Question about GFCI's

[bthielen]

I rewired my garage. For lighting I provided receptacle outlets on the ceiling because my plan was to use 4? 120v plug-in fluorescent shop light fixtures that I hung from the ceiling. I used a GFCI receptacle to protect the circuit from ground faults and when I powered up the lights the GFCI receptacle would trip. If I unplugged the lights from the circuit everything was fine so I know I don?t have a ground fault in my circuit. I had been using these fixtures in that garage for years without issue but the old circuit was not GFCI protected. Can these fluorescent fixtures cause nuisance trips of the GFCI when they are firing up?

Thanks,

Bob

 

[480sparky]

If the lights are creating a ground fault, then they are not producing any 'nuisance' tripping.......... The GFCI's are doing the job they are intended to do.

 

[jumper]

I agree with the dude getting shocked. GFCIs are pretty much rock stable these days. Your problem lies else where, unless you miswired them accidently.

 

[bthielen]

I understand the functionality, purpose, and reliability of GFCI's. There was nothing to wire wrong. These are purchased fixtures with cord and plug. Both of them caused the GFCI to trip. This raised a question. First, what are the odds that two different fixtures have the same problem? Both fixtures function normally when plugged into non-GFCI protected circuits. I inspected the interior wiring, ballasts, and starters and found nothing out of the ordinary such as scuffed wires or indications of excess heat damage.

I'm not a field electrician and so I was wondering if it is possible that startup of these types of lights have been known to cause nuisance trips because somehow there is enough current difference between the grounded and ungrounded conductors to trigger the GFCI circuitry.

Thanks,

Bob

 

[jumper]

I have connected a very large array of equipment including flourescent and halogen lights to GFCI and they all worked unless the equipment was faulty. GFCIs are used extensively in construction without a problem.

They used to be cranky, but are pretty much rock solid now. Did you try applying a different load on the receptacles?

 

[jumper]

If you are thinking harmonics, forget it. Not in a 120/240 resi system.

 

[charlie b]

I suggest bringing the lights into the kitchen or bathroom, and plugging them in to a receptacle that is protected by a GFCI. This would confirm whether there is a problem with the lights. The lights could have a small leakage current, enough to trip a GFCI, but not enough to cause a standard 20 amp breaker to trip. The fact that they work on a non-GFCI circuit does not mean that they are working properly, or that they are safe.

 

[K8MHZ]

inspected the interior wiring, ballasts, and starters

I had been using these fixtures in that garage for years

Starters? On a 4 foot fixture? I think it's time for new fixtures. I'll bet there is some continuity between the neutral and the ground on both fixtures.

 

[cadpoint]

Since their ten hears old, I'll bet there is no ground at all in the fixture much less a three prong plug...

:jawdrop:

 

[bthielen]

These are those 4' dual lamp fluorescent fixtures you can buy at the local hardware store. They do have a three-prong plug with the grounding wire bonded to the fixture. At any rate, I decided not to use them and installed a different fixture instead. Just thought it was pretty coincidental that both fixtures caused the same problem and my limited knowledge of ballasts and such I thought maybe there's something about that. Thanks for the replies.

 

[handy10]

Since their ten hears old, I'll bet there is no ground at all in the fixture much less a three prong plug...

:jawdrop:

I don't believe that the GFI could trip if the lights are not grounded. There would be no place for the current to leak except into the neutral.

 

[cadpoint]

I don't believe that the GFI could trip if the lights are not grounded. There would be no place for the current to leak except into the neutral.

A GFCI internals do not sense anything about the ground wire, IE it's not needed in determining or cause a GFCI to function.
Here this is an action or java link that turns off!

My statement was based on age of the lamp fixture where a bond was not a requirement and at that time a change in the Code required a bonding screw within 6" of the ballast, and a ground / bonding wire were required to make a three wire assembly.

Age of the ballast is the biggest culpert, weather, un-heated space, illumanition under incorrect tempertures, take your pick.

 

[bthielen]

It has been my understanding that a GFCI worked by measuring the difference in current between the grounded and ungrounded conductors. If there is no alternate grounding path to reduce the current on the grounded conductor how can a GFCI measure these differences? Maybe I misunderstand how GFCI circuits work.

