Faintly-glowing LED lights - any idea what the problem is?

[ABB125]

I have some LED battens (fluorescent tube replacement type things) which glow faintly and continuously when switched off. Does anyone have any ideas as to what the cause might be? I've replaced the switch, and there aren't any parallel runs between different circuits that might induce a current (there's only one place where the lighting and socket circuits cross perpendicularly).

I'm happy to provide more details if needed. Thanks very much.

 

[RichT54]

Are they powered directly from the mains, or from a transformer? If the latter, I wonder if there is a smoothing capacitor in the transformer that could retain sufficient charge to allow them to glow?

 

[ABB125]

Are they powered directly from the mains, or from a transformer? If the latter, I wonder if there is a smoothing capacitor in the transformer that could retain sufficient charge to allow them to glow?

Direct mains supply, in the old-fashioned junction box arrangement (the power supply, switch wire and load wire all connected in a junction box).
I know there are some capacitors in the lights themselves (note to anyone attempting similar work: just because the circuit is switched off doesn't mean you won't get a shock when touching the wires!), but the glowing stops when the whole circuit is switched off.

 

[headshot119]

Direct mains supply, in the old-fashioned junction box arrangement (the power supply, switch wire and load wire all connected in a junction box).
I know there are some capacitors in the lights themselves (note to anyone attempting similar work: just because the circuit is switched off doesn't mean you won't get a shock when touching the wires!), but the glowing stops when the whole circuit is switched off.

I believe this issue comes about when the neutral line is switched rather than the live. My parents had this problem in there kitchen when they first got LED lights put in.

A bit more information on Stackexchange here

 

[ABB125]

I believe this issue comes about when the neutral line is switched rather than the live. My parents had this problem in there kitchen when they first got LED lights put in.

A bit more information on Stackexchange here

I'm pretty sure it's switched live, but I'll check again. Thanks

 

[Cowley]

I recently put an LED light in the van and it did the same thing. Upon reading the instructions unlike a normal bulb it did actually matter which way around the positive and negative were, once swapped around it worked fine.

 

[ABB125]

I recently put an LED light in the van and it did the same thing. Upon reading the instructions unlike a normal bulb it did actually matter which way around the positive and negative were, once swapped around it worked fine.

Interesting.

I've checked, and I've definitely got switched live (unless I somehow managed to get the wires the wrong way round in the (not-)fuse box I'll check tomorrow, but I really can't imagine myself doing something like that! I'll also check how I've connected the individual fittings, but again I can't imagine doing anything other than what the instructions said!

 

[dm1]

The switch itself can act as a capacitor and allow a very small current to flow (in the order of 10 uA or so), which since LEDs are so efficient can be enough to emit a faint light. Lots of the better LED lamps have a resistor in place to prevent this, but the cheap ones often don't.

 

[John Webb]

Is there a radio or TV transmitter anywhere in your vicinity? Just conceivable that if there is, the house wiring is acting as an aerial and inducing sufficient voltage to operate the LEDs in the way you describe.

EDIT: Sorry - just read post #3 again - please ignore my suggestion!

 

[mikeg]

Capacitive coupling? Similarly it can result in CFL bulbs giving the occasional flash as it charges and then discharges. Usually caused by the switched wire being run exactly parallel to a permanent one. Can also occur for the reasons in post 8.

 

[ABB125]

The switch itself can act as a capacitor and allow a very small current to flow (in the order of 10 uA or so), which since LEDs are so efficient can be enough to emit a faint light. Lots of the better LED lamps have a resistor in place to prevent this, but the cheap ones often don't.

These lights were fairly expensive, so I'd hope they have a resistor, but perhaps not!

Is there a radio or TV transmitter anywhere in your vicinity? Just conceivable that if there is, the house wiring is acting as an aerial and inducing sufficient voltage to operate the LEDs in the way you describe.

EDIT: Sorry - just read post #3 again - please ignore my suggestion!

The lights are in an outbuilding with it's own (not-)fuse box. There is a repurposed router in the building acting as a WiFi source, though I'm not convinced it's powerful enough to cause glowing lights!

Capacitive coupling? Similarly it can result in CFL bulbs giving the occasional flash as it charges and then discharges. Usually caused by the switched wire being run exactly parallel to a permanent one. Can also occur for the reasons in post 8

The switch wire is very short (maybe 15cm long) and separate from any other wires. I might try rearranging the wires in the junction box to reduce any crossover between different wires, to see if that makes any difference.

 

[RichT54]

Have you checked with a volt meter whether there is any voltage showing between the neutral and earth?

 

[ABB125]

Have you checked with a volt meter whether there is any voltage showing between the neutral and earth?

Unfortunately I don't have a voltmeter!

 

[Domh245]

Unfortunately I don't have a voltmeter!

Not even a multimeter? I'd strongly suggest getting one if you can, they're not expensive and come in vey handy when doing electrical troubleshooting!

 

[RichT54]

Unfortunately I don't have a voltmeter!

I remember reading somewhere (sorry can't find the link) of a similar case where there was a problem with the neutral. Not a break in the circuit, but a poor connection somewhere that introduced enough resistance to allow the neutral to 'float' which caused it to have a sufficent voltage difference to earth to light some LEDs. If that is the case, finding the point where the poor connection is located would be rather tricky without a meter.

 

[ABB125]

Not even a multimeter? I'd strongly suggest getting one if you can, they're not expensive and come in vey handy when doing electrical troubleshooting!

I've never been able to work out how to use a multimeter; I probably should learn though!

I remember reading somewhere (sorry can't find the link) of a similar case where there was a problem with the neutral. Not a break in the circuit, but a poor connection somewhere that introduced enough resistance to allow the neutral to 'float' which caused it to have a sufficent voltage difference to earth to light some LEDs. If that is the case, finding the point where the poor connection is located would be rather tricky without a meter.

Ok, thanks. I'll have another look at the connections later.