I didn't do the wiring in this house, this has been happening since I moved in.

Basically, about 50% of the time that I turn off the bathroom fan, the GFCI outlet pops and needs to be reset. Once that happens, the fan no longer turns on until I reset the outlet. This only happens when shutting the fan off, not when turning it on or while it's running.

Does this likely indicate a problem with the wiring, the fan, or the outlet? I'm wondering what I should look at first.

  • Does hitting TEST on the GFCI cause the fan to stop working? Jan 27, 2019 at 16:20
  • Could be "any of the above". First step: Turn off the breaker(s) for anything related - i.e., outlet & fan (obviously) and any lights that have switches in the same location. Then remove the cover plates and pull out (but do not disconnect anything) the outlet, the fan switch, and any other switches in the same box. Take pictures of all the wires you can see and add them to your question. Jan 27, 2019 at 16:21
  • What kind of switch is servicing the fan? Where is the fan located in the bathroom?
    – noybman
    Jan 27, 2019 at 17:52
  • Yes, hitting TEST on the GFCI outlet causes the fan to stop working. The fan is controlled by a switch in the same switch box as the light switch, though the lights are not affected by the GFCI outlet (the light is on a separate circuit with other lights). The fan is in the ceiling
    – Chris.B
    Jan 28, 2019 at 3:00

3 Answers 3


I have seen this with fans several times, the fan motors create inductive kickback when spinning down this is enough to trip some GFCI's, if there are no outlets on this switched side of the circuit move the wiring to the line side of the GFCI. I have even seen motor circuits on a separate branch circuit cause this problem but they were usually larger 3+ hp motors in those cases both that I saw changing brands of GFCI did the trick, (the GFCI tested good I have a variable load to test so I know they were good just a bit more sensitive).

  • Thanks, the kickback makes sense. So there isn't likely anything wrong with the outlet, it's just an annoyance?
    – Chris.B
    Jan 28, 2019 at 3:06
  • 1
    Yes this is one of the problems that can happen with both AFCI's and GFCI's. If the light and fan are the only things connected (no vanity outlets) I would move the fan and light to the line side of the GFCI and this may stop the issue. If the light is on a vanity with an outlet that outlet would need to be changed to a GFCI but this usually fixes the problem of nuance tripping.
    – Ed Beal
    Jan 28, 2019 at 14:11
  • OK great, that sounds simple enough. The lights are actually on a different circuit with the other lights on the floor, so it's only the fan I need to worry about.
    – Chris.B
    Jan 28, 2019 at 15:38
  • My fan isn't over the shower but it's close, so I wonder if the electrical code requires it to be on the GFCI?
    – Chris.B
    Jan 28, 2019 at 15:43
  • I think my electrical code (Ontario) actually requires the bathroom fan to be downline from the GFCI. Can you recommend a model/brand of GFCI that won't trip as easily from an inductive kick?
    – Chris.B
    Jan 28, 2019 at 15:49

It's not the outlet. It's a ground fault.

Here's what is happening. When you have a big magnetic coil, and power is flowing through it, it makes a magnetic field. When you suddenly interrupt that power, the magnetic field collapses, and this has a small force behind it. It is the nature of coils to resist changes in current flow. It does this by turning that small force into voltage, which increases until the desired current flows, or the force just runs out of energy. This is called an "inductive kick" and everything in the circuit must be insulated for it.

When the insulation fails, the high voltage will ignite an arc that the lower voltage will sustain. It can also leap across switches! In DC power over about 30 volts, this arc is Very Nasty Business because the arc will never stop.

However, with AC power, AC switches polarity 100-120 times a second (8-10 milliseconds). It is a sinewave, and part of the time it is near zero volts. The arc will self-extinguish at the next zero crossing. But in the meantime, the arc will flow enough current (potentially a lot of current) to trip the GFCI.

However... If you happen throw the switch at the right instant, you are near a zero-crossing for the voltage or magnetic field, so there's no magnetic field and no inductive kick. So no voltage spike.

So, no leakage and no GFCI trip.

Insulation is failing. This will get worse.

I would start by changing the switch because they are cheap. But prepare for a fan replacement, or at least take it down and clean it. Damp dust makes a great leakage path.

Note that switches have ratings, and their rating for a motor (inductive) load is much lower than a plain load. They may also have a similar "ballast" rating, because HID and older fluorescent lights have ballasts with a big inductive winding. Their tungsten rating has nothing to do with this.

  • 1
    I wonder if Meggering the fan would be useful for proving out the fault in this situation? Jan 28, 2019 at 23:51

I would try just replacing the GFCI switch first. Easy inexpensive way to see if that is all or something more complex. Do be sure to work safely and turn off breaker before changing out.

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.