 

[handy10]

It has been my understanding that a GFCI worked by measuring the difference in current between the grounded and ungrounded conductors. If there is no alternate grounding path to reduce the current on the grounded conductor how can a GFCI measure these differences? Maybe I misunderstand how GFCI circuits work.

You have said what I attempted in my post. I don't believe that the GFI would trip unless there is a path to ground that is different from the neutral. It would
have been interesting to plug the light into a two prong cheater and see if the GFI tripped. Of course, I am assuming that the light is not hung on chains
from a water pipe.

 

[dkidd]

A GFCI internals do not sense anything about the ground wire, IE it's not needed in determining or cause a GFCI to function.
this is an action or java link that turns off!

Not quite.

A GFCI also detects neutral to ground connections.

View attachment 5926

The link posted is only a ground fault detector circuit.

 

[cadpoint]

Not quite.

A GFCI also detects neutral to ground connections.

View attachment 5926

The link posted is only a ground fault detector circuit.

I'll stick with my answer a grounded wire is not needed for a GFCI to work, a grounded wire if present helps to clear or gives a
fault a path but in the end it is not required for a GFCI to operate, so I'll raise you one. Note the author!

Contractor Talk #25 post

 

[handy10]

I'll stick with my answer a grounded wire is not needed for a GFCI to work, a grounded wire if present helps to clear or gives a fault a fault a path but in the end it is not required for a GFCI to operate, so I'll raise you one. Note the author!

Contractor Talk #25 post

If cadpoint is saying that a GFI does not need a ground wire to operate, then we are in agreement. If he is saying that the device (fluorescent lamp) connected to
the GFI receptacle does not need a return path for leakage current (leakage current is required for the GFI to trip), then we do not agree.

 

[dkidd]

I'll stick with my answer a grounded wire is not needed for a GFCI to work, a grounded wire if present helps to clear or gives a
fault a path but in the end it is not required for a GFCI to operate, so I'll raise you one. Note the author!

Contractor Talk #25 post

I never said that a neutral was needed for it to work. I was refuting the statement that only hot to ground would trip it.

As to your link, I don't see your point.

 

[kwired]

If you are thinking harmonics, forget it. Not in a 120/240 resi system.

The harmonics are still there they just don't add up in the neutral to the same effect as they do on a wye system.
They also are probably not a problem for the GFCI on either system.

It has been my understanding that a GFCI worked by measuring the difference in current between the grounded and ungrounded conductors. If there is no alternate grounding path to reduce the current on the grounded conductor how can a GFCI measure these differences? Maybe I misunderstand how GFCI circuits work.

They do monitor grounded and ungrounded for equal amount of current (within 4-6mA otherwise they trip).

You likely have purchased 'cheap' fixtures that are letting some current return on the equipment ground conductor and causing the unbalance needed to trip the GFCI. Try megging the fixture from either lead to equipment ground, that will likely tell you quickly that is the problem. These fixtures often are not worth the metal they are made out of- if that is possible.

 

[charlie b]

I am a little unclear on whether the previous answers are in agreement or in conflict with each other. So without taking time to straighten it all out in my mind (that mind being insufficiently infused with coffee this morning), I will say this. Feel free to substitute ?garage walls? for ?human body,? in the following paragraph. But I will say that this discussion of ?human body? is the reason GFCIs exist in the first place.

Current flowing in the following ?leakage path? will cause a GFCI device to trip.

• From the ungrounded (?hot?) conductor, via damaged insulation or some other failure within the light fixture, to the external metal case of the fixture.
• From the case to the hand of the person touching the fixture.
• Through that person?s body to the ladder he is standing on, in order to reach the light.
• Through the ladder to the floor.
• Through the concrete garage floor to the dirt below.
• Through planet Earth to the location of the ground rod associated with the building?s main service panel.
• Into the ground rod, and up the Grounding Electrode Conductor to the main panel?s neutral bar. At this point, we can say that the current has returned to its source.

Please note that this path does not involve the grounded (neutral) wire, nor does it involve the Equipment Grounding Conductor (EGC). Please note also that this path can carry enough current to electrocute the person on the ladder, but not enough current to trip the circuit breaker. As I said earlier, this scenario is the reason GFCIs are required in certain areas.