light fixture trouble shooting

Status
Not open for further replies.
C

copper123

Senior Member
I installed two new electronic ballast fluorescent fixtures in a older home the other day.(One feed, with a chase nipple between, bonded together) They were hung in a room that used to have a spa in it. I powered them up from a existing light fixture rough in. The 120 volt branch circuit that feeds the lights and receps are on a 20 amp GFCI breaker. After I installed the new fixtures, I could not get the GfCI breaker to hold. I started doing some troubleshooting, and found that every time I would take the ground wire off the fixtures, so they were ungrounded, the GFCI would hold. I did some checking with my meter, and found that I had 120 volts from my fixture frame to my branch ckt ground when the lights were energized and lit. My thought on this was that the 120 volts was purely magnetic induction from the ballast of the fixture inducing voltage on the metal parts.
The neutrals and grounds in the panel are all bonded together also. My thought is that I am creating some type of ckt with the magnetic induction. When I don?t bond the fixture I don?t have a secondary. Thus, my hot to neutral ratio back at the GFCI breaker sees a balanced current between, hence no trip. When I ground the fixture I have current leakage somehow and I don?t have a balance between my hot and neutral. Even though its on a GFCI I can?t leave the fixtures ungrounded, and am reluctant to put standard breaker in its place until I get to the bottom of it.
Any other thoughts would be great. Thanks
 
Re: light fixture trouble shooting

"The neutrals and grounds in the panel are all bonded together"
If this is a gfci breaker then there is a problem.Your neutral from the light outlet goes to the neutral lug on the gfci breaker not the neutral bar.
 
Re: light fixture trouble shooting

sorry Jim, I didnt clarify that.
The nuetral of the GFCI branch ckt is landed on the breaker. All others are bonded together.
 
Re: light fixture trouble shooting

Let's say that your fixture has a connection somewhere between the hot leg and the metal parts of the fixture. This could be in the ballast itself or somewhere in the wiring. When you remove the ground the energized metal parts are now at 120VAC to ground but since you have no ground connection the breaker holds and everything works. Now you connect the ground to the metal parts and it trips. This is not because of a GFCI issue but because of a connection between the hot and the ground. My guess is that if you were to wire the fixture correctly and connect it to a standard non-GFCI breaker than you would still have the same problem.
 
Re: light fixture trouble shooting

Did you check your connector in the light fixture to be sure it's not cut through the insulation?

If you disconnect the ballast hot, do you still get the 120 on the fixture frame (with the wire hot)?

Did you ohm from the hot of the ballast to the casing? How about the neutral to the casing?

Let us know what you determine.
 
Re: light fixture trouble shooting

HI guys,

Lets see.. I ohmed out my hot conductor to the fixture case and didnt get a reading, so I figured from there I didnt have a short to ground. Also I metered from my nuetral to the case and didnt get a reading. The fixture is floating on the rock ceiling, no bonding to systems at all. Thanks for the input.
 
Re: light fixture trouble shooting

I once found out the hard way if you remove a stanard 120v t12 ballast from a fixture while it is still hot, you are going to find your self trying your a** off to let go of the ballast.
 
Re: light fixture trouble shooting

I think the problem has to do with the rapid start ballast. I had a problem like this some time last year. Someone had installed two (2-lamp) fluorescent fixtures off the load side of a GFI receptacle. If only one fixture were on the circuit the GFI held. Once I tied the second one in it tripped the GFI receptacle. It seems these ballasts generate enough noise to trigger the internal GFI circuitry
 
Re: light fixture trouble shooting

Thanks goldstar,
This could be possible. Also, this GFCI breaker is pretty old, not sure how sensitive it could be. I have noticed over the last couple of years the nuisence tripping has gotten much better as the technology has changed.
 
Re: light fixture trouble shooting

If you followed any of the posts from awhile back about the theories of why (or if) a fixture must be grounded to work, the only plausable explanation to support it is that the neutral from the fixture ballast must have continuity to the sheet metal of the fixture. The only way this could happen is through the EG conductor back through the neutral/ground bond at the service where the two are connected together.

