In the '99 NEC this section reads--"Where the box is mounted on or at the surface,..."

In the '02 and '05 NEC this section reads--"Where the box is mounted on the surface,..."

What is the intent of this change?

If a drywall ring is flush with the wall, is it acceptable to ground the receptacle by direct metal to metal contact with the ring?

These changes sort of tend to not allowing this anymore.

Thanks

No. A self grounding receptacle or a bonding jumper must be used.
Don

Yes, a bonding jumper MUST be used because if you remove the receptacle from the box it would no longer be bonded. Same rule applies to switches for the same reason.

Both versions also say (in effect) "unless grounded as in 250.146(A) through (D)"

C & D only apply to certain situations, but "B" is required for "Flush Type" boxes. Whatever a flush type box is? I wouldn't consider a 4s ring in plaster a "Flush Type", even if flush. Or a masonary, even if flush. A metal cut-in yes... (Installed so it is flush.)

Or as Don pointed out 'self grounding' receptacles.

There is no requirement that the device must remain bonded when removed from the box.

Why not?

A flush box is one that is not recessed below the plane of the finished wall. Tile or plaster rings become part of the box, the same as box extensions become part of the box. If the tile ring is flush, and is properly connected to the base with two 8-32 screws, then the receptacle can be properly bonded through the two 6-32 screws made up tight with "self-grounding type receptacles".

would 250-86 exception 3 apply that states a metal elbow underground in a nonmetallic run that is isolated and 18 inches below ground shall not be required to be grounded.250-84 is service but than 250-86 is other conductor enclosures and raceways.i posted this in the wrong topic. can the moderator move it to the gec topic

[This message has been edited by dannynova (edited 12-17-2005).]

Don, I agree that is the intent, is there any documentation that supports this? As older versions of the code had less than perfect language (imagine that). Even the handbook doesn't have good examples to show the intent. (Plus the '02 doesn't have a line next to the wording change. Maybe I have a '99 with printing errors?)

Now 404.9(B) says general use snap swithces are considered effectively grounded if "..the switch is mounted with metal screws to a metal box.."

A general use snap switch by NEC definition appears to be what is regularly used in most buildings to control the lights.

Is ShockME77 wrong about switches and grounding/bonding?

Steve, you've hit on the issue behind this code change. Is the strap in solid contact with the box?

I almost always find a small gap between the strap and the mud ring; even if the box was installed correctly, with the correct mud ring. Maybe the "texture" aplied to the wall is what causes this. It is also quite common to see various things used to 'fill the gap' between the strap and the mud ring, and give the secure mounting we want.

This makes the mounting screws, by default, the ground path. Now, I can personally recount interested parties in the '60's worrying about that fiber retension washer, and whether it might interfere with a good connection. There is a camp out there who are convinced that the mounting screws cannot be relied upon for a ground path, under any circumatances.

Add to this the common use of wiring methods that use plastic boxes and supply a ground wire- it has to go somewhere. So, we've seen the code go through a number of revisions that, bit by bit, are eventually going to mandate that all devices be attached with pigtails to wires in the box. It's creeping in.....

In a parallel developemnet, the code a few cycles back began to require the use of two screws to attach devices to "industrial" or "surface box" covers. So, now you have the strap clamped firmly to the cover with screws/nuts in two places. Under those circumstances, it's kind of hard for the "pigtail" crowd to claim an unreliable ground!

Now, I will admit to lacking enthusiasm for some of these "improvements" to the code- but that's a whole different discussion.

The whole point of the slight change in code language was to require the use of pigtails.

Steve,
The following is from the panel statement on comment 5-257 in the 2001 ROC.
This is a comment on proposal 5-282 that eliminated the words "or at" from this code section. You can find links to these documents here .
Don


[This message has been edited by resqcapt19 (edited 12-17-2005).]

Grounding switches? Anyone?

Bob, I looked it up after reading some of the other posts and re-read 250.146 (B) and I have a much better understanding of this particular code. I always knew to use self-grounding receptacles, especially in old houses (which I work in alot), but I never knew why I wasn't required to install a bonding jumper. Thank you.

--Ron