I'm installing two new bathroom light fixtures. First one went in fine. Second one will blow the wall receptacle based circuit breaker the moment I screw in the bulb (in fact three different bulbs tried).

So for clarity both fixture circuits run through this wall receptacle based circuit breaker. It has a little green light to indicate it is "on". I measured the new fixture and found 115-120v on the light socket. So with no bulb in I can turn on the fixture and measure the voltage, but screwing in the bulb (three different bulbs, which then work fine in the first fixture) causes it to trip. This happen if the first fixture is turned off or if it is turned on. Help?

I thought if I had a short I would never be able to measure the voltage at the socket, or that it would trip when I turned on the switch without a bulb. Doesn't trip.

I thought maybe something wrong with fixture, so returned it and got another, same problem.

This one has got me.

Wall socket

enter image description here

enter image description here

So I opened up the box and found that the light fixture (and switch) is wired to the line connections on the GFCI, and the true "hot" wires are wired to the load connectors on the GFCI. In the following diagram just reverse the labels for Load and Line on the GFCI.enter image description here


I opened up the switch and outlet under fixture number 1. Here I see the HOT lines going to a standard outlet and switch. These lines are then fed through the back of the wall over to the box under fixture number 2, where as I said earlier they go to the LOAD on the GFCI and my problem fixture (#2) is wired to the LINE terminals.

So my assumption is that due to this Fixture #1 never really had GFCI protection(?) Not sure I understand why it turned off when the GFCI tripped.

Can I fix the fixture #2 problem by just switching the incorrect(?) Load and Line terminal wires? and shouldn't this also mean my fixture #1is also now GFCI protected?


  • Hello, and welcome to Stack Exchange. Would you post a picture looking into the suspect socket? Nov 13 '18 at 0:05
  • @DanielGriscom, the fixture has been replaced as per second last sentence
    – jsotola
    Nov 13 '18 at 0:09
  • what type of breaker are you using?
    – jsotola
    Nov 13 '18 at 0:10
  • 2
    It isn't a short - it is a ground fault. On a new fixture, that would typically mean something wrong in the wiring - e.g., hot/neutral reversed, ground touching neutral, etc. Not sure why the meter isn't triggering it, but it may use less current than needed to trip the GFCI. Double check all the wiring. If you can't figure it out, post a picture of the guts of the box - i.e., all the wires going into the GFCI/receptacle and the switch, as well as any other junctions involved. Nov 13 '18 at 1:14
  • 1
    That receptacle thing is not a circuit breaker and it doesn't care how much current flows. It has one job, sniffing out a specific condition called a ground fault, that often gives people dangerous shocks. You can hook your whole darn house to ground fault detectors, but you'll find every ground fault in the place if you do. Your ground fault detector may also be miswired, it is particularly easy to mess up when you have a light switch next to a receptacle. Nov 13 '18 at 4:11

The GFCI is miswired

It's quite common for GFCIs to be miswired in bathrooms where a light switch is right next to a receptacle.

I bet if you look at the backside of the GFCI, you will find that the "For Wizards Only" warning tape has been removed from the LOAD terminals, and a wire has been connected to one of them but not the other.

I would speculate it is the white wire on LOAD, and the black supply wire loops through the switch and then hops to the GFCI, or vice versa.

This is typically done when a novice installer has hooked up the GFCI, but has a wire left over. They see a spare screw behind the "Not For Novices" tape, and go "this is the easiest way to solve my problem, what could go wrong?" In your case they never hooked up a fixture up there, so they were never aware of the problem.

Since lamp current bypasses the GFCI on one leg, but goes through it on the other leg, currents are unequal and the GFCI trips.

The answer is to rewire the switch and GFCI relationship so either a) LOAD is not used, or b) whatever is attached to LOAD is wired entirely off LOAD and does not partially bypass it.

  • if this is true, then how is it that “first one went in fine” ? Re-read OP’s first two sentences only. He seems to be saying fixture #1 works just fine, and the problem presents itself when a light bulb is screwed into number 2. We really also need to know if these are replacement fixture— how did this work before?
    – Tyson
    Nov 13 '18 at 16:10
  • Probably on the neutral side: the first one is returning current where it's supposed to, the second one is not. I'm speculating it's a case of "extra-terminal-itis" where the installer thought LOAD terminals were spares like on normal receptacles, needed a spare so he used it, and never tested. Nov 13 '18 at 16:23
  • This is a new electrical install when we had to replace the drywall. Both fixtures and wiring are new and have never been turned on before. Nov 13 '18 at 20:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.