What might have happened to cause this fault?

[cjard]

I'm wiring up a light in a bedroom and I was hoping to make it 2 way but after I assembled the circuit I tested it and found avoltage on L1 and L2, of 230v and 100v, flicking the switch flipped the voltages round.

The wiring system was organised in the following way:

This is an attic bedroom, and a spur was run from an existing lighting ring so I have a single entry point presenting 240v at the switch. I joined this live to common and passed the neutral straight into the 4 core.

Half way along the 4 core, I've split the insulation open and cut the two possibly-live cores, putting them onto L1 and L2 of the slave switch, picking on one exiting live core, to go onto the light:

[code:1]
L1-----L1
LIVE ---COM/ \COM------
L2-----L2 LIGHT WILL GO HERE
NEUT --------------------------
[/code:1]

The light isnt in yet so I was expecting to see a nice and simple outcome to my tests, that L1 / L2 on the master would have 0/230 depending on the switch position.

Seeing that about 100v is present on the core that should be 0, and also having observed that if I put the voltmeter between the live supply and e.g. a nail driven into the wall, I also get 100v I'm suspecting that at some point on the cable, both cores leak into e.g. the plaster of the wall, but not each other.. If a screw had been driven through the wires, bridging them and earthing them I would have expected the bridge to cause 230v to be presented on each regardless of the switch position, but given that they both present 100v, i'm thinking that somehow both cores have been grounded and the leak from one core to the other is 100v..

What is really confusing me though.. I've currently got the circuit like this:



[code:1]
L1-----SL1 LL1------
LIVE ---COM/
L2-----SL2 LL2------
NEUT ------------------------------

No light in circuit - should just be a wire in the ceiling terminating in (no connection) as in diagram
SL = to-switch live core
LL = to-light live core
[/code:1]

The master switch is switched so that L1 is getting 230. The slave switch has been removed and all I have are the cut wires sticking out the wall (with the neutral uncut). Putting the multimiter between the following gaps I get the reported voltages

SL1:SL2 = 120v (up from before - it's raining outside??)
SL1:LL1 = 230v
SL1:LL2 = 230v
SL1:NEUT = 230v
SL2:LL1 = 85v
SL2:LL2 = 90v
SL2:NEUT = 100v
LL1:LL2 = 0v
LL1:NEUT = 0v
LL2:NEUT = 0v

It's almost like someone wound a screw through all 3 cores up in the ceiling on the light side of the circuit, because every core functions as a neutral to the live 230v.. But the voltages definitely differ when the LL cores or the neutral are used with the leaking 120v voltage from the SL2.

I have done a DC continuity test and there is no continuity between any of the cores, so I don't think a screw has been wound through it.. Multimeter was set to max ohm scale for this test, to catch any connection having a resistance up to 2000 megaohms; all core connections showed as infinite resistance

Can anyone suggest what's gone wrong/where I should look for the fault? I suspect there might be two wiring faults; one on the light side that makes all cores function as neutral, and another on the run between the switches which presents about 100v on a line that should be 0v

Thanks in advance for reading my lengthy and boring post!

cj

 

[Spark123]

Firstly doing the thing with the nail in the wall isn't a good idea.
What you are seeing is capacitive coupling of the cores inside the cable. This can only be measured with a high impedance volt meter such as a digital multimeter. If you have a normal lamp inserted and have them configured so the lamp is off, the voltage should fall to zero. Might be a slight issue if you use a compact fluorescent lamp as it may flash.
I haven't a clue what your LL1 and LL2 are about!

 

[cjard]

So there might actually be no fault at all? In other words, if i just connected a length of 3 core up like this:

[code:1]
LIVE-----nothing
nothing-----nothing
NEUT-----nothing
[/code:1]
i.e. the earth core is doing nothing except lying alongside a long length of live core connected to a live supply, then a voltage may be observed between the earth core and the neutral?

Mmhh.. *Recalls physics lessons whereby a live wire generates an em field that when another wire lies within, a current is induced in that other.."

I'll connect it up to a light and see what happens. At the worst, I'll end up with a bulb that is half bright or full bright!

Why is bridging a live to a nail in the wall (it's only an earth in the same way that i'm an earth when I touch the handle of my electric screwdriver) a bad idea? I'm using a digital voltmeter, and I presume the internal resistor is massive enough to prevent me getting a shock.. i've even found it quite handy to use myself as an earth (holding the black prong of the meter and poking around with the red) during this investigation..

Thanks for the info! i'll go finish the wiring!

 

[Spark123]

You should never have an earth core disconnected, but in theory you may see a voltage in a core left floating and run in parallel with another core caused by capacitance.
In order for a voltage to be induced by a magnetic field you need to have a current flowing in the cable to create the magnetic field.

 

[cjard]

I haven't a clue what your LL1 and LL2 are about!

