Jumper on auxiliary panel

[Robert Huffman]

If there is a jumper in a subpanel between the ground and neutral bars, the ground can carry voltage when it is not intended to. If this was done in an older house with no ground wire but only the hot and neutral wires, is the issue mute or are there other issues of which I am not aware?

[Jerry Peck]

If there is a jumper in a subpanel between the ground and neutral bars, the ground can carry voltage when it is not intended to.

It's not voltage that is on it, it could carry ground fault current (and, yes, a conductor is a "resistance", thus current through a resistance does create voltage across the resistance).


If this was done in an older house with no ground wire but only the hot and neutral wires, is the issue mute or are there other issues of which I am not aware?

If the older house is ungrounded (hot, neutral, but "no ground path", i.e., no ground wire, no metallic conduit, no etc), then that creates a problem with mettalic boxes.

I'll see what David says before I look back at my old codes to see how metallic boxes were addressed, but I suspect that they were addressed like metal conduits were.

In the 1897 NEC, metal conduits were not required; however, when conduits were installed, the metal conduits were required to be grounded.

Thus, without looking up the old codes, I suspect metallic enclosures for panels were likely required to be grounded.

David, what say you?

[Gunnar Alquist]

This will be an interesting thread. I have seen a fair number of mid-century homes ('40s & '50s) with the older cloth-wrapped NM (no EGC) branch circuit conductors and no obvious EGC between interior (not service equipment) panel enclosures and the service equipment enclosure.

If I remember correctly, an integral EGC was not required in a cable between the service equipment and other panels until the mid or late 1960s. My last house was built in 1963 and used teeny EGCs in the branch circuits (smaller than #14, so maybe #18?). In that case, the bare feeder EGC was run separately between the service equipment and this interior panel. In this case, the EGC was run to bonded terminals inside the panel enclosure.

Is it likely that a separate EGC was attached to the exterior of the (not service equipment) panel to connect it to the service equipment in these earlier systems? That doesn't seem reasonable. I do not recall seeing any additional conductors that could be a feeder EGC.

[david shapiro]

Thanks, friends, for the tip of the Hatlo hat.

Most commonly, I'm sure, when any of us sees what I think Robert is describing, the subpanel started out as the service panel, was converted by using the original service cable as a feeder, and that was plumb illegal until adoption of the 2023 NEC made the changes found in 250.140(B). Since when until then? To quote from Chapter 28 of Behind the Code, ". . ..In 1910, a three-year-old National Bureau of Standards (NBS) appropriation led to a solution: insulated feeders. . . ." But it took years to get this nailed down into the code. If this became a subpanel within the last oh, 85-90 years, violation. Unless it was an legit subpanel, they used a grounding raceway containing a neutral, and then neutral and ground could be separated, confirming that the raceway path is intact.

The safety issue definitely is not moot. Frequently in houses like those to which Gunnar alludes, ones wired using non-grounding cables, everything's locally bootlegged to the grounded conductors, in violation of 250.130(C). The main out, 250.86 Exc.1, says that existing enclosures wired by methods that legitimately lacked grounding conductors need not be grounded, but its terms are procrustean. (2) Any run over 25 ft, no exemption. (3,4) Accessible by people or to metal lath, ground (dirt, concrete, . . .), grounded metal, no exemption.

Now grandpa isn't mentioