GFCI outlet + wall switch wiring

electricalgfciwiring

I'm trying to install a GFCI outlet into a bathroom that currently has a regular (non-gfci) power outlet, but I'm struggling.

My bathroom electrical has a single wall outlet (standard 2 plug receptacle), a ceiling fan, a wall light, and 2 wall switches to operate them. All of this seems to be on the same circuit breaker, as when I trip one breaker switch, all 3 of these electrical items cease to function.

GFCI model: Leviton GFNT1-3W
https://www.amazon.com/gp/product/B01B8JE4YO/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

I removed the existing wall outlet and the box had 2 cables inside (2 wires ea). Using a voltmeter, by putting the probes across the white and black wire of each cable, only 1 cable showed voltage, so I determined this to be my LINE cable. To clarify, the LINE wires are both part of the same romex cable, and come in on the right side of the box. The other romex cable comes in on the left side of the box, so they appear to be distinctly different cables.

So, I installed these LINE white and black wires to the LINE terminals on the GFCI, and the other 2 white and black wires to the LOAD terminals. After turning the circuit breaker back on, the GFCI would not reset. When I would press reset, the led would light for a split second, and a lamp I plugged into the GFCI would turn on for a split second, but then the GFCI would trip immediately.

I then removed the LOAD wires, leaving just the LINE wires connected to the GFCI, and tried again. It now operates properly and will test & reset and expected. So, I assume I have some issue with my load circuit.

To troubleshoot, I voltage tested more wire combinations and found something perplexing.
For this test, no wires were connected to anything, so no GFCI connected at all.

With test probes across:

  1. LINE-black + LINE-white = 125v
  2. LOAD-black + LOAD-white = 0v
  3. LINE-black + LOAD-white = 125v
  4. LINE-black + LOAD-black = 0.1v

#3 & #4 were very unexpected to me. #4 starts out at a few volts and then gradually decreases down to 0.1v over the course of a few seconds. If I disconnect the probes, and then try again a few seconds later, the voltage jumps back up to a few v and then ramps down again. (Maybe there's a capacitor in the ceiling fan causing this, and its getting charged by my voltmeter probe tests?)

Up until this point, I ASSUMED my load wires ran to my wall switch, which then controlled the fan and light. I felt this way because with the LOAD wires disconnected, the switches cant turn the fan or light on, supporting my theory that they feed the wall switch. But I don't understand how this can be true given my volt meter readings on #3.

btw – the wall receptacle I'm trying to replace was wired with both black wires on one side, and both white wires on the other side. The little copper tab between the 2 white wire terminals and also between the 2 black wire terminals was intact and was not broken/removed, so I imagine the circuit behaved as if the 2 black wires were twisted together, and also the 2 white wires twisted.

I'm confused and not sure exactly how to proceed at this point. Any ideas how to proceed with troubleshooting, ultimately to wire this properly?

Best Answer

Prior to "With test probes across", you are plainly describing a situation where the downline has an actual ground fault. The GFCI is tripping because it's doing its job; same as when a Radon detector goes off because there's Radon.

With the wires disconnected as you have them, Load hot and neutral may be connected to each other (through the load(s)), but they should be totally isolated from ground and any hot or neutral in any other circuit. If you put a hi-pot "megger" insulation tester with one lead on neutral and the other lead on ground, it should read infinity ohms, or at least several megaohms.

The problem is, it does not.

This is due to either

  • a faulty appliance (which I gather are all hard-wired)
  • faulty wiring in the walls, screw through a cable, that kind of thing.
  • neutral contacting bare ground somewhere in a box - normally this is undetectable since neutral and ground are near the same voltage, but GFCIs suss it out.
  • Neutral from this part of the circuit being bridged to another circuit's neutral -- in any given box, combining all grounds is correct, but some people confuse neutral and ground, and combine all neutrals too. Wrong.
  • Someone is bootlegging ground. The usual reason is they want to run a fan and light under separate control, but there is only /2 cable in the walls.

I would say your measurement #3 points to one of the last three.