8

My parents got a ceiling light for my parents for Christmas, which I was tasked with installing. I don't know why I bothered to check the voltage before putting the new fixture on (and after taking the old one off,) but I did, and it read 123V when on (good, obviously,) but it read 38V when turned off.

I disconnected all of the wires and checked the switches, and they worked perfectly fine (complete discontinuity when broken). I did break one of the switches, however, when releasing the wires that they'd stabbed in the back (20 year-old plastic is sorta brittle, I guess,) so I got new 3-way switches and not surprisingly, this did not fix the problem.

I removed the wires from the switches, except I kept the line attached to the first switch and tested the switched terminals vs. ground. I got 123V on the active terminal and <.5V on the inactive terminal -- checked out just fine. As soon as I connected the wires that lead to the box with the second switch (which was not connected,) I got 37V on the wires that were supposed to be disconnected from the line. (Actually, one reads 55V when disconnected, and the other reads 37V when disconnected.)

Unfortunately, I do not have an insulation tester, but it would seem as though the insulation has failed on those wires (assuming no stupid mistake was made by a previous installer.) However, I am open to comments about why this might be happening. Any thoughts? Anyone ever had this happen before? Etc.

Tester101
  • 130,386
  • 77
  • 314
  • 605
Michael
  • 3,776
  • 4
  • 26
  • 28
  • 2
    which lines are you seeing the voltage on? Travelers, neutrals, etc.??? Seeing 40 to 60 vac is common with open neutrals. Did you ID the source lead and isolate it? Check travelers to neutral and ground at source switch? Can you draw the circuit and detail where you are seeing the phantom voltages? – shirlock homes Dec 27 '11 at 12:24
  • See also: http://diy.stackexchange.com/questions/10856/several-outlets-dont-work-on-same-circuit – BMitch Dec 27 '11 at 14:00
  • Make sure that the neutrals aren't being switched. – Brad Gilbert Dec 28 '11 at 01:47
  • Take a look at the article on [Stray voltage](http://en.wikipedia.org/wiki/Stray_voltage) on wikipedia. – Tester101 Dec 28 '11 at 18:17
  • @shirlock homes The voltage is present almost everywhere. As soon as I hook-up the travellers, one will read 123V and the other 37V or 55V (to both GND and neutral) depending on which is disconnected. This continues on to the luminaire itself, which has voltage (to both GND and neutral) even when turned off. The neutral is unbroken at all times. – Michael Dec 28 '11 at 19:36
  • Also, I am merely updating an existing installation. I could draw how I *think* it is wired (your typical three-way installation), but I don't know if anything else exists behind the walls and in the attic, such as a jbox. – Michael Dec 28 '11 at 19:38
  • Does the light turn off, or simply dim? – Tester101 Dec 28 '11 at 20:36
  • @Tester101 It turns off completely, but it has had a CFL in it for quite some time, which may require a certain voltage to light up. However, it also turned off completely back when it had an incandescent in it. – Michael Dec 28 '11 at 20:52
  • Have you measured the current on the "disconnected" traveler? You could also check the continuity (with them disconnected from the switches), to make sure they are not connected to each other. – Tester101 Dec 28 '11 at 21:03
  • Could you measure these voltages with a 100 watt incandescent bulb screwed into the luminaire socket at the time of testing, with the switches in each of the various combinations of on and off? Measure both wires (black hot and white neutral) relative to ground at the luminaire. I'm assuming you are North America though your profile gives no location. – Skaperen Dec 29 '11 at 05:30

3 Answers3

6

Do some research on capacitive coupling, and Inductive coupling. This should explain why you are seeing voltage on the "dead" traveler.

If you load up the traveler that is not carrying current; with a test light for example, you'll likely see this voltage go away.

Working in a residential setting, this voltage is likely harmless. Phantom voltages, however, can be lethal when working with distribution lines and other high voltage situations.

You may still want to have a licensed electrician come out and take a look, just to be sure this is actually the case.

Tester101
  • 130,386
  • 77
  • 314
  • 605
  • I got similar results recently testing with a regular volt meter and no load - the voltage goes away once you put any sort of load on it. – Grant Oct 24 '12 at 18:50
  • Awesome answer (and question). I was having the same "issue" and it was driving me crazy. – matteo Mar 28 '21 at 13:52
2

As was mentioned above, check your neutral source. I've seen the same situation and what cured it for me was tightening down the screw holding the neutral to the bus bar in the breaker box (after turning the box off of course). While I was at it I found a couple other screws holding down neutrals that were loose by a screw turn or less. It is always a good practice to check voltages before installing and/or replacing - one never knows who has been working on the electrical. Or as in the case of an older home, what condition the wiring, outlets, etc. are in. An odd voltage is usually a clue to something amiss.

Niall C.
  • 20,709
  • 17
  • 90
  • 131
PJaeger
  • 21
  • 1
1

I have experienced the same thing where when off certain brand of led bulb would glow on a hall light controlled by 2 3 way switches. I did all the outlined tests and everything checked out ok.

I tried an experiment with ~30’ of 14-3 rolled out straight. Connected the black and neutral only with no load and then measured a voltage on the third (red)conductor while both it’s ends were disconnected.

I believe this is simply due to induction of the changing electrostatic field. Any load, a second led, high ohmic resistor or neon bulb pulls down this voltage.

Bob Hyde
  • 11
  • 1
  • 1
    in your case, and likely the OP's too, it's due to capacitive coupling, not inductive – P2000 Jan 09 '23 at 21:49