Shouldn't Light Fixtures Have Bare Grounding Wire Connected?

Discussion in 'Electrical Forum discussion & Blog' started by DavidSeon, Dec 11, 2012.

  1. DavidSeon

    DavidSeon New Member

    Messages:
    47
    Location:
    MS Gulf Coast
    Doing bathroom remodel I notice incoming bare grounding wires of standard 3-wire cable (12ga black, white, bare) is not connected to green screws or green grounding wires supplied with the fixtures, just hanging loose or pushed back in the box. Doing a quick check, I find the same lack of grounding connection at several other fixtures (ceiling light, outdoor porch light, etc), although I haven't checked all of them. In one case, the incoming bare wire is cut back to the insulation and not even available for connection, apparently not used on purpose.

    This seems like a no-brainer safety feature to me, and requires such a trivial amount of extra effort, I'm beginning to wonder if I'm missing something. Is there ever a reason NOT to connect the bare grounding wires?

    Thanks
  2. leejosepho

    leejosepho DIY scratch-pad engineer

    To: Old Retired Computer Programmer -- Obsolete and loving it
    From: an obsolete hobbyist posting in Windows 98 (but I had to use Firefox here)

    I would suspect someone was being lazy, so no, you are not missing anything other than a bit of protection in the event of a lightning strike and a GFCI in the bathroom since I do not think one will work at all without a ground wire.
    Last edited: Dec 11, 2012
  3. nukeman

    nukeman Nuclear Engineer

    Messages:
    711
    Location:
    VA
    Is it a metal junction box? I assume it is wired with NM cable (no conduit that could be acting to complete the bonding), correct?

    The main purpose of that bare wire is to bond the junction box/fixture. If it was bonded and the box became energized (wire nut came loose and the hot touched the box, or some other fault), that wire would carry current back to the panel (essentially a short between the hot and neutral) and would cause that breaker to trip. Without that bonding, the box/fixture could become energized and could potentially shock someone if they touched it.

    As for why they didn't do it, jw or another electrician may be able to answer. Without metallic conduit or another means of bonding, it sounds like they took a shortcut, but perhaps we need more details to tell for sure.
  4. jwelectric

    jwelectric Electrical Contractor/Instructor

    Messages:
    2,540
    Location:
    North Carolina
    My question would be what year was the house wired?

    My comment would be that the grounding conductor needs to be connected to any grounding conductor in the fixture or the screw on the fixture.

    As to the GFCI it will work with or without the grounding conductor.
  5. jadnashua

    jadnashua Retired Defense Industry Engineer xxx

    Messages:
    22,152
    Location:
    New England
    FWIW, one accepted use of a GFCI is to swap a 2-prong to a 3-prong (i.e., grounded) receptacle. It makes it more convenient if you have a 3-prong plug, but does not actually provide ground. It is legal if downstream receptacles are fed properly from the load side and it's marked as not having an equipment ground. I'm in the process of doing this at my mother's house as last time I was here, I found some of the 50+ year old receptacles have worn out, and we're not ready to rewire the whole house with grounds, so I'm adding GFCI's and grounded outlets.
  6. cacher_chick

    cacher_chick Test, Don't Guess!

    Messages:
    3,249
    Location:
    Land of Cheese
    Every metal junction/device box should have a threaded ground screw or ground clip installed in the box for connection of the grounding conductor. If the fixture has no ground wire or screw, it was probably made before the mid-60's when it was not a requirement.
  7. kreemoweet

    kreemoweet New Member

    Messages:
    371
    Location:
    Seattle. WA
    The only legitimate reason I can think of for not properly connecting the Equipment Grounding Conducter in a cable would be if the EGC
    was not actually "grounded".
  8. DavidSeon

    DavidSeon New Member

    Messages:
    47
    Location:
    MS Gulf Coast
    No boxes for the vanity fixtures, they just punched a hole in the drywall for the cable. I picked up a couple of plastic ceiling boxes with hanger bars for making a neater installation of the new fixtures. My concern was, as you described, if the exposed metal of the fixture became energized, the breaker wouldn't trip if the grounding wire wasn't connected.

    Should have mentioned the house was built in 1979 with 12/2 NM-B with ground and 5-15R three-wire grounding receptacles.

    That's the kind of off-the-wall thing I started worrying about when I saw how consistent they were with not connecting the grounding wires, like maybe they knew something I didn't. Although, the grounding wires were connected to all the outlets on the same circuits, if that means anything.

