3 way switch weirdness

[Malamute]

A friend had a ceiling light that stopped working. Basic multi-meter showed both 3 way switches had power and it alternated which terminal when they were switched. OK... Check power at ceiling box, 120v, switches seem to work fine with alternating on-off and both switches activated alternately. Track light doesnt work. Has 120v at the button that twists into the track, but the light wont work, new bulb....tried same light in another track, it worked. She got a new track, it wont work.....after some puzzlement, I kept the meter on the track when switching the actual track fixture on, the 120v goes to 10v and no light. What the WTF?

Is it a weird switch issue/failure? Just replace both switches and drive on?

ive seen them fail by not passing power through, and only one working, but both seemed OK when using the basic analog meter, never seen one fail with a load but otherwise appearing OK.

Ive disassembled the actual spot type fixtures, nothing seemed wrong, the manual switch worked fine, all contacts seemed intact and properly arranged.

[rdtompki]

loose wirenut/connection on return would be my guess.

[Malamute]

Thanks, that had occurred to me, but will have to wait until I can go back over.

[tadawson]

If you can meter across segments, switches, etc. you can get a better idea where the fault lies by finding the point where voltage is lost and should not be. IE, if the light is mounted to a grounded box, try reading hot to neutral as well as hot to ground. As noted, since you say that position does not matter, this fault is clearly on a single wire segment, and could still be a swich, since I think they use a single contact to connect to either pole.

(Myself, I'd check the switches *first*, since they are the only non-passive devices involved. Short of rodents, wire rarely just fails sitting there undisturbed.)

[Malamute]

I checked the fixture at both ground at box (obviously wired to ground wire) and neutral, same result.

I didnt check the switches under load, just that they were passing 120v and seemed to operate normally before going dowstream to the fixture. I told her to get 2 new switches, in the interest of time.

Ive seen breakers get sort of weird, half of a main 220 breaker fail (half the house works the other half doesnt, or not at full power), and even brand new GFCI outlets fail right out of the box, just not seen a switch fail in a load situation. Not saying its not possible, just puzzled me, hence asking if its possible. Some electrical things have voodoo or something going on.

[Wheeler]

Is this ceiling fan and track light on the same set of 3-way switches? Sounds like you've lost a neutral somewhere, probably a bad tap or burnt up wire nut. I'd check the taps in the box supporting the fan, those are usually a failure point, especially in the older ones that didn't have support hooks to hold the motor while wiring.

[whomever]

"I kept the meter on the track when switching the actual track fixture on, the 120v goes to 10v and no light"

Dunno if this is your problem, but when you have two wires running parallel with one energized and the other not, the energized one will induce a potential on the non-energized one. You can see, say 90V, but it's at negligible amps if you connect a load, so connecting any load drops the voltage a lot.

3 way switches are prone to exhibiting this when run with 3 conductor romex, for example. I usually see maybe 80V dropping to 0 on load, but those values probably depend on the lengths involved etc.

Not an EE, may have garbled the explanation, just scratched my head over it the first time I encountered it and asked a cluefull friend.

[Caballoflaco]

I have seen some of the bad Chinese sheet rock cause corrosion where wires were connected into switches and outlets. However, if that’s the case here you should pretty easily be able to see verdigras and corrosion on any exposed copper or brass.

We had to strip back all the wiring to clean copper and replace all of the affected switches and outlets.

[Malamute]

Is this ceiling fan and track light on the same set of 3-way switches? Sounds like you've lost a neutral somewhere, probably a bad tap or burnt up wire nut. I'd check the taps in the box supporting the fan, those are usually a failure point, especially in the older ones that didn't have support hooks to hold the motor while wiring.

No fan, just the single ceiling box with a track light on it.

Ill have to look in the switch boxes again, I was mainly looking to see if the switches had power and seemed to be functioning normally, which they did, and power was getting to the ceiling box, so I closed them back up.


