Mains LED lights glow in the dark

[Barge Maria]

I wonder if one of you electrickery experts can give me a steer please.

 

230 volt led bulbs in metal kitchen light fitting

 

Work fine, all other use elctrics work OK as well

 

When switched off they still give a faint glow whilst in fitting

 

When unscrewed there is no glow

 

I can't seem to detect any voltage from the turned-off fitting when using a multimeter although access is tricky (deep plastic collar on fitting)

 

Any ideas anyone?

 

Many thanks - apologies if this email seems curt but typed on small tablet in haste while on car ferry!

 

Tim

[by'eck]

Wild guess but light switch would break live only so they may be receiving low voltage between neutral and earth. Raw LED's only need 2-3 volts to strike.

[Nickhlx]

Or a high-impedance, capacitive pickup from adjacent cables....

 

Nick

[nicknorman]

Are you sure it isn't just afterglow? Warm white LEDs tend to use a fluorescent material to convert the blue or UV LED light to lower frequency warm light, and when the power is removed some types do continue to glow faintly for quite a while. Depends on whether they stop glowing abruptly when you disconnect them, or whether it is just that by the time you have disconnected them they have stopped glowing anyway.

[Justme]

As mentioned above this can be due to long adjacent cable runs inducting a small current.

 

Another possibility is that your install is not correctly N & E bonded & you have a small potential across them.

 

Not enough to light a normal bulb but will light an LED.

 

Does it happen on your inverter, genny, land line or a combination of all 3?

 

To rule out the induction turn off all the MCB's except for the light you want to test.

 

If it goes out its induction from the other cables, if it does not then its the potential.

[Radiomariner]

Are you sure it isn't just afterglow? Warm white LEDs tend to use a fluorescent material to convert the blue or UV LED light to lower frequency warm light, and when the power is removed some types do continue to glow faintly for quite a while. Depends on whether they stop glowing abruptly when you disconnect them, or whether it is just that by the time you have disconnected them they have stopped glowing anyway.

This is a most probable answer. In my kitchen, the white LED's glow for about 2 minutes after switching off. The bedroom reading lights (warm white LED's) glow for about 30 minutes after being switched off.

Both are from the same type and manufacturer with 60 surface mounted LED's per lamp.

Gave me quite a scare when after fitting these I walked into the darkened bedroom to find an eerie glow at eye height a few feet away.

 

Edited to add: - If they stay glowing all night. try removing the lamps and rotating them 180 degrees in their sockets. I don't think it'll make a difference but it might be that the electronics in the lamp allow one connector to be sensitive to the power on the neutral line.

Edited March 2, 2015 by Radiomariner

[John C]

_{Had exactly the same happen with the LED's in my kitchen at home, the faint afterglow would diminsh after about half an hour. An electrical check showed no problems. These LED's where bought off Ebay really cheap, since then I have bought more expensive (from Screwfix) that do not "afterglow". I know of a couple of other installations that did the same, better quality LED's stopped it. So it could just be cheap LED's. Do the electrically learned here think this to be so?}

Edited March 2, 2015 by John C

[by'eck]

This is a most probable answer. In my kitchen, the white LED's glow for about 2 minutes after switching off. The bedroom reading lights (warm white LED's) glow for about 30 minutes after being switched off.

Both are from the same type and manufacturer with 60 surface mounted LED's per lamp.

Gave me quite a scare when after fitting these I walked into the darkened bedroom to find an eerie glow at eye height a few feet away.

 

Edited to add: - If they stay glowing all night. try removing the lamps and rotating them 180 degrees in their sockets. I don't think it'll make a difference but it might be that the electronics in the lamp allow one connector to be sensitive to the power on the neutral line.

 

That may be an answer since direct mains powered LED's can simply have a capacitor, resistor and diode to provide the low voltage DC, with afterglow caused by residual charge in capacitor. It seemed the OP was confident that the glow dissapeared as soon as they were removed from socket though - maybe time for some verification.

 

ETA: FWIW we have BC and ES LED bulbs throughout the house with no such issues. I notice that although the 6:1 reduction in power for same light output compared with incandescent may seem attractive, when I first started using LED replacement bulbs on 12 volt system, the power saving for same light output was closer to 30:1.

