Led lights glowing slightly when switch is off

sparks3000

Hi there as stated led lights 230volt installed in a hall are glowing when switch is off .
There is a live from board to light switch and neutral from board to lights , so no looping .
I have checked the 1g1w switch for dampness which could create a path and that is not the problem. I have checked the earth at board and carried out a loop impedance test and meggered the circuit. All these have shown no issue .

I then tried running a separate neutral and lights still glow. Using a clip on ammeter I have a tiny current at the lights under all these scenarios which leads me to think that a emf is being induced in the feed to these lights. How do I rectify this problem I'm thinking wiring a traditional incadescent in parallel would solve this or is there some sort of electronic circuit available . Thanks in advance for any solutions

brightspark

When you say you meggered the circuit, did you megger between the switch wire and any other live cables? (Obviously done with the LED removed)

What voltage do you have across the LED when it is switched off?

Does the light glow continuously or does it eventually fade completely?

2011

I bet they stop glowing when the neutral is disconnected. Right?

sparks3000

brightspark wrote: »

When you say you meggered the circuit, did you megger between the switch wire and any other live cables? (Obviously done with the LED removed)

What voltage do you have across the LED when it is switched off?

Does the light glow continuously or does it eventually fade completely?

I meggered between switch wire and all lives. All lamps and LEDs removed
Voltage was tiny the current was.001 amps or 1mA. The light glows continuously with the switch off

Bruthal

2011 wrote: »

I bet they stop glowing when the neutral is disconnected. Right?

That would be expected in every scenario that they light. They will probably stop glowing when the sw is disconnected too.

A capacitance/inductive link is probably the cause. Wrong polarity is another possible cause although unlikely to be the case here.

sparks3000

2011 wrote: »

I bet they stop glowing when the neutral is disconnected. Right?

Correct

Bruthal

sparks3000 wrote: »

I meggered between switch wire and all lives. All lamps and LEDs removed
Voltage was tiny the current was.001 amps or 1mA. The light glows continuously with the switch off

Did you try disconnect the sw from the lights Themselves?

sparks3000

Bruthal wrote: »

That would be expected in every scenario that they light. They will probably stop glowing when the sw is disconnected too.

A capacitance/inductive link is probably the cause. Wrong polarity is another possible cause although unlikely to be the case here.

Polarity is correct . I think it is a problem with induced voltage . What do you think would correct this

sparks3000

Bruthal wrote: »

Did you try disconnect the sw from the lights Themselves?

I did . They still glowed

Bruthal

sparks3000 wrote: »

Polarity is correct . I think it is a problem with induced voltage . What do you think would correct this

Running twin and earth from source to switch, then to lights might do it. Or single core run in separateseparate /away from other circuits.

Adding the incandescent in parallel will work, but thats sort of defeating the purpose.

Try disconnect the switch wire from the switch completely, and see does that have any effect. If they still glow, induction/capacitance is very likely.

Try switching off every MCB, starting with the affected one first off and back on, and see can another circuit be identified as the cause of induction/capacitance, while leaving on the circuit for the affected lights.

Bruthal

sparks3000 wrote: »

I did . They still glowed

You disconnected the sw directly from the lights themselves?

How many are there?

Bruthal

Put a small relay into the setup. Light switch that on and off. Relay switch L + N.

Or keep them as they are as a likely free night light.

brightspark

It's a slight cheat but apparently a small neon will solve it too?

http://www.davidsavery.com/index.php/free-advice/117-why-do-my-led-bulbs-glow-when-off

Hoagy

Or one of these from Maplins. ( read the review)

Bruthal

Hoagy wrote: »

Or one of these from Maplins. ( read the review)

They will work as well. Installed in a metal box might be a good option and flex to any light on the switched group.

2011

Bruthal wrote: »

That would be expected in every scenario that they light. They will probably stop glowing when the sw is disconnected too.

No, the lamp is glowing when the switch wire is disconnected. The issue is noise on the neutral. The solution is to disconnect the neutral when the lamp is off.

sparks3000

Got this sorted the switch live was ran to a junction box hidden under insulation and I found two feeds for the spotlights here.
I also found 4 neutrals joined here . So originally these led lights were two separate circuits . Two of these neutrals were for each circuit . The third neutral went directly back to board . The fourth was brought to a junction box outside for what was once an outside light and from here another neutral was brought back to the board . This was the loop which allowed the LEDs to light when switched off . It's easier to draw than describe
Many thanks guys for your help

Bruthal

2011 wrote: »

No, the lamp is glowing when the switch wire is disconnected. The issue is noise on the neutral. The solution is to disconnect the neutral when the lamp is off.

So induced voltage is eliminated as a cause by that you believe?

sparks3000

Bruthal wrote: »

So induced voltage is eliminated as a cause by that you believe?