Yes, ive seen a number of ceiling fans set on non-fan rated boxes.

[UNK]

If I suspect a switch I hold a screwdriver by the the shank and tap the toggle with the rubber coated handle.

[UNK]

A friend had a ceiling light that stopped working. Basic multi-meter showed both 3 way switches had power and it alternated which terminal when they were switched. OK... Check power at ceiling box, 120v, switches seem to work fine with alternating on-off and both switches activated alternately. Track light doesnt work. Has 120v at the button that twists into the track, but the light wont work, new bulb....tried same light in another track, it worked. She got a new track, it wont work.....after some puzzlement, I kept the meter on the track when switching the actual track fixture on, the 120v goes to 10v and no light. What the WTF?

Is it a weird switch issue/failure? Just replace both switches and drive on?

ive seen them fail by not passing power through, and only one working, but both seemed OK when using the basic analog meter, never seen one fail with a load but otherwise appearing OK.

Ive disassembled the actual spot type fixtures, nothing seemed wrong, the manual switch worked fine, all contacts seemed intact and properly arranged.

Figured it out yet?
Im thinking something in the hot side wiring. First thought is a switch or breaker most likely switch. I think Id try turn circuit off, pull switches out of wall still wired, turn circuit back on then once I had low voltage Id go to the last 3 way and check output side then input side and keep going back till I found the problem point. Ive seen this a few times and its always been at a switch or at a fuse.

[Malamute]

Figured it out yet?
Im thinking something in the hot side wiring. First thought is a switch or breaker most likely switch. I think Id try turn circuit off, pull switches out of wall still wired, turn circuit back on then once I had low voltage Id go to the last 3 way and check output side then input side and keep going back till I found the problem point. Ive seen this a few times and its always been at a switch or at a fuse.


I havent been able to address it yet, her schedule hasnt had an opening. The plan at this point is to replace both switches, see what happens, and go from there. Ill report back on results.

[UNK]

I havent been able to address it yet, her schedule hasnt had an opening. The plan at this point is to replace both switches, see what happens, and go from there. Ill report back on results.

If one was bad id replace both since the are both about the same age but one could get a lot more use than the other. Id still try tapping on the toggle before i replaced anything to try and isolate it. If that didnt show anything id do the voltage check as outlined above. Hell it could go all the way back to the breaker. Is this a fuse or breaker panel?

[Malamute]

If one was bad id replace both since the are both about the same age but one could get a lot more use than the other. Id still try tapping on the toggle before i replaced anything to try and isolate it. If that didnt show anything id do the voltage check as outlined above. Hell it could go all the way back to the breaker. Is this a fuse or breaker panel?


Roger that.

Nothing else seems to be operating in a flaky fashion. Theres likely a number of things on the circuit.

House is maybe 20 years old, so breaker panel. I think shes been in it about 6 years, it worked when she moved in, she doesnt recall when it stopped working, she just asked me about it recently.

[Wheeler]

Take a keyless fixture with you with an incandescent bulb in it. Wire that to the wires coming from the ceiling box and see if it works. If there’s a bad rap or you think you’re getting some sort of inductive reading that will tell you pretty quick. Might save you from having to replace switches and chase ghosts.

When you replaced the track did you replace the power head as well?

[Malamute]

Take a keyless fixture with you with an incandescent bulb in it. Wire that to the wires coming from the ceiling box and see if it works. If there’s a bad rap or you think you’re getting some sort of inductive reading that will tell you pretty quick. Might save you from having to replace switches and chase ghosts.

When you replaced the track did you replace the power head as well?


Yes, replaced the power head when replacing the track. Both old and new ones seemed fine (120v at the contacts) by basic analog tester check for voltage. Both sets of lights had LED bulbs, which didnt illuminate at all. Ill check with an incandescent next time i look at it.

I wired the female end of an extension cord to the wires in the ceiling box, plugged a radio in,...and the voltage dropped to 10v when the radio was turned on.

