I’m not an electrical expert by any means, but in replacing some illuminated switches in my house, this is what I discovered. The old and simple kind of illuminated switches, which used a tiny neon tube, pass a small amount of current through to the bulb. There might be a way to build the illuminated switch so that doesn’t happen, but that’s not how they were designed.
When I was replacing these, about 5 years ago, I looked for drop in replacement illuminated switches that would be compatible with, at the time, CFL bulbs, and found they basically didn’t exist. None of the $5 glowing switch types were compatible. I believe there were much more expensive LED plate covers, but I decided to just replace everything with non-illuminated switches.
What you’re missing is that you are talking about ‘switch loops’, where the power wired (including the neutral) come to the light fixture, and then a loop (in the hot wire) runs down to the switch & back to the light.
These were classed as depreciated (outdated) 7 years ago, in the 2011 National Electric Code. Since then, both the hot 7 neutral are required at the switch box, and from there proceed to the light fixture.
But even so, I think it will work. The neon bulb completes the circuit when the switch is off, and allows enough current to flow to light the neon bulb (but not enough for an incandescent bulb to glow).
I have a couple of Leviton branded rocker switches with a neon light that have a neutral terminal to operate it, but it lights up in an on position (I use them to control a fluorescent strip in the closet so I can see if it got left on with the door closed.
With the new code you can still use switch loops, but you need three conductor Romex instead of two- no more using the white wire as a hot. You connect the white wire to the other neutrals at the light fixture and then cap it off at the switch location if you’re not installing an electronic or illuminated switch that requires it.
Don’t get hung up on the term “switch loop”. Every (switched) light circuit has, as one of it’s components, a “switch loop”. Some people feel the only time this term is applicable is when the hot supply for the switch is fed from the fixture, but this is simply incorrect. A switch loop is merely the control side of a switch, it matters not the geography of the hot supply.
As Mdcastle said: the NEC does not disallow a fixture-sourced power supply for a switch. What you are referring to simply requires the neutral to be extended from the fixture into the switch box. In other words, you now must pull a 3 wire cable instead of a 2 wire cable as in the olden times.
What any of this has to do with my original comment is beyond me. Semantics and the NEC aside, I question the contention that the current drawn by a switch illuminator somehow finds its way through the light fixture with the switch open (off). Unless…
Perhaps the odd performance(s) you describe is related to the fact (in the old days) that the illumination for such switches used ground instead of neutral to complete their circuits. This stray current flow in the ground system might be the culprit that causes weird things to happen to LEDs or other modern electronics installed in older light circuits.
But, that;s exactly how the switch illumination works.
A Neon lamp in the switch is wired across the switch terminals. When the switch is ON, there is zero volts across the terminals. When the switch is OFF, there is 120V across the terminals, and the lamp lights up (there is a high-value resistor in series with the lamp to limit the current to a few mA). This trickle of current is enough to cause LED lamps to flicker or glow very dimly.
An update. I removed the LEDs and replaced them with clear Reveal HD Light 60 Watt replacement, 43 Watt usage, Enhanced Spectrum Halogen Bulbs. Things are back to normal. But to make sure, I set-up my iPhone to do a selfie of the ceiling it to record video while I left and room and then returned. The lights with no motion turn off, stay off, and nothing is dimmed like with the LEDs. When I returned the lights came back on.
Who knew that an iPhone would be so useful in testing lights. I think I’m going to put it in the fridge next to confirm the light there really goes off when the door is closed.
Next house, when having new electrical work done, I’m going to insist on making things LED compatible.
I have a hunch that the current flow is SO SMALL that the incandescent filament will not produce light, but current could be enough to turn the electric meter VERY slowly…just like the electric company want to see to pad their bill. The LED does light up because it has a lower threshold of light to current - but the low power might be damaging to its circuit.
