Author Options:

Residential 3-way switch problem with leaking voltage? Answered

I noticed a slight glow in the light when it was switched off. LED bulbs in a ceiling fan/light fixture. House built in 1905, wired whenever? using all-fabric insulation type wiring but not knob and tube .  Removed the load from sw 2 and tested travel wire voltages and got 95 v on one and 25 v on the other. Flip sw 1 and the voltages trade wires along with the switch position change.  The switches tested OK but replaced both anyway and got duplicate readings when repeating tests. Isolated the travel wires and got no continuity  (infinite resistance) using all scales of an analog meter as well as a Fluke digital multi-meter.. We checked for voltage feedback from another circuit and even used an extension cord to a different circuit in order to use a different ground wire. With the load connected, the travel wire being used shows 120 v but we still get 25 v on the other wire. I discovered this because when I bought the house only sw 2 worked the light, and one travel wire was disconnected at sw 1 and taped  off.. I have not opened the ceiling box at the fixture because the problem exists when the wire to it is disconnected.The previous owner said he didn't remember what sw 1 went to because it hadn't worked for many years.  
So we eliminated worn insulation or nail into the cable because either one  would cause a resistance reading below infinity. All we can think of is a legacy doorbell. Neither of us remembers working one, but I know some used 120 volts and a relay.  People have told me that relay contacts can cause this problem when they get worn.
Does anyone have any ideas? I'm retired on a fixed income and tearing walls open is an absolute last resort. 
Thank you in advance for any advice I can get!


1. Your house should be rewired as a matter of urgency.

2. More then likely you have a neutral fault allowing a small leakage current to feed back to the LED lights which do not draw much current

3. Measuring voltage on disconnected wires will usually give a false reading as the wire acts as a long aerial and it picks up stray current and this appears as a voltage on a modern meter which has a very high input impedance.

Assuming your in the UK - you may qualify for a grant to rewire.

The use of the term "Knob and tube" implies the OP is American, I don't think the UK used the system.

Very old UK wiring - fabric covered - used the ceramic screw pot type of connector. Sometimes Bakelite.

However as you say the OP is over the other side, USA wiring is a bit of a mystery to me, Why use aluminum? Never mind 110 volts.

I don't think Al was ever permitted in low voltage wiring in the UK, but I think it may still be used for high voltage transmission lines, I'm not sure.

Yep - Steel cored aluminium is still used on the National grid system in the UK. High voltage lower current, aluminium to improve the conductivity of the steel which is used for strength.

Still in use for High Voltage Transmission lines here with a steel core for strength.

I thought that might be how its done. Wonder how they make good joints ? That always seemed to be the big problem with Al in the USA. Any damp, the joint rotted, the resistance soared and you had a fire hazard.

Outdoors - maybe they have special crimping tools ?

They do use special crimping tools but that is also why we get 6 or more blackouts a year. Keeps my generator in good running order the ice storm we had last week had us on the generator for 48 hours.

I think rod and tube just about predates that. I've never seen it in an installation though

I am "across the pond" in the USA. so I won't get any help from the Gov't. But thanks for the tip, maybe it will help someone else.

Thanks for your reply.
The house was checked and GFI outlets were installed in bathrooms and kitchen. before I bought it.
All other outlets are 3-prong. I have not opened the breaker box (upgraded from fuses) to see if all cables coming in have a neutral as well as a separate ground. Given the older wiring, someone could have "cheated."
Would removing the fan/light and getting easy access to that box allow me to further isolate this problem? The neutral connections in the switch boxes appear to be soldered and taped.
The high impedance is the reason I duplicated all tests with an old analog meter, and esp. the continuity tests.
From what you said, maybe I should bite the bullet and have a contractor come in and give me a diagnosis. The friend who helped me so far is a journeyman but is more into industrial wiring. We must not have done a proper check for a neutral fault.


2 years ago

Leaking you say, .... well be just be sure to screw all the bulbs nice and tight to stop the gas I mean electricity from leaking out :-)

I'm inclined to think its just very low leakage currents and induced currents in the wires. Put a high value resistor, like 100K to ground on the travel wires, and remeasure, try 10K and 1K. If the voltage drops rapidly with increasing load, its purely high impedance, and likely just leakage, if it holds up, its more serious.