DIY Chatroom Home Improvement Forum

DIY Chatroom Home Improvement Forum (http://www.diychatroom.com/)
-   Electrical (http://www.diychatroom.com/f18/)
-   -   New GFCI Breaker Tripping w/ 3-way Flourescents (http://www.diychatroom.com/f18/new-gfci-breaker-tripping-w-3-way-flourescents-6721/)

BigJimmy 02-25-2007 06:38 PM

New GFCI Breaker Tripping w/ 3-way Flourescents
 
I have been in the process of re-wiring my entire (very old) house for several months now. One of the first projects was to install lighting in the basement (prior to that, you had to screw in incandescents as you walked from front to rear, burning fingers on your way out). A finished room in the middle of my basement creates a natural corridor from the stairs back to the laundry so I installed several 1'x4' T8 Flourescent fixtures with HPF electronic ballasts and included 3-ways at the staircase and the rear door to the basement (the difference was dramatic, incidentally).

When I completed the initial install, I did not have any GFCI breakers so I simply powered the entire circuit from a standard breaker. Well, I picked up a few GFCI's the other day and set out to retrofit the lighting branch breaker today. When I got everything installed, I turned on the breaker, the lights came on and then the GFCI tripped (15A, 1P, same as previous breaker). Tried resetting again and the same thing happened. Recalling that I nicked one of the travellers during installation (all #12 THHN in EMT), I flipped the nearby 3-way figuring if it held then that would be the cause. Sure enough, when I switched the lights to the other traveller, the breaker reset and held (before I go further, the nicked wire is in a 4-11 and I made sure that it was clear of the box until I'd have time to repair it). So, I tried flipping the 3-way back a forth just to verify that the problem was repeatable (it was).

Well, after monkeying around for a good hour running a replacement traveller and removing the old, I reset the GFCI and it tripped again! However, now it was tripping on the half of the circuit that proved good before the replacement! So, I tried a few more permutations of the 3-ways and then all of a sudden, it held. Disgusted, I removed the breaker and performed a continuity test of all of my wires to each other/ground. All good. Still baffled, albeit visibly PO'd, I remembered that the first fixture, closest to the stairs, is often slow to start (the rest come to life immediately). For the heck of it, I disconnected the ballast in this unit only and powered everything back up. All fine, both sides of the 3-way.

Now, I reconnected the fixture and tried again, this time with my wife watching the flxture to note any delay in its starting. It seems that when it starts immediately, the GFCI is happy but if there is any delay, the GFCI trips. At this point, I have the following questions:

1. Do GFCI breakers and Flourescent fixtures not play well together?
2. I heard somewhere that GFCI's can misbehave if placed on an excessively long circuit. Is this one way or round trip length.
3. Any other ideas would be welcome!

Thanks,
Jim

Stubbie 02-25-2007 08:14 PM

Gfci's and highly inductive loads like those of fluorescent lights do not go well together. There is no requirement to gfci lighting loads. There is no requirement to gfci finished basement outlets. Bedroom outlets require afci. Put the lights back on a regular inverse time breaker. You wont be left in the dark at least.


stubbie

BigJimmy 03-10-2007 09:56 AM

As a follow up...

1. Yes, I realize that a GFCI is not required for basement lighting however this circuit also supplies an outdoor switched light. Although the exterior build uses weatherproof boxes, conduit, fittings, I wanted to protect against moisture-related ground faults.

2. The problem is not the inductive nature of the loads. Although the current will indeed lag the voltage, the GFCI only cares about the residual difference in current magnitude between the hot/neutral. Any phase lag in current is absolute and will continue to cancel in the torroid responsible for tripping the breaker.

3. Further investigation revealed a very minor nick in a wire feeding some bathroom lighting in the basement which was also part of the circuit. Once I corrected the problem, no more tripping.

4. Excessive wiring length in a GFCI-protected circuit can cause false tripping due to capacitive coupling between hot and neutral wires which only becomes non-trivial in "long" circuits. Therein, the wires behave like plates of a capacitor and capactive leakage current can in fact rise to the point where the GFCI thinks that there is a ground fault (Ihot<>Ineut).

5. Although the mystery is solved, I did opt to change the design so that a standard (non-GFCI) breaker is used in the panel as the OCP device and the outdoor light circuit is fed from a GFCI receptacle placed outdoors. I prefer to have receptacles and lights on separate circuits however in this instance, a GF on the lighting circuit would only trip the receptacle and not the entire basement lighting circuit.

Regards,
Jim

Stubbie 03-10-2007 06:55 PM

Its been a known fact that it is indeed the inductive nature of fluorescents that can, at times, false trip gfci's much like motor loads. You said your wiring was good so the fluorescents was a likely candidate. You will never see fluorescents spec'd in commercial work for gfci. What I dont understand is that you have come back with a very knowledgable reply to your own original post. You also answer all the questions of your original post as if you knew the answers to begin with...so whats the deal??

