grounded plugs/no ground pin

Status
Not open for further replies.
jute

jute

Senior Member
Location
SO CAL
I have a condo in that I am changing all outlets, switches, fixtures, etc...There is the old cloth romex feeding all the receptacles and are the old 2 prong without a ground pin. The problem I have is the ground is ran through the back of the j-box and connected by a screw on the top on the outside of the j-box so the only way I can get to the ground is to remove the j-box which I don't want to do... They are all wired this way and need to know what I need to do to make it right...The options I see is to use self-grounding receptacles or GFCI's spaced throughout the house. Aren't all receptacles self-grounding now so I wouldn't need to attach a ground to the receptacle itself becuase the receptacle is grounded to the metal j-box by the mounting screws which the ground wire is attached to?? Suggestions appriciated...Thanks, JB
 
Last edited:
didn't edit my post fast enough...couldn't I also add a ground wire and screw it into the back of the j-box inside??
 
Jute,
Good practical question. :smile:

In concept, and that is what this forum is about, you need to ground or immulate a ground for the EGC. Pigtailing to the Jbox, assuming the existing EGC connection is under 1 Ohm to the Jbox, should work. Would you trust it? I would put GFCI's at the first receptical in each circuit. If the existing wiring leakage is too much, then you would have to install more GFCI's in the circuit. Old 40's romex is a royal pain for developing leakage currents, so do not ever bend it, and always use the liquid tape to keep the remainder of the joint insulation from fall off. I have even crimped butt-splices onto stranded THHN for pigtails, just to keep from breaking up the old insulation. I believe in the liquid tape. I'd rather work with K&T than 40's romex. :smile:

To conclude, IMO the problem is that the existing EGC screwed to the outside of the Jbox is an unknown as to having good contact and low resistance.

If you can slide your Simpson probes in the crack above the Jbox and measure EACH one at < 1 ohm to box, then maybe you will have an adequate EGC to trip the OCPD, and just pigtail to the Jbox. Otherwise, for safety, I'd use the GFCI's, and even try more spaced out on the same circuit. Leakage is a very 'IFFY' thing. Be ready for surprises. :wink:

Good practical question. :smile:
 
jute said:
didn't edit my post fast enough...couldn't I also add a ground wire and screw it into the back of the j-box inside??
Click to expand...
If I didn't know where that scabbed-in EGC came from, or where it's going, or what condition it was in, I would be inclined to go the GFCI route too.
 
glene77is said:
Jute,
Good practical question. :smile:

In concept, and that is what this forum is about, you need to ground or immulate a ground for the EGC. Pigtailing to the Jbox, assuming the existing EGC connection is under 1 Ohm to the Jbox, should work. Would you trust it? I would put GFCI's at the first receptical in each circuit. If the existing wiring leakage is too much, then you would have to install more GFCI's in the circuit. Old 40's romex is a royal pain for developing leakage currents, so do not ever bend it, and always use the liquid tape to keep the remainder of the joint insulation from fall off. I have even crimped butt-splices onto stranded THHN for pigtails, just to keep from breaking up the old insulation. I believe in the liquid tape. I'd rather work with K&T than 40's romex. :smile:

To conclude, IMO the problem is that the existing EGC screwed to the outside of the Jbox is an unknown as to having good contact and low resistance.

If you can slide your Simpson probes in the crack above the Jbox and measure EACH one at < 1 ohm to box, then maybe you will have an adequate EGC to trip the OCPD, and just pigtail to the Jbox. Otherwise, for safety, I'd use the GFCI's, and even try more spaced out on the same circuit. Leakage is a very 'IFFY' thing. Be ready for surprises. :wink:

Good practical question. :smile:
Click to expand...

I thought I should probably go with the GFCI spaced throughout but would I need to measure ohms from the ground wire scabbed to the top of the box or could I measure from the box itself? Also are the self-grounding receptacles necessary if I install GFCI's? Thanks again for the replys...JB
 
Take a close look at 406.3(D) and 250.130(C).

Actually, thinking on this more, with the EGCs being in place, you might run afoul of the separation of receptacle EGCs required by 406.3(D)(3)(b) and (c).

I guess that leaves you with verifying the integrity of that existing EGC.
 
jute said:
would I need to measure ohms from the ground wire scabbed to the top of the box or could I measure from the box itself?
Click to expand...

If you use GFCI's then you do not need a EGC, 406,
because the function of the GFCI is to detect the very fault current that the EGC carries back to the panel. The GFCI monitors for 5mA of fault current going back to ground (possibly through a persons body) and does not care where this ground located. By NEC, the GFCI can replace the function of the EGC.

IF you are measuring resistance from the EGC to the Jbox,
THEN measure from the EGC wire to the Jbox.

IF you are using GFCI's
THEN don't bother with the EGC. :smile:
 
George Stolz said:
Take a close look at 406.3(D) and 250.130(C).

Actually, thinking on this more, with the EGCs being in place, you might run afoul of the separation of receptacle EGCs required by 406.3(D)(3)(b) and (c).



I only have the NEC 1999 here and it doesn't have an art 406? but the way it sounds is that the EGC must be connected to the receptacle continuously and using the box for continuity is a no go?