Yes but thanks for the info it will be useful in the future . I returned it to a two gang switch so two separate circuits again. As they say every day is a school day. Looking back on the day the megger results for the neutral was an indication because it was very slow to reach infinity but looking at it now there was a lot of water in the junction box outside so possibly acted as a weak conductor but enough to create a circuit for 1 milli amp

Bruthal

The cause is almost always induced or capacitive voltage in the floating switch wire. Its more noticable in 2 way switching.

Disconnecting neutral from lights will cause them to go out in most circumstances. An intact neutral from light to neutral bar is unlikely to have anything on it causing a glow, when the other side of light is connected to nothing except a floating length of wire. But the floating wire can have induced voltages from nearby live wires causing small current flow through the high impedance LEDs, and onto the intact neutral.

Either way, disconnecting the neutral I would expect to turn off any light, be it fully switched on, or on from induction, or capacitance. So disconnecting the neutral doesn't certainly point to 1 single cause.

2011

Bruthal wrote: »

So induced voltage is eliminated as a cause by that you believe?

I think it is unlikely, yes.
I know you are thinking of capacitive coupling. This would be enough to register on a meter as you discussed before. However I do not think that there would be enough power to partially light a lamp.

The reason that I say this is I had exactly the same issue with LED lamps not so long ago. I remember discussing this with you over a coffee when we worked on the same job several months ago. LED lamps were glowing quite brightly on DOL starters when the switch wire was disconnected. The problem was on the neutral. Once we switched the neutral with the switch wire the problem went.

Bruthal

Like I said though, disconnecting the neutral will eliminate just about all reasons for lights emmittin light. It doesn't confirm just 1.

I had already suggested s double pole relay to eliminate the problem. As would dp switch. It doesn't confirm that its not capacitive link, or is neutral noise. Any thing the neutral is doing has to be relative to what the connection on the other side of the lamp is doing.

If it is capacitive coupling, ground the floating sw will likely kill the glow. Even touching the floating sw is likely to do it.

2011

The current induced in a cable due to capacitive coupling will depend on many factors but mainly:

1) The length of run of cable that is in close proximity to other cables that are carrying a current. In a house the lengths of cable runs are unlikely to be very long (<100m).

2) The combined magnitude of the current in the "other cables". Again the currents are not likely to be that large (domestic installation).

3) The properties of the cable insulation. This insulation would have been selected on the basis of the application.

If the OP wants to know definitively if it is capacitive coupling or not it is easy to find out; simply turn off every MCB in the board except for the effected lighting circuit. If the glow was caused by capacitive coupling the problem would go away when this is done because there is no other circuit present to "couple" with.

Bruthal

2011 wrote: »

The current induced in a cable due to capacitive coupling will depend on many factors but mainly:

1) The length of run of cable that is in close proximity to other cables that are carrying a current. In a house the lengths of cable runs are unlikely to be very long (<100m).

No current need be flowing in the live cable for capacitive coupling except for the capacitive coupled current itself. It often happens in 2 way switching circuits, with strappers following identical paths. No current is flowing in either strapper with the light off, except the actual capacitive current.

2) The combined magnitude of the current in the "other cables". Again the currents are not likely to be that large (domestic installation).

How much current is going to flow from the "noisy" neutral, through the lamp, and into a floating wire? We wouldnt expect much it it was a live to the lamp, and a floating wire on the other terminal.

3) The properties of the cable insulation. This insulation would have been selected on the basis of the application.

The insulation is not going to have much bearing on capacitive coupling except for the distance it maintains between conductors. It doesnt prevent it, so its not overly relevant.

If the OP wants to know definitively if it is capacitive coupling or not it is easy to find out; simply turn off every MCB in the board except for the effected lighting circuit. If the glow was caused by capacitive coupling the problem would go away when this is done because there is no other circuit present to "couple" with.

It would have been better to first switch off the affected circuit, then switch it back on if still glowing. Then off with each other one, leaving them off.

But presumably if it is some neutral problem, they would stay glowing with everything off.

The real test to check for capacitive coupling, or induction, is to touch the floating switch wire, or earth it into an independent earth rod. It only needs to be something like a tent peg etc into the grass. That would usually be enough to eliminate any coupling voltage.

Anyway, you missed my point overall. Id be surprised if capacitive coupling was enough to get them to glow. I tested it before on numerous lamps, and never got it to happen. It does happen with neon indicators easily. But id be more surprised if anything on an intact neutral into a lamp, and other side is a floating wire caused it either.

Its not capacitive coupling v neutral noise that im posting anyway. It is simply that removing the neutral is highly likely to kill all emitted light from any lamp, no matter what the cause of glowing. As in, its not proof of any cause. If its neutral noise, disconnecting the neutral will kill it. If its capacitive coupling, disconnecting the neutral will also kill the glow.So the neutral disconnection is not definitive imo.

Bruthal

Interestingly enough, with nothing better to do, I just connected a single led directly to the landing light fitting with the light turned off, and the led is lighting. What will happen if I remove the neutral?