We were talking about rapid start magnetic ballasts but this theory may lend credence to your GFCI problem and electronic ballasts in that there still may be some current flow through the ground. The reason you wouldn't be able to measure any resistance from ballast case to the hot or ground leads is that there probably is capacitive coupling that would block DC but pass AC.

I would try a new GFCI breaker and see what happens. More likely it could be noise generated by the ballast that is tripping the GFCI. Either way, if you checked your work and found no N-G connection then I can't see a problem with replacing the GFCI breaker with a standard one.

-Hal
 
Re: light fixture trouble shooting

There is a MOV connected L-G {Line-to-Ground) on the Line side of the AC Input Filter (Electronic High Frequency Output switch mode Ballasts),
and have seen MOVs connected L-G at each "end" of the Autotransformer - where the "Line-In" connections are terminated (Linear Magnetic Ballast).

These TVSS devices would certainly trip a GFCI device of 6ma rating, since there would be as much as 25-100ma flow during initial starting / connecting to AC Line.
The amount of time for the Lamp/Ballast assemblidge to become stable may exceed 30 cycles (? second on a 60 Hz system).

I made some custom job site "Temp Lighting", which randomly trip the "Locally mounted" GFCI Receptacle during initial startup.

Reset the device and it holds steadily.

One reset is normal, but there have been occasions where it tripped back out as soon as I reset, but after the second reset, it held.
These dual resets occured when someone turned on, or off, a heavy load power tool - like a Circular Saw ("Skil-Saw").

The "Temp Lights" are made up of two separate 4' 2 Lamp Fluorescent strips (mounted back to back), attached to a custom made stand - made of ?" EMT.
Each has a 50' 12/3 cord whip, which lands at a single gang deep bellbox, mounted to the side framing of the stand.

They look (and work) really cool!

A 20 Amp duplex GFCI is mounted in the box, and the fixtures are connected to the LOAD side of the GFCI Receptacle.
Some models have a cord whip, which gets plugged into the GFCI Receptacle, as opposed to being hardwired through the LOAD side.

Fixtures use F40 T12 Lamps, driven by Magnetic Ballasts (previous versions), or F32 T8 Lamps, driven by Electronic Ballasts (current versions).

Trip issues are a bit more frequent with the Cold Cathode, Instant Start, Electronic Ballast driven T8 Lamps, as opposed to the Hot Cathode, Rapid Start, Linear Magnetic Ballast driven T12 Lamps.

Fully traced circuitry to verify no Ground Faults existed on either of the Line Conductors (Grounded "Noodle" Conductor and Ungrounded Conductor) at any location on the temp light stands - and to verify Ballast output circuitry was not Ground Faulted either. Found no Ground Fault issues.

Just some FYI and 2?

Scott
 
Re: light fixture trouble shooting

Bumping the topic, as to find out just whaddaheck happened so far, or if the original posted unfriendlyness, between GFCI and Ballasted Lamps, has been resolved... and how?! :)

Like my descriptions? :eek:

Scott
 
Re: light fixture trouble shooting

What you have is capacitance coupling between the bulbs and the fixture. In order for a fluorescent bulb to start under all circumstances such as below 70 degrees Faherheit, there must be a grounded antenna near the bulb to help the initial ionization wavefront propogate from one end of the bulb to the other. In fact, most high pressure sodium fixtures will flat refuse to ignite if not grounded.

The reason why an ungrounded fluorescent fixture will work is that any big enough chunk of metal will act as a ground for alternation current. That is how antennas on airplanes work particularly at the shortwave radio frequencies that are used when crossing oceans.

Amp checks ( using a true RMS ammeter ) on 120 volt electromagnetic rapid start ballasts like you have show that about 10% of the hot lead current flows back through the equipment ground. Most of this is high frequency ( harmonic ) garbage and your GFCI should be programmed to ignore currents that are not line frequency.

Try a different brand of GFCI.

Oh, I hope that those temp lights for a jobsite are enclosed in some way. Breaking the bulbs can get to be quite expensive.

[ April 16, 2005, 03:14 AM: Message edited by: mc5w ]
 
Status
Not open for further replies.
Top