    Is there a way to verify that the grounding path is adequate with my Fluke meter, resistance maybe? Have a lot of experience troubleshooting, voltage and signal tracing on live military electronic equipment, just not building wiring. I'm comfortable that I can keep myself safe, or at least know when I'm not.
  9. jadnashua

    jadnashua Retired Defense Industry Engineer xxx

    Messages:
    22,152
    Location:
    New England
    Assuming you can do it safely, open the panel at the breakers and see if they connected the grounds to the ground bus bar. If they did, I'd certainly want to maintain that path for any device that has that provision.
  10. leejosepho

    leejosepho DIY scratch-pad engineer

    Only if the neutral is broken. Like Jim had mentioned in reference to GFCI in a two-wire circuit, the ground can be optional and things will still work as they should.
  11. nukeman

    nukeman Nuclear Engineer

    Messages:
    711
    Location:
    VA
    That is not true. The neutral would not have to be broken for the metal to become energized (if not bonded) or for whether the breaker would trip. It takes 0 current to energize the housing/junction box (would have a potential of say 120v, but no current (think of it as something with infinite resistance)). The breaker has no idea that the housing/box is energized. Bonding the box/fixture would create a dead short back to the panel where the neutrals are connected to the ground bar. Now you are putting that 120v potenial across something of little resistance (just the resistance of the wiring out to the fixture and back). This will cause a large current to flow and that breaker will trip.

    A GFCI is another beast. They work by comparing the current on the hot and the neutral. If they are enough out of balance, the GFCI determines that there is current leaking somewhere (like through someone's body) and will trip. It will serve this function with or without the grounding conductor attached.
  12. big2bird

    big2bird IBEW Electrician

    Messages:
    141
    Location:
    Anaheim, Ca.
    No boxes? No grounds connected? Hack job. Every fixture gets a box. Everygound is boded to the box, the mounting strap, and the fixture pig tail.
  13. big2bird

    big2bird IBEW Electrician

    Messages:
    141
    Location:
    Anaheim, Ca.
    The first receptacle in the circuit string after it leaves the panel is sufficient. The rest can be normal 3 prong receptacles fed thru the first GFCI in the string. The just must be marked as such.
  14. DonL

    DonL Jack of all trades Master of one

    Messages:
    4,332
    Location:
    Houston, TX

    You can measure the resistance from the unconnected wire to neutral , it should be near 0 ohms if it is connected at the other end.
  15. nukeman

    nukeman Nuclear Engineer

    Messages:
    711
    Location:
    VA
    ^But, remember to cut the main breaker before you do this. The neutral could have voltage on it from other devices on that circuit. Voltages do not play nice with your meter when checking for ohms. :)

    Turning off only the breaker that this circuit is on should be good enough for this test, but I would do the main, just in case. With the other issues you have found, you don't know if they tied neutrals from other circuits together (outside the panel) or did any other funny stuff.
  16. DonL

    DonL Jack of all trades Master of one

    Messages:
    4,332
    Location:
    Houston, TX
    Not a bad idea nukeman.

    I always check for voltage before trying to use the ohm meter on unknown wires.

    Turning off the power may be safer.

    I also use a light bulb for testing, but JW says that is a no-no.

    It is best to be safe when playing with electricity.
  17. leejosepho

    leejosepho DIY scratch-pad engineer

    I stand corrected, and I really do know better...but when I do make mistakes, they do tend to be big ones!
  18. DonL

    DonL Jack of all trades Master of one

    Messages:
    4,332
    Location:
    Houston, TX
    It is best to have a Ground even when using a GFI.

    The GFI will work without a ground, But if a ground is connected and you have a Appliance fault, The GFI and maybe the breaker will trip.

    That is better than a persons body being needed to make a ground fault trip.
  19. DavidSeon

    DavidSeon New Member

    Messages:
    47
    Location:
    MS Gulf Coast
    Just to be safe I cut the power at both the outdoor main panel and the indoor subpanel before removing the subpanel cover

    Inside Breaker Panel.jpg

    I can see the bare grounding conductors fastened to the ground bar on the right, and the white neutrals all connected to the neutral bar on the bottom. The grounds and neutrals are not connected together in the subpanel, but my understanding is they are bonded together at the outdoor main panel. I don't understand why the cable coming in at the bottom has a green conductor instead of a bare wire, but maybe just what they had on hand?

    This may be goofy but just to satisfy myself, I connected a 25 foot length of zipcord to the ground rod outdoors by the main panel, and brought the other end inside to the unconnected bathroom fixture wiring. I get 1.2 ohms resistance from the bare gounding conductor to the ground rod, and 1.3 ohms from the neutral to the rod; the zipcord reads 1 ohm by itself.

    I'm thinking the main wiring for the panels out to the transformer is correct but if I'm misinterpreting something, please let me know. Because of the issues I've seen, I'm planning to check all of the fixture wiring connections; apparently they used hacks for that.

    Thanks for all the help.
    Dave
  20. DonL

    DonL Jack of all trades Master of one

    Messages:
    4,332
    Location:
    Houston, TX
    Your measurement sounds about correct.

    There is nothing wrong with Green.

    The wire just cost more, in most cases.

    The Ground does not have to be Bare.
    Last edited: Dec 14, 2012
Similar Threads: Shouldn't Light
Forum Title Date
Electrical Forum discussion & Blog Kitchen lights dim when oven heats Yesterday at 4:53 PM
Electrical Forum discussion & Blog Question regarding wireless light switch Oct 15, 2014
Electrical Forum discussion & Blog Down Light's keep going out Oct 12, 2014
Electrical Forum discussion & Blog Light Bulbs Failing Simultaneously Sep 19, 2014
Electrical Forum discussion & Blog Adorne Under-Cabinet Lighting System Aug 20, 2014

Share This Page