LED always glowing

[mrcharly-YHT]

We've replaced some bulbs in this house.

One bulb glows faintly when off. It is on a double switch (two switches, either end of a corridor).

I'm thinking that this could be because:

Live and neutral are swapped somewhere.
One of the switches is on the neutral wire (rather than live).

Any other suggestions?

[grams]

One bulb glows faintly when off. It is on a double switch (two switches, either end of a corridor).

Induced voltage in the two-way wire path. Some bulbs glow, some don't. Switch to a different bulb if it annoys you.

Live and neutral are swapped somewhere.
One of the switches is on the neutral wire (rather than live).

I don't think it's possible to build a two-way light switch where both of these things aren't true.

[ppg]

One bulb glows faintly when off. It is on a double switch (two switches, either end of a corridor).

Induced voltage in the two-way wire path. Some bulbs glow, some don't. Switch to a different bulb if it annoys you.

Yes - we changed a row of outdoor garden tungsten bulbs for LEDs and they flickered annoyingly.
Replacing just one with a 40W filament to provide a higher loading cured the problem.

[Kim]

One bulb glows faintly when off. It is on a double switch (two switches, either end of a corridor).

Induced voltage in the two-way wire path. Some bulbs glow, some don't.

This.

Solution (other than a different lamp that puts a slightly heavier load on the circuit) is a bit of extra load in parallel with the lamp *at the fitting*.  A suitably-valued resistor (which could be a tungsten lamp) or capacitor would do it.  A mains-rated snubber capacitor would be a good choice...

Ah, here we go, this sort of thing's available off the shelf for just the purpose: https://www.tlc-direct.co.uk/Products/DNCAPLOAD.html

[Asterix, the former Gaul.]

Apparently telephone lines can be tapped by induction.

[Kim]

Apparently telephone lines can be tapped by induction.

Sure.  All you have to do is get your current transformer between the two wires so their respective fields don't cancel each other out.

[grams]

Back in the day they sold transformery things you could glue to your phone base to record calls. I'd love to know if those worked, and if any were ever sold to non-weirdos.

[Kim]

Back in the day they sold transformery things you could glue to your phone base to record calls. I'd love to know if those worked, and if any were ever sold to non-weirdos.

Your phone handset has a loudspeaker in it.  Paper cone attached to a coil of wire, wiggling in the field of a magnet.  Put another coil inline with it, and the varying magnetic field creates a voltage, just like a microphone would.  Commercially available as a little device with a suction cup for ease of attaching to random telephones.  Popular with journalists and the like.

At some point in the days when everything was orange and brown, some clever person realised that if you equipped a hearing aid with such a coil of wire, and a switch labelled 'T' to select it rather than the usual microphone, the user could hold the telephone handset up against their hearing aid and hear the call without any risk of squealing from acoustic feedback.

This quickly became a standard feature, and another - slightly less clever - person came along and reasoned that if you filled a room with a varying magnetic field, you could send audio from a microphone or television or whatever directly to people's hearing aids, improving the signal-to-noise ratio by eliminating background noise.  The 'Induction loop' was born - literally a piece of wire around the edge of a room, service counter or similar, fed by current modulated by baseband audio.

Unfortunately, it stopped being the 70s, and the real world started to fill with devices that produce stray audio-frequency magnetic fields:  CRT screens, DC power supplies, dimmable lighting, etc. etc.  Hearing aid users learned to ignore the 'T' setting, and people forgot all about that loop thing.  The microphone got unplugged by the cleaners, but because everyone likes to help the disabled, the sign indicating that a loop was available remained in pride of place.

Now the telecoil is an optional mode that you have to ask the audiologist to enable when programming the hearing aids.  And most people don't.  Hearing aid manufacturers are, however, in a race to see who can come up with the most irritating implementation of The Devil's Other Radio^[1].

Meanwhile telephone calls can easily be recorded by writing a copy of the data stream to a file.

And LEDs come in blue now.


_{[1] Bluetooth.}

[Feanor]

Back in the day they sold transformery things you could glue to your phone base to record calls. I'd love to know if those worked, and if any were ever sold to non-weirdos.

Analog phones have a curious device in them called a Hybrid Transformer.

