1

We have a working GFCI outlet in the basement near a tub that the washing machiner empties into.

We'd like to install an overhead utility light, controlled by an on/off switch, protected by the GFCI circuit. I figured one terminal of the on/off switch must be connected to BLACK LOAD on the GFCI and one terminal of the on/off switch goes to the light fixture. The light fixture's WHITE would return to WHITE LOAD on the GFCI receptacle. But this doesn't work. The GFCI always pops.

What is wrong with this diagram? Thanksenter image description here

P.S. The GFCI trips as soon as power is returned when the breaker switch is flipped in the breaker box.

P.P.S. So, I removed the light fixture and the GFCI still tripped when power was returned from the breaker, and re-tripped when I pressed Reset. However, I tried moving the WHITE wire coming back from the (missing) fixture onto the Line WHITE connector, from its previous position on the Load WHITE connector. And now, no more tripping. The whites could be pigtailed inside the box --right?-- so this is not unsafe?

Improve this question
12
  • Does the GFCI trip (pop) as soon as you reapply power to the circuit, or only when you try to turn the light on?
    – ThreePhaseEel
    Aug 16, 2015 at 19:44
  • When you flip the breaker on and the GFCI trips, is the light switch on or off? (does that make any difference)
    – Craig Tullis
    Aug 16, 2015 at 20:51
  • A GFCI works by comparing the current on the hot and grounded/neutral conductors (there's a little inductive coil around each conductor inside the GFCI). If the current isn't identical on both conductors, it trips. Presumably, you either have a defective GFCI, or you have a short or maybe some kind of inductive current situation putting current on (probably) the neutral/grounded conductor. So, regardless of whether the light switch is on or off, there's always some difference in current between the hot and neutral.
    – Craig Tullis
    Aug 16, 2015 at 20:57
  • @Craig: I've tried two different GFCI receptacles, thinking one was possibly defective. You don't see any problems with the wiring diagram? That should work, all things being perfect?
    – user41524
    Aug 16, 2015 at 21:02
  • 1
    @Tester101: replaced the six-foot length of romex leading to the light fixture, and bingo! No tripping when the WHITE goes to Load White on the GFCI. It wasn't a new piece of romex, but a little piece left over, which was a bit kinked in places. Thanks for the help.
    – mr blint
    Aug 18, 2015 at 12:18

1 Answer 1

Reset to default
2

GFCI devices work by measuring the current flowing on the ungrounded (hot), and grounded (neutral) conductors. They do this by running both the conductors through a current transformer (CT), which produces a current on the secondary winding whenever there's a difference in current between the two primary conductors. So as long as both the ungrounded and grounded conductors are carrying the same current, there will be no current on the secondary of the CT. If there was a ground-fault, the current on the conductors would be different and a current would be induced on the secondary of the CT.

If there's no current flowing on the circuit conductors, then there's no way to induce a current on the secondary of the CT. Therefore, there's no way to trip the GFCI. So if you've removed the fixture, the GFCI should not trip.

If this diagram is accurate...

enter image description here

The only way the GFCI could trip, is if current was being introduced somewhere. Like if one of the conductors was shorted with a conductor from another circuit, or a different part of this circuit.

There has to be more to this situation, than the information you've provided. Simple explanations might include:

  • If the switch is in a box with other switches, and the grounded (neutral) from the GFCI was interconnected with the other grounded (neutral) conductors in the box.
  • A couple cables are stapled together, and the staple has shorted some of the conductors together.

Since you say that removing the white (grounded (neutral)) conductor from the LOAD terminal fixes the problem, I'd guess that the problem lies with that conductor.

If you have a really accurate ammeter, you could clamp it on the white wire and see if there's any current on it. To do this, you'll have to connect the wire to the line terminal (as you described in a comment). You'll only do this temperately, while you're taking the reading.

Alternatively, you could trace the wire and look for damage, or interconnection with other wires.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.