Stubbie

BigJimmy 03-11-2007 11:37 AM

Quote:

Originally Posted by Stubbie (Post 36513)
Its been a known fact that it is indeed the inductive nature of fluorescents that can, at times, false trip gfci's much like motor loads. You said your wiring was good so the fluorescents was a likely candidate. You will never see fluorescents spec'd in commercial work for gfci. What I dont understand is that you have come back with a very knowledgable reply to your own original post. You also answer all the questions of your original post as if you knew the answers to begin with...so whats the deal??

Stubbie

Stub-

In my follow up post, I did note that after some further investigation, I found a place where I had inadvertently ragged on some conductor insulation that was apparently creating a minor GF.

In response to your other question, although there are a lot of very intelligent people out there who will gladly share their knowledge/experience, I've seen a lot of bogus information being shared as well. In fact, I used to subscribe to the alt.engineering.electrical ng which was supposedly home to electrical professionals, including degreed engineers, some with P.E.'s. After one particular engineer argued with me over and over that no current flows in a neutral conductor, I stopped visiting! At any rate, I think I posted the original on a Sunday and didn't have an opportunity to check back until almost a week later. Since I am curious by nature (I too am an engineer), I did quite a bit of research, consulting with collegues/electricians and in general, thinking and drawing circuits (trying to remember my vector math about killed me, BTW!). This is where I gleaned my further knowledge and I thought that following up on my original post with my gained experience could perhaps be of benefit to other members. After all, based on the fact that there are 80 views and you and I are the only ones dicussing it, the topic does not seem as cut and dry as some others.

So, the original questions were, at that time, legit questions. On the other hand, it is not my nature to simply take advice from complete strangers and assume that it is correct or appropriate without conducting my own research. Don't get me wrong: I am not saying that I have not learned a lot of great things from people via the internet and this site in particular; I have. But sometimes, even though the responses may be technically correct, learning the reasoning behind them (i.e. codes, etc) helps to explain why the answers are sound.

On last thing too. Most people engaged in electrical engineering, design, installation, etc., use the NEC as well as other reference texts (IEEE color books are fantastic). Often times I have heard "the NEC doesn't require you to do that." You mention that "You will never see fluorescents spec'd in commercial work for gfci" and I agree. Then again, the NEC doesn't require GFCI protected receptacles in a four-year old's bedroom but you'll notice that my guy's got 'em. Why? Cheap insurance (I remember almost setting fire to my room when I was attacking a set of line-voltage decorative lighting with screw driver as a kid).

I hope this answers your question. I wasn't trying to look like some smart ass, know-it-all. I just wanted to share what I learned in hopes that it could help someone else. I look forward to reading and contributing to further discussions in this chatroom.

Take care,
Jim :thumbsup:

Stubbie 03-11-2007 06:30 PM

No problem Jim....that sounds reasonable to me. I didnt mean to offend I was just a little confused. I have been involved in other posts where the whole thing was a setup to allow someone to show the world how smart they are at other peoples expense and time.


Now to add to what you have been discussing. The most common distance in my experience that is taught in the trade for gfci parameters is 250 feet. This seems to be the most agreed upon distance that capacitive coupling begins to show its effects for false tripping.

The gfci's manufactured by most major suppliers today have circuitry for the most part that prevents false tripping during operation of motors or fluorescents. So it isnt a very good probability that a gfci would be at fault for these situations in todays world. Though I like to read about these physics I am not going to pretend to be well versed in understanding the why of it. I've taxed my brain power enough just trying to understand the avoid using gfci areas. At any rate no harm done and I understand now where your coming from.

Stubbie

Yoyizit 12-18-2008 01:34 PM

If the flourescent ballast has more than 0.9F capacitance to ground from either input lead it will leak 4 mA to ground and trip a GFI.
You can measure this capacitance with a C-meter, or a DVM on the volts scale.
900' of Romex will also give you 0.9F. And there may be other stuff on the same circuit that is using up some of your 4 mA "budget".

AllanJ 12-29-2008 09:40 AM

OT: Wouldn't the capacitive coupling between conductors given a long length of run, together with capacitive coupling depending on the physical arrangement of the ballast within the (grounded) housing of a fluorescent light fixture, together with the inductance (L) of the ballast all create "tuned LC circuits" that may result in resonant behavior (seen by the GFCI as mini surges that are differences in hot versus neutral current as the lights are turned on or off) that might trip the GFCI?

In your case, you described finding at least two nicks in insulation. Did both of them involve metal to metal contact between ground and circuit conductors?

>>> ... high from behind ...
If the voltage is a little high or a little low, that is not the reason for tripping of a GFCI.

Gigs 12-29-2008 02:29 PM

Quote:

Originally Posted by AllanJ (Post 203779)
together with the inductance (L) of the ballast all create "tuned LC circuits" that may result in resonant behavior

Yes this is called self-resonance... but the resonant frequency will be very high considering the small values involved, and so usually easy to ignore.


Quote:

>>> ... high from behind ...
If the voltage is a little high or a little low, that is not the reason for tripping of a GFCI.
I am 99% sure that message was from a spam bot. They are making smart spam bots now that post babble that almost sounds like a valid message.


All times are GMT -5. The time now is 05:51 PM.


Copyright 2003-2014 Escalate Media LP. All Rights Reserved