Shouldn't Light Fixtures Have Bare Grounding Wire Connected?

**DavidSeon** · 12-11-2012, 10:30 AM

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

**leejosepho** · 12-11-2012, 02:10 PM

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.

**nukeman** · 12-11-2012, 02:28 PM

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.

**jwelectric** · 12-11-2012, 04:56 PM

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.

**jadnashua** · 12-11-2012, 05:10 PM

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.

**cacher_chick** · 12-11-2012, 07:10 PM

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.

**kreemoweet** · 12-11-2012, 09:15 PM

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".

**DavidSeon** · 12-12-2012, 01:01 AM

Originally Posted by nukeman

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.

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.

Originally Posted by jwelectric

My question would be what year was the house wired?

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

Originally Posted by kreemoweet

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".

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.

**jadnashua** · 12-12-2012, 07:00 AM

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.

**leejosepho** · 12-12-2012, 08:41 AM

Originally Posted by DavidSeon

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.

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.

**nukeman** · 12-12-2012, 10:29 AM

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.

**big2bird** · 12-12-2012, 03:10 PM

Originally Posted by DavidSeon

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.

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.

**big2bird** · 12-12-2012, 03:12 PM

Originally Posted by jadnashua

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.

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.

**DonL** · 12-13-2012, 04:02 AM

Originally Posted by DavidSeon

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.

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.

**nukeman** · 12-13-2012, 05:06 AM

^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.