[Wheeler]

Yes, replaced the power head when replacing the track. Both old and new ones seemed fine (120v at the contacts) by basic analog tester check for voltage. Both sets of lights had LED bulbs, which didnt illuminate at all. Ill check with an incandescent next time i look at it.

I wired the female end of an extension cord to the wires in the ceiling box, plugged a radio in,...and the voltage dropped to 10v when the radio was turned on.

You have a bad connection for sure. While you have the switches out check the neutral taps. Most residential electricians bury those in the switch box rather than the fixture box.

[UNK]

Take a keyless fixture with you with an incandescent bulb in it. Wire that to the wires coming from the ceiling box and see if it works. If there’s a bad rap or you think you’re getting some sort of inductive reading that will tell you pretty quick. Might save you from having to replace switches and chase ghosts.

When you replaced the track did you replace the power head as well?

Whats a rap?

[Borderland]

I havent been able to address it yet, her schedule hasnt had an opening. The plan at this point is to replace both switches, see what happens, and go from there. Ill report back on results.

Switches are cheap. They're built that way.

[Wheeler]

Whats a rap?

Typo. Should have been “tap.”

[farscott]

Yes, replaced the power head when replacing the track. Both old and new ones seemed fine (120v at the contacts) by basic analog tester check for voltage. Both sets of lights had LED bulbs, which didnt illuminate at all. Ill check with an incandescent next time i look at it.

I wired the female end of an extension cord to the wires in the ceiling box, plugged a radio in,...and the voltage dropped to 10v when the radio was turned on.

The voltage drop with a load means there is a high resistance connection in the current loop (hot and neutral). High resistance connections lead to fires as the losses are the current squared times the resistance. So finding this is key. The circuit should be switched OFF until the problem is resolved.

Another way to think about this. If you measure 10VAC with the radio connected and measure 120V without it, 110V is being dropped across the wiring, switches, and connections. That is a very high resistance connection.

A couple of ways to tackle this. The way I like is simple but labor intensive. With the breaker OFF and the multimeter set to resistance, I probe across every connection (screw terminal, other terminals, switch, wire nut). A good connection should have fractions of an ohm of resistance. Check every one. Start with the switches and any wire nuts. Look and smell for signs of heat discoloration or burning. Anything close to an ohm is suspect and should be redone or replaced, but you are looking for a much higher reading. If the connections are all good, the copper wiring is suspect and should be checked. Usually it is a bad switch or bad connection. I have had bad outlets where the female terminal in the outlet was too large, leading to a poor connection. Found it by smell.

The other way to do this is to use a thermal imager. With the circuit energized and with the load attached, look for the hot spot. Have done this tracking down issues in the walls.

[Malamute]

I havent updated yet, i went over last weekend, replaced one 3 way, went to the other one....wait, its not a 3 way....its a 4 way. What the WTF?

Had a consultant available, he said it means theres another switch in the circuit, we hunted around and finally located it, a dimmer. Messed with it some, it seemed to be flaky, so I replaced it with the other 3 way. The dimmer didnt have an obvious indication of which was the common, so I made a guess. i was wrong, but with the switch in one position, it all worked without the voltage crashing with a load,....except the track still isnt making connection to the light fixture. We ran out of time, so its on hold until she has time open again.


In light of the previous post,
The voltage drop with a load means there is a high resistance connection in the current loop (like a dimmer?) it appears at least the house isnt going to conflagrate.


Thanks for all that offered information and advice, and thanks especially to my consultant for being available and helping figure this out.

[farscott]

The dimmer is likely the source of the voltage drop. The presence of a dimmer changes the issue and may eliminate the high resistance connection concern. Modern dimmers work by pulse-width modulating the input AC waveform, so the resulting output waveform has a voltage proportional to the dimmer setting. Older dimmers increased resistance to decrease the light brightness and were designed to dissipate power across the resistance.