You mean, youre not sure why they function as perfect neutral/earth and a full 230v develops between them and the SL1 when I connect them up...

or was it just the LL terminology that was confusing? In the second diagram I'm merely saying that I have 3 participating cores coming out of the master switch.. one neutral and an L1 / L2 occasionally-live pair.. because the wire that goes to the slave and then onto the light has these 3 participating cores, it sticks out of the wall in a loop.. i've cut the two cores that run to L1 and L2 so that i can wire another 2 way slave switch in... i will, of course, then only use one of the cores on the leg out to the light


Here are some colours:

[code:1]
With switch in:
L1--red---L1
LIVE --brn--COM/ \COM------red------------+
L2--yel---L2 X--yel--X LIGHT
NEUT --blu----------blu---------------blu------------+

Without the switch (testing purposes, just cut wires hanging our the wall):

L1--red---SL1 LL1--red-------------
LIVE --brn--COM/
L2--yel---SL2 LL2--yel-------------
NEUT --blu----------blu---------------blu-------------
[/code:1]

As can be seen, right now I have a t the slave switch two red and two yellow hanging out the wall.. they used to be joined but I cut them.. I will eventually clip off and dispose of the yellow core to the light because it will not be used. SL2 (yellow) back to the main switch is used because sometimes it will be live.. So SL means "switch-connected live" or "a live core that goes back to the main switch" and LL means "light-connected live" or "a live core that goes on to be connected to the light"

If there is a standard terminology for these (other than L1 and L2, because they didnt really help me explain things) do let me know what it is so i dont cause as much confusion in future!

thanks!
cj

 

[cjard]

You should never have an earth core disconnected, but in theory you may see a voltage in a core left floating and run in parallel with another core caused by capacitance.
In order for a voltage to be induced by a magnetic field you need to have a current flowing in the cable to create the magnetic field.

Of course! I remember now that capacitors conduct AC but not DC. Thanks for the reminder about the EM field.

Given that cores should not be disconencted.. when I do away with the yellow core that goes to the light, should I tie it to earth at the light end:

[code:1]
L1--red---L1
LIVE --brn--COM/ \COM------red------------+
L2--yel---L2 X--yel----ERTH LIGHT
NEUT --blu----------blu---------------blu------------+
ERTH ----------------------------------------ERTH
[/code:1]

Thanks again!

 

[Spark123]

It is the terminology of what you are doing which is a little confusing, you shouldn't really have bare live terminals hanging out of the wall even for testing, or do you mean you have them in a chock block / switch etc.
There isn't really a need to run a 3c+e to the lamp from the 2nd switch. Instead of leaving the spare core floating it is better to connect it to earth.

 

[conny]

I'm using a digital voltmeter, and I presume the internal resistor is massive enough to prevent me getting a shock..

i've even found it quite handy to use myself as an earth (holding the black prong of the meter and poking around with the red) during this investigation..

erm,

you are not an electrician are you?

 

[plugwash]

I'm using a digital voltmeter, and I presume the internal resistor is massive enough to prevent me getting a shock..

i've even found it quite handy to use myself as an earth (holding the black prong of the meter and poking around with the red) during this investigation..

erm,

I suspect the internal resistance of most DMMs is indeed enough to prevent a shock, not that it is something I would advise assuming.

 

[Spark123]

I missed that bit, doing that it somewhat worse than using even a neon screwdriver, set the wrong range or get it mixed up with current and you will be in direct contact with 240v.
Definately a no no.
You even have a handy neutral and earth to connect the black probe to!

 

[securespark]

i've even found it quite handy to use myself as an earth (holding the black prong of the meter and poking around with the red) during this investigation..

If you are on a timber 1st floor, that won't work because you are well separated from true earth.

And in any case, if it did, this would be a very bad idea.

This is probably adding to your confused figures.

 

[aptsys]

set the wrong range or get it mixed up with current and you will be in direct contact with 240v.

That's one way of doing an RCD test... and at least you'll be able to see the trip current

 

[Softus]

If there is a standard terminology for these (other than L1 and L2, because they didnt really help me explain things) do let me know what it is so i dont cause as much confusion in future!

Given that cores should not be disconencted.. when I do away with the yellow core that goes to the light, should I tie it to earth at the light end

For a start you could avoid writing things like "tie it to earth", which you appear to think sounds impressive, but is worse than meaningless.

You very first sentence is a model of confusion:

I'm wiring up a light in a bedroom and I was hoping to make it 2 way but after I assembled the circuit I tested it and found avoltage on L1 and L2, of 230v and 100v, flicking the switch flipped the voltages round.

There is hardly one noun or verb that doesn't need explaining:-

"wiring up" ?
"light" ?
"hoping" ?
"assembled" ?
"circuit" ?
"tested" ? (!)
"found" ?
"flicking" ?
"flipped" ?

I'm amazed by the politeness of the professionals on this topic. The work you're doing is clearly illegal, because it's stunningly incompetent and unsafe.

Have you ever heard of BS7670, let alone picked up a copy of it?

 

[VanSolo]

Have you ever heard of BS7670, let alone picked up a copy of it?[/quote]

BS7670?

 

[Softus]

Indeed so - it's the BS number for the testing of idiots.