These are a thing that mix ( and un-mix ) the 4-wire ( microphone, speaker ) to the 2-wire phone line.
They will also mix back a certain portion of your microphone to the earpiece, which is more 'natural' than not hearing your own voice mixed with the other end's voice.

This transformer emits a magnetic field which can be picked up with an external coil.

[Kim]

Ah, I missed the 'base' part of that.  Clever things, those.  Of course, sidetone went out of fashion a couple of decades ago, but the modern phones that do it probably use op-amps.

[grams]

This transformer emits a magnetic field which can be picked up with an external coil.

Yes, this is the thing.

Before I wasted my life on the internet I wasted my life perusing the Innovations catalogue* and the telephone base with a coil randomly stuck to it was always a "hang on, what?" object of particular boggling.

Neither the Innovations catalogue nor this particular type of telephone pickup seem to exist on the modern internet.

(* I wasn't alone. You'll note one went away as soon as the other gained popularity)

[Diver300]

One bulb glows faintly when off. It is on a double switch (two switches, either end of a corridor).

All practical implementations of two-way switching end up with live and switched-live wires side by side, with no earth wire between them, the whole way between the two switches*. Any capacitive leakage between the two will put a tiny current through the load.

Many LED lamps have a capacitor inside them, in series with the rest of the circuit, to limit the current. The capacitive leakage is effectively making the lamp's capacitor much smaller, resulting in much less current and light, but not doing any harm.

It's going to cost approximately nothing to leave the lamp glowing.

There are various designs of LED circuit. Some won't glow when there is leakage current, while some will flash annoyingly

*some implementations will only have that for some switch states.

[Feanor]

One bulb glows faintly when off. It is on a double switch (two switches, either end of a corridor).

All practical implementations of two-way switching end up with live and switched-live wires side by side, with no earth wire between them, the whole way between the two switches*. Any capacitive leakage between the two will put a tiny current through the load.

Indeed. time for my old hand-sketch from some time back...


IMG_20151107_181740[1] by Ron Lowe, on Flickr

[mrcharly-YHT]

Cheers all.
I've learnt something.

the bulb in question is in a hall light - so it isn't really a bad thing that it is a dim night light.

[Kim]

If you've got a LED lamp that glows politely (rather than flashing) at low currents like this, you can deliberately use that as a night light by putting a small capacitor across the switch.

Big Clive discussed this recently:  https://youtu.be/ISTB0ThzhOY?t=1342

[Beardy]

LEDs glowing on a two way lighting circuit is as discussed down to the inductance and capacitance of the extra run of wires creating a tuned circuit, which again, as discussed can be unbalanced by providing extra resistance, inductance or capacitance in the circuit. Better quality LED ‘bulbs’ have an extra resister in the circuit to suppress this issue, and poorer quality LED ‘bulbs’ will glow even on shorter single switched circuits.

Basic telephony has concepts such as cross talk (signal ‘bleeding’ from one pair to another via inductive coupling) which the less scrupulous can use nefariously to eavesdrop on others telephone conversations. There are easier although less clandestine ways to achieve the same goal. Both methods are becoming obsolete though as the analogue ‘local loop’ gets replaced by digital systems. Another important basic telephony concept is ‘side tone’ which is the deliberate feedback of your own voice in your earpiece. It was learned very early on in telephone use that people are used to hearing themselves talk and when you’ve got a telephone pressed up to your ear the natural audio feedback is blocked. The GPO/PO/BT had a research department dedicated to such matters. Over the years as telephone manufacture and sales has been relaxed some manufactures have left this out (it requires design and extra components because it needs to be balanced properly) and have subsequently had lots of returns as faulty. Even mobile phones have sidetone designed in.

[Regulator]

Forget all that well-intentioned but complete nonsense above....




..it's ghosts.  You know it is!




 

[Diver300]

LEDs glowing on a two way lighting circuit is as discussed down to the inductance and capacitance of the extra run of wires creating a tuned circuit,

I don't think it's that complicated.

It's just the capacitance between adjacent cores in a cable producing leakage current.

I have accidentally created a tuned circuit at mains frequency, but that was a with a huge inductance, a relay coil, and a suppressor capacitor. You need loads of inductance to get a tuned circuit at 50 Hz. Accidental tuned circuits happen far more easily at higher frequencies.

I agree that a resistor will be fitted inside many lightbulbs to stop the glow.