I guess that leaves you with verifying the integrity of that existing EGC.
Click to expand...
Can I measure fromt the j-box itself since I can't get to the EGC scabbed to the top??

Thanks, JB
 
My entire house was wired this way when I bought it. If you're sure that each metal box is grounded I would just install self-grounding devices.
 
jute said:
18 views and no comments?? Did I word it wrong??
Click to expand...
Just an aside note: JB, a lot of the posters here are full-time working people, and we're here voluntarily. You might see a question viewed 50 times before someone with a contribution has the chance to read and respond.

You posted this "complaint" one hour and eight minutes after your first question. It may take a full 24 hours for everyone to see every question posted, depending on their individual schedule. Please be a bit more patient.
 
LarryFine said:
Just an aside note: JB, a lot of the posters here are full-time working people, and we're here voluntarily. You might see a question viewed 50 times before someone with a contribution has the chance to read and respond.

You posted this "complaint" one hour and eight minutes after your first question. It may take a full 24 hours for everyone to see every question posted, depending on their individual schedule. Please be a bit more patient.
Click to expand...

That was a bit of sarcasm and definately taken the wrong way...I've been a member of this forum since 2003 and have a whopping 145 post (I think). I read alot more than I post but have always considered this forum and it's members to be the best. I've asked for help on how to proceed on certain projects to what kind of tools are preffered. This forum has pulled me out of the mud more than once and is always more than appreciated. I value this forum and have always bragged on how fast I get responses and usually get a code article to back them up, just like yesterday morning. So again THANKS to everyone for the help and in no way was my comment intended as a "complaint"... JB;)
 
Last edited:
realolman said:
I can see one on a circuit, but I'm not getting the reasoning behind the multiple ones.:-?
Click to expand...


What would happen if there was (12) duplex receptacles on a circuit and had (3) GFCI'S spaced on #1 > #6 > #12 positions then #12 detected a ground fault. Would all (3) shut off or just #12 GFCI?? just a thought...JB
 
Last edited:
realolman said:
I can see one on a circuit, but I'm not getting the reasoning behind the multiple ones.:-?
Click to expand...

Real and Jute,
Thanks for questioning my method.
IF:
___In practice, in a 30's house, with 40's repair and additions,
with lots of leakage through the very old rubber insulation,
and having carbonized soot covering the K&T and 40's romex in the attic,
THEN:
___GFCI's may tend to trip more often that not.
FIX:
___I use them downstream to split up the leakage, the furthest one will trip first.
And sometimes you just have to cut the lines and re-string with NM-B.

Maybe I should not have mentioned this fix method,
as it may never happen to you, and just brings up a question.

I have had occasion to separate parts of a K&T circuit,
cutting the wires, making partial circuits, which can be re-fed.
I then run a new HomeRun in NM-B to a jBox,
then spider-web out to the partial K&T circuits.
This can avoid some leakage,
and can also avoid arcing/loose joints at the telegraph wraps.

And now I suppose someone is going to ask me
what a 'telegraph wrapped' joint is! :smile:

Always be ready for surprises in Old Work.
Always express cognitive flexibility! :;):
 
Last edited:
glene77is said:
Jute,
Good practical question. :smile:

In concept, and that is what this forum is about, you need to ground or immulate a ground for the EGC. Pigtailing to the Jbox, assuming the existing EGC connection is under 1 Ohm to the Jbox, should work. Would you trust it? I would put GFCI's at the first receptical in each circuit. If the existing wiring leakage is too much, then you would have to install more GFCI's in the circuit. Old 40's romex is a royal pain for developing leakage currents, so do not ever bend it, and always use the liquid tape to keep the remainder of the joint insulation from fall off. I have even crimped butt-splices onto stranded THHN for pigtails, just to keep from breaking up the old insulation. I believe in the liquid tape. I'd rather work with K&T than 40's romex. :smile:

To conclude, IMO the problem is that the existing EGC screwed to the outside of the Jbox is an unknown as to having good contact and low resistance.

If you can slide your Simpson probes in the crack above the Jbox and measure EACH one at < 1 ohm to box, then maybe you will have an adequate EGC to trip the OCPD, and just pigtail to the Jbox. Otherwise, for safety, I'd use the GFCI's, and even try more spaced out on the same circuit. Leakage is a very 'IFFY' thing. Be ready for surprises. :wink:

Good practical question. :smile:
Click to expand...

I believe you mean emulate.

Where did you find the requirement for this?



glene77is said:
If you use GFCI's then you do not need a EGC, 406,
because the function of the GFCI is to detect the very fault current that the EGC carries back to the panel. The GFCI monitors for 5mA of fault current going back to ground (possibly through a persons body) and does not care where this ground located. By NEC, the GFCI can replace the function of the EGC.

IF you are measuring resistance from the EGC to the Jbox,
THEN measure from the EGC wire to the Jbox.

IF you are using GFCI's
THEN don't bother with the EGC. :smile:
Click to expand...


This is too broad of a statment to make on a forum such as this where so many people are just lurking. It would be better if you were to be more specific in nature when you make a statement such as this.
 
Status
Not open for further replies.
Top