Tester Screwdriver" although they seemed to be all the rage in the 60s and early 70s.

They are always my first 'go to', ever since they appeared on the market. The beauty is, they work without any contact, or exposing anything live. The trick, as always, is knowing what they can and cannot do, knowing what to expect.

 

[ebee]

The trick, as always, is knowing what they can and cannot do, knowing what to expect.

Yes Harry, that is why they should NEVER be relied upon as a "Test for Dead" which is the most important part of all.
That is why I am not against them once a proper test for dead regime is established first then use the volt stick afterwards as a double check if you like.
Never use them as the only test at all.

 

[ebee]

Not forgetting to disconnect one of the switch connections , just incase

Exactly, that is most important that you power down completely then remove this connection and make safe before you power up again (I would put both the conductors (feed and return a.k.a L and SWL ) into two different choc blocks/wagos etc rather than just disconnecting then a choc block on just the one.

 

[securespark]

I was explaining it to my wife but she said "stop" I have no idea what you are talking about

I quote regs to Mrs S in the hope of getting her in the mood, but it doesn't work....

Fair enough so I shared it with you lot

Now I realise so do I, but not with the same intention!

 

[morqthana]

If one tried to ascertain which of the permutations 'worked' (or didn't work') when the power w as turned on, some of options would result in a BANG!

Which would tell you things.

 

[ebee]

Which would tell you things.

Well yes it certainly would help narrow things down, I have known quite a few "ordinary persons" take a bit advice from someone who knows a bit better and end up with a tell tale bang (Actually more of a click or a thwack these days but the BS3036 rewireable fuses could be quite good for BANGS ).

Some folk even embrace the good old "Bang Bang 240" test as an intuitive method of fault finding

 

[Harry Bloomfield]

And 'tis the most likely method I'd use, however historically when I even hinted at anything similar I've been shot down in flames, to the extent I asked mods to remove the whole section of the threads... the thought of "working live" or using anything other than a 2 prong tester... heaven forbid.
In more recent times we get the odd "Junk" comment from one member or chuck it in the bin but they are certainly a little more acceptable now.

As said, a volt stick, is certainly my go to/my first line diagnostic tool. If you know, and understand their limitations, they are invaluable - you can certainly rely upon one to test for live, what you cannot safely do, is rely on one to prove dead.

I regularly use one for finding breaks in flexible cables, and to confirm there is both a live and neutral available in a flex, plus some sort of earth, in a three core - you can follow the individual wires, in the flex, from the outside.