I have a set of 4 downlights that emit a faint light (visible at night) even though they are switched off. The lamps are rated at 16W and are conected to an LED driver rated to output between 20 and 90V DC. The driver is compatible with dimmer switches although it is just connected to a standard light switch. I have checked the switch and is wired to switch the live with neutral passthrough which can apparently cause this issue if it is wired with neutral switching. I have also checked the sockets in the roof and they are also wired correctly. When switched off, the sockets have 5V AC (240V AC ON) so I think the induced AC from parralel circuits is what is causing this issue. It isn't practical to rewire the light circuit to reduce this and would like to try to salvage these lights as they perfrom well otherwise.

Is there something I can add to the LED driver circuitry to block or mop up the stray power being delivered to the LED lamp when switched off?

 

A fairly common problem, even with the older CFLs, which would slowly charge their input filter cap until the inverter triggered and you would get a brief pulse from the fluoro tube.

But anyway, for LEDs, yes, it is usually capacitive leakage through the long runs of paralleled cables, I have a LED fluoro tube in the laundry that never turns off completely. Probably the simplest solution is a resistor across the driver input to drain the leakage current. The current is low, so maybe start with something like a 22k, 2W (for 120V mains, higher for 230V) and see how that goes. Increase the resistance value until the problem comes back and then drop back to highest value that worked. Just remember that when the light is on, the resistor is connected directly across the mains, so it will need to be suitably rated for voltage and power.

 

You could use a two way switch, and in the off position the load connects to neutral via a 10K resistor, so in effect it's shorting the fitting out, discharging any stray voltages,i have done this in the past .

 

You could use a two way switch, and in the off position the load connects to neutral via a 10K resistor, so in effect it's shorting the fitting out, discharging any stray voltages,i have done this in the past .

That's a nifty solution. I think my light switch doesn't have a two way switching option but will recheck. Would using a double pole switch that also disconnects the neutral aslo solve this and wouldn't need a resistor to bypass?

 

That's a nifty solution. I think my light switch doesn't have a two way switching option but will recheck. Would using a double pole switch that also disconnects the neutral aslo solve this and wouldn't need a resistor to bypass?

It's the same if you disconnect the neutral, the stray voltages induced in the wires, a resistor is better, try values from 1K to 10K to see which is best.

 

Thank you for that diagram, super helpful! I checked my switch again and it does have an extra terminal. Unfortuantely wiring in resistors the way sugested doesn't make a difference which is baffling. I tried from 10 to 10K ohms without any difference. Super weird. What does work is seperating the neutral connection from the mains and even a tiny air gap stops the glow. I'll buy a double pole switch and see if it works.

 

You could use a two way switch, and in the off position the load connects to neutral via a 10K resistor, so in effect it's shorting the fitting out, discharging any stray voltages,i have done this in the past .

I've used that before, but without the resistor. If your switch is in the route of the cable (rather than being looped-in from the light fitting, it gives you a convenient terminal to join the two neutrals together.
As anything with a switched-mode supply will also have a bridge rectifier, there is no danger in discharging a big capacitor when connecting live-out directly to neutral. SPDT Light switches are always break-before-make, however, it might not be a great idea if the luminaire has a large power factor correction capacitor.

 

I have the same issue with a light fixture that had an associated dimmer. Previously incandescent lights that I switched to LED lights. In the full off position the LED lights were still minimally lit but enough to still light up the room. I figured it was because the dimmer didn't really go all the way to 0% pulse width, maybe only 5%, but that was enough to keep the LED's lit because of their efficiency. I haven't tried a dimmer specifically for LED circuits, maybe that is the solution?

 

Thank you for that diagram, super helpful! I checked my switch again and it does have an extra terminal. Unfortuantely wiring in resistors the way sugested doesn't make a difference which is baffling. I tried from 10 to 10K ohms without any difference. Super weird. What does work is seperating the neutral connection from the mains and even a tiny air gap stops the glow. I'll buy a double pole switch and see if it works.

Are the live and neutral not reversed by any chance by mistake?

 

Live and neutral are correct on my multimeter. As part of my testing I connected to the live but disconnected the neutral and the glow was bright enough to see in a sunny room. Isn't capacitive coupling neat

 

Just an update. Have installed a double pole light switch and this has fully resolved the issue. Easy and fairly inexpensive fix and fully compliant.