Can’t believe this is the first time that I have run into this one.
Recently had a pre sale home inspector inspect a clients house and two gfi’s with no equipment ground would not trip using the “Ideal” type gfi tester. His OR mine. So I replaced the two gfis assuming they were defective. Still no go.
Talked to another electrician last nite and he said you need to tie the neutral to the equipment ground lug on the gfi thereby jacklegging it. Of course make sure you mark it NO EQUIPMENT GROUND.
Other thoughts?
Replies
"Talked to another electrician last nite and he said you need to tie the neutral to the equipment ground lug on the gfi thereby jacklegging it. Of course make sure you mark it NO EQUIPMENT GROUND."
What are you trying to DO? Make the external tester work or provide the safest system.
The internal "selftest" is just a resistor that goes from the hot side, AFTER the current transformer, to the neutral side, BEFORE the current transformer. It exactly dupplicates the fault condition that the GFCI is protect against.
Unlike smoke detector or CO alarms that only simulation the signal that the smoke or CO would cause and don't 100% test the sensor, the self test on a GFCI does FULL TESTING.
The only use for the plug in tester is to verify protected receptacels downstream from a GFCI.
If you really, really, want to prove that the gfci (or downstream recptacles) work then you need to do a simiulate REAL CONDITIONS. Get a grounding socket adaptor and on the little ground strap that is supose to go under the "grounded" coverscrew connected a test lead and run that to a real ground someplace.
If you connect the "ground" to the neutral connect and lose the neutral upstream what will happen to anything that is plugged into theat socket?
Bill there is nothing upstream
"Bill there is nothing upstream"
No CB or fuse pannel????
"What is the unneccessary hazard if it is labeled NO EQUIP GRD and there is no load downstream?"
Who is going to understand what that means and the risk envolved if you bootleg the ground?
And why bother with that when you have a perfectly good test button on the GFCI.
Are they hooked up right?
GFCI's won't test if the incoming leads are attached to the LOAD side.
Check the back of those units. The contacts are marked LINE and LOAD.
You SHOULD NOT tie the neutral to the GFI ground (this creates an unnecessary hazard), but the outlets should be labeled ungrounded. The plug-in testers won't work in an ungrounded outlet. If you want to test an ungrounded GFI with a plug-in tester, use a "suicide adapter" and ground the ground lug of the adapter to something metal nearby.
What is the unneccessary hazard if it is labeled NO EQUIP GRD and there is no load downstream?
If you tie the neutral and ground together, and a neutral fault develops upstream, then the neutral can be at line voltage, and that will put the ground (and any "grounded" device plugged into the GFI) at line voltage. I consider this to be an "unnecessary hazard".
If you feel you must provide a ground for the GFI, the better (but still not legal) solution is to ground it to a pipe or some other convenient grounding object.
Dan,
You're wrong about how to provide a grounding conductor, and wrong in a very important way. You said "the better (but still not legal) solution is to ground it to a pipe or some other convenient grounding object"
Not legal, and not effective either. Very dangerous, actually. The idea with equipment grounding is to provide a very good path back to the source of the circuit, not to earth.
A low-impedance path (via the equipment grounding conductor) back to the breaker panel will allow the fault current to open the breaker. Having the branch circuit equipment grounding conductor tied to only a "grounding object" like a separate ground rod will not do the job. The earth is rarely a good enough conductor to pass enough current back to the grounding electrode at the panel to clear the fault.
It used to be you could legally grab a ground from a water pipe, the idea being that the pipe was electrically continuous throughout the house and was tied into the panel ground buss.
That was when men were men and water pipes were metal. Now using a water pipe as a grounding conductor is allowed only in industrial facilities where the pipe is in the open and can be inspected, and there is zero chance of someone making a repair with a section of PVC pipe.
Metallic water piping in contact with the earth for at least ten feet is still the preferred system grounding electrode. It's purpose is not to clear a branch circuit fault, but to act as a drain for high energy surges like lightning or high-voltage power line crosses with the distribution lines.
Learn and live. Grounding and bonding is one of the more complex areas of electrical installation, but that won't save you. Electricity does not care whether you got it right or not, whether you got it 90 percent right, or whether it was an honest misunderstanding.
Work safe--
Cliff
I fully understand the purpose of the grounding conductor. Ideally the ground connection should go back to the panel, but remember we're talking about a GFI outlet, and the only other option we're considering is to leave it ungrounded. By attaching the gound wire to a "reasonable" grounding object, at least SOME additional protection is provided, should the GFI fail, and, in addition, devices such as surge protectors will be rendered at least partially effective (whereas they would be totally ineffective with an ungrounded outlet).
Dan,
Good intentions, but I disagree.
E = IR, do the math. Ain't no way a EGC pathway with an earth return will offer any useful protection to clear a fault. At least not with the typical soil impedance in North America.
As far a being better than nothing, I can't see it. With heavy use or line surges, a GFI will give up the ghost. I see it often. When it does, the face receptacle is still hot, even with the fancy new models. So now you have what looks like a grounding receptacle (GFI, even), with no GFI protection AND no meaningful EGC protection. Will this matter for most of the tools and appliances plugged into it? No, because they're double insulated. So why kludge together an inherently unsafe "solution" to a non-problem, that'll let you down when you need it most?
Remember, this all started because an inspector didn't realize why a plug-in polarity/GFI tester didn't work, and a dangerously unimformed "electrician" recommended bootlegging a ground just to allow the tester to trip the GFI. If it's that important to provide a test of the GFI other than the "test" button, you could use a solenoidal tester--go from hot at the outlet face to an external ground (I use either an extension cord plugged into an outlet with a good EGC, or a 14 gage stranded THHN lead that I've clipped to the ground/neutral bar at the main panel), The low impedance of the Wiggy will trip the GFI.
You might have a point about surge protection...if the "grounding object" is a very good grounding electrode. If it isn't, and there's a surge dumped to say, the water pipe you gave as an example, and the water pipe is not well bonded to ground--well, you've just energized the water piping. Besides, there are very good reasons not to have any secondary grounding electrodes in a building unless they're all bonded together and to the service ground/neutral point. My thinking is, if you need surge protection, run a new circuit with an EGC from the panel.
Work safe,
Cliff