Edited March 2, 2015 by by'eck

[Radiomariner]

This is a most probable answer. In my kitchen, the white LED's glow for about 2 minutes after switching off. The bedroom reading lights (warm white LED's) glow for about 30 minutes after being switched off.

Both are from the same type and manufacturer with 60 surface mounted LED's per lamp.

Gave me quite a scare when after fitting these I walked into the darkened bedroom to find an eerie glow at eye height a few feet away.

 

Edited to add: - If they stay glowing all night. try removing the lamps and rotating them 180 degrees in their sockets. I don't think it'll make a difference but it might be that the electronics in the lamp allow one connector to be sensitive to the power on the neutral line.

Just tried turning a lamp 180 deg's. No change. With hind sight I should have known: With a total 14 similar lamps installed you'd think at least one would be the right way around!

[Nickhlx]

We have a CFL that is on an electronic timer switch which is a simple two wire (in series with the bulbs ) device - the "leakage" through the switch was enough to make that glow when "off", as the standby few mA of current was enough to cause it to glow faintly...

 

Nick

[Barge Maria]

Thanks for all this folks.

 

Just to confirm:

 

No adjacent cables

Not cheap LEDs

Glow stays all night

 

I shall investigate N/E bonding upon my return to the boat.

 

Tim

[Keeping Up]

We have a CFL that is on an electronic timer switch which is a simple two wire (in series with the bulbs ) device - the "leakage" through the switch was enough to make that glow when "off", as the standby few mA of current was enough to cause it to glow faintly...

 

Nick

 

You can buy a capacitor to put across the CFL, which stops that glow, and indeed is necessary for some timers to operate correctly.

[Barge Maria]

Quick QN.... To check the N/E bonding, is it as simple as turning power off and checking resistance between the conductors? I have a Victron isolation trannie in the system but have no reason to suspect this.

 

No doubt someone will guide me!

 

Cheers

 

Tim.

[by'eck]

Quick QN.... To check the N/E bonding, is it as simple as turning power off and checking resistance between the conductors? I have a Victron isolation trannie in the system but have no reason to suspect this.

 

No doubt someone will guide me!

 

Cheers

 

Tim.

 

If you have an isolation transformer between shore power inlet and boat distribution system, it must have its nominal neutral output bonded to earth and boat hull . See here - lack of this which is a safety breach, may explain your issue.

 

Note that the incoming earth is connected to transformer core, since its an appliance and should be earthed, but is otherwise isolated from boat.

Edited March 2, 2015 by by'eck

[Barge Maria]

Thanks. As far as I know the Victron has the correct N/E bond internally and it is clearly earthed to the hull... Gurt big earth wire!

 

These replies are much appreciate iated btw.

[by'eck]

Thanks. As far as I know the Victron has the correct N/E bond internally and it is clearly earthed to the hull... Gurt big earth wire!

 

These replies are much appreciate iated btw.

 

If you have an isolation transformer (IT), its entirely separate from the Victron. As you suggest the N/E bond on the latter is usually automatic but only applies when its inverter is running - generally when shore power is not present.

 

You need to check the IT wiring with shore power disconnected of course, since with shore power present the AC power simply passes through the Victron.

[Barge Maria]

No its a Victron IT. I didn't mention inverters at all.

[by'eck]

No its a Victron IT. I didn't mention inverters at all.

 

Apologies for incorrect assumption - its just that when Victron's are mentioned on this forum its usually one of their Combi's (combined charger/inverter) being referred to. Still worth checking in the IT manual though for that all important N/E connection.

[Barge Maria]

No worries! I knew what the assumption was and why.

 

The manual/PDF shows a jumper for the N/E link in the IT. I shall need to..er...refresh my memory when I get back on Wednesday.

[forsberg]

Leakage in the switch seems most plausible, carbon build up or humidity.

A non electronic analog volt meter might not be able to detect voltage if very low current.

[by'eck]

Leakage in the switch seems most plausible, carbon build up or humidity.

A non electronic analog volt meter might not be able to detect voltage if very low current.

 

Happens when using mains switches to control 12 volt LED's since you don't get the self cleaning action on the contacts, but not mains ones I would have thought.