GFCI tripping when opening a switch

[Natfuelbilll]

Unbalanced load? That would be normal neutral (or phase) current flowing on the ground.
Could it be if the switcher has his foot in a metal pan, and is touching the metal frame of the fixture - then switches the light off would be a parallel path for the subject ground current - that the installation of a gfci makes safety sense?

[ELA]

Natfuelbilll,
How often does the GFCI trip? Is it totally intermittent or can you make it fairly repeatable?

If you can make it repeatable and want to try an experiment you might try the following:

Install a long length of NM temporarily (partial box you have around > 50ft) between the load side of the GFCI and and the feed side of the light switch.
Then see if the occurrence rate goes down?

As has been mentioned earlier one possible culprit are the transients generated when the fan switch opens the ballast load. If you add some impedance between the source of the transients (the fan switch) and the GFCI then the RFI generated when the switch opens will be attenuated some.
{ The concept being -why a panel mounted GFCI breaker might not trip but a local GFCI outlet that is so close to the transient generation might.}

A better test would be to install an EMI filter between the two if you have one.

Just a suggestion if you are curious enough to look into it further and to satisfy your request for a technical view of the problem.

[Natfuelbilll]

Sounds easy enuf. I'll give it a try.

[gndrod]

Originally Posted by Natfuelbilll

Sounds easy enuf. I'll give it a try.

Thanks for the update. Another possible test is to swap out the fluorescent for an incandescent load and see what happens. Also, is there another simultaneous active inductive load downstream from the GFCI? rbj

[gar]

091108-2026 EST

Natfuelbilll:

In an earlier post I ask if you could provide some other numeric description for your Leviton GFCI. If it is not a recent unit, then could you try a Leviton 7899?

Another question. Do you have an EGC conductor going to the fluorescent? If so, then open the EGC and see if the tripping still occurs. If the tripping ceases with out the EGC connection and the fixture is essentially insulated from ground, then there is probably internal breakdown or a conductive path on turn off in the fluorescent fixture.

I just tried a cheap 4' electronic ballast unit with an EGC and I could not trip a 7899.

Is the ballast magnetic or electronic?

.

[jmellc]

Different components, similar problem. I once had GFI's tripping in a kitchen when LV track lights were tuned on. Not every time, maybe every 3rd or so. I looked into filters, etc and talked to someone at Hubbell. Truthfully told her I didn't remember if they were Hubbell GFI's or some other brand. She'd heard of a similar case and advised to be sure the GFI's were on different phase from track lights. Sure enough they were same phase. I moved GfI's to A and track lights to B. That took care of it. Somehow, the LV tf was affecting the GFI's. GFI receptacles are much more subject to nuisance tripping of any kind than GFI breakers.

[gar]

100117-1059 EST

I have tried to cause transient tripping of a Leviton 7899. I have not been able to create a false trip. This device uses the National chip, there is substantial low pass filtering on the supply to the SCR anode, supply to National, and input to National.

Using a different phase for your low voltage lamps and the GFCI provides substantial transient isolation.

I would like to know if you used a Leviton 7899 on the same phase and branch circuit and within a few feet between the GFCI and the lights if there would be any false tripping.

.

[Tommy Hobbs]

(1) Where is the GFCI located
(2) Where is the Fixture located
(3) GFCI Receptacle, I assume, right
(4) The Fixture / Switch must be connected to the Load side of the GFCI if
Receptacle, can it be removed from the Load side & Relocated to the
Supply side, just wondering

[Natfuelbilll]

Symptom stopped after I installed a new fluorescent lamp in the fixture.