Another grounding question

Status
Not open for further replies.
You may have a totally unrelated issue, is the ground from the uf cable terminated to anything at the source end? such as cold water, ground rod ect? If you are able to read voltage from earth to the box, turn off the breaker or remove the fuse feeding the receptacle. If the voltage goes away, the issue is with that installation, if it does not, Turn off other circuits in the panel one at a time until the voltage goes away. There is probally a high resistance fault to ground on one of the circuits, and the box may have a better ground than you think, and may be providing a return path for the fault.
 
erickench said:
If there is no metal raceway and no equipment ground wire then you would have to use the neutral otherwise there is no effective fault current path. If the hot wire touches the box then you run the risk of someone being electrocuted. It could be that the metal yoke of the receptacle is electrically connected to the box and because it's a GFCI it has no grounding terminal. That may be the reason for the current flow to the box. Rubber washers might solve the problem.
Click to expand...

I thought that the yoke was part of a rated assembly and bonded thus creating a continuation bond in a GFCI, the green screw through to the yoke is what get "bonded", as there installed, modern day.

The chassis of a GFCI is electrical isolated by design. while the frame might well carry the components they are manufactured to be isolated!
 
I would unscrew the receptacle from the box and then check and see if there is a tingling sensation. If so then that's where the current is coming from. The GFCI is probably old and there is a short somewhere inside it. As for the ground, if there is a metal raceway or a ground wire going back to the service then by all means use it. All I'm saying is that a neutral should be used as a last resort if you don't have anything else.:cool: The earth and a persons body in series is a resistance. A wire with very low resistance running from the box back to the service would shunt out any current flowing through a persons body to the earth. The fact that we are feeling a tingling sensation means that the neutral is tied to the grounding electrode back at the service as it should be.
 
Since the house has no ECG's, if the outside recept in a metal box is fed with UF from another recept. , I think it would be a good idea to make THAT recept the GFCI, rather than the outside one, so personnel protection was provided on the entire circuit including the box, rather than just what was plugged into the GFCI.

I think that would also eliminate whether or not the problem was from this circuit or another source.....assuming the GFCI actually works.
 
Last edited:
realolman said:
Since the house has no ECG's, if the outside recept in a metal box is fed with UF from another recept. , I think it would be a good idea to make THAT recept the GFCI, rather than the outside one, so personnel protection was provided on the entire circuit including the box, rather than just what was plugged into the GFCI.
Click to expand...

that is assuming that it is fed from a recept and not a j-box in the basement.
 
realolman said:
However it's done... maybe even put in another recept in the house... my thought is to put the entire outside portion of the circuit on a GFCI.... with a useful grounding conductor.
Click to expand...

while your at it, extend the circuit all the way to the panel and drive a groundrod...:grin:
 
erickench said:
Because there is no ground you cannot bond the box. It's ungrounded...
Click to expand...

Maybe but in error/violation, A system is ungrounded only by the fact it's supply is not connected to earth, intentionally. If a supply is a grounded system (XO or a phase connected to earth) it remains a grounded system even without the existence of an equipment ground. Fault current is not trying to return to earth it is trying to return to its source.

Earth is not an acceptable fault current path because it resistance is unknown at any given moment, especially in premises wiring ohms law will prove earth as an ineffective fault path.

Neutral current is not allowed on anything other than a conductor.
 
erickench said:
I would unscrew the receptacle from the box and then check and see if there is a tingling sensation. If so then that's where the current is coming from. The GFCI is probably old and there is a short somewhere inside it. As for the ground, if there is a metal raceway or a ground wire going back to the service then by all means use it. All I'm saying is that a neutral should be used as a last resort if you don't have anything else.:cool: The earth and a persons body in series is a resistance. A wire with very low resistance running from the box back to the service would shunt out any current flowing through a persons body to the earth. The fact that we are feeling a tingling sensation means that the neutral is tied to the grounding electrode back at the service as it should be.
Click to expand...

eric, as most have already said, this is not the way to go, not even as a last resort, never.
 
erickench said:
... As for the ground, if there is a metal raceway or a ground wire going back to the service then by all means use it. All I'm saying is that a neutral should be used as a last resort if you don't have anything else.:cool: ...
Click to expand...
I'm sorry, but under NO CONDITIONS should a grounded conductor ever be used as an EGC on the line side of the main bonding jumper or the system bonding jumper. There is NO SAFE way to ever do this and the hazard that is created by doing this is much greater than the one you are trying to eliminate.
 
mikeym76 said:
......and ground to green. However, placing a knee on the physical ground and one hand on the box you will feel a slight tingle.
Click to expand...

I just noticed that there's a ground jumper running from the receptacle to the box. Look's like someone tried using the wrong type of GFCI with a metal box. I would replace the GFCI with one that is marked: "No Equipment Ground". Then replace the metal box with a nonmetallic one. You're running a great risk using an isolated metal box. If a hot wire touches it then it will become energized. The shock you'll receive will be much greater then the one that you are getting now.:smile:
 
I think the exception of 250.140 that allows the frames of electric ranges to be grounded to the grounded circuit conductor does not apply to this situation.

But it does seems to indicate that it is not all that dangerous to do what erickench is suggesting; and in fact must be safer than leaving a potentially energized metallic part isolated.

My own feeling is to not do what erickench suggested.... It's just a receptacle ...fix the thing right.

ultramegabob said:
while your at it, extend the circuit all the way to the panel and drive a groundrod...:grin:
Click to expand...
 
The jumper is running from the GFCI receptacle to a box that has no return path. For all we know this could be causing the problem. Your only supposed to use GFCI receptacles marked: No Equipment Ground for this type of situation. It's outlined in NEC 406.3(D)(3) as being the rule whenever a nongrounding type receptacle is replaced with a GFCI. The UF cable has only two conductors and therefore no equipment ground. If there is no metal raceway then you're going to have a problem. You might as well use a nonmetallic box.:roll:
 
erickench said:
The jumper is running from the GFCI receptacle to a box that has no return path. For all we know this could be causing the problem. Your only supposed to use GFCI receptacles marked: No Equipment Ground for this type of situation. It's outlined in NEC 406.3(D)(3) as being the rule whenever a nongrounding type receptacle is replaced with a GFCI. The UF cable has only two conductors and therefore no equipment ground. If there is no metal raceway then you're going to have a problem. You might as well use a nonmetallic box.:roll:
Click to expand...
There is only one type of GFCI receptacle. There is not one that is factory marked "no equipment ground". That is a field applied sticker that is used when an old two wire, non-grounding receptacle is replaced with a GFCI in a box where there is no equipment grounding conductor available.
Unless this is a very very old installation, the UF will have an EGC within the jacket.
 
erickench said:
Don, what does 12/2 UF mean?
Click to expand...

in UF and Romex the first number is the gauge of the wire, the second number is the number of insulated conductors in the cable, they do not include the ground wire in the number for reasons unknown by me, technically you need to specify 12/2 with ground, but in modern days, the ground is just assumed to be in new cable.
 
Status
Not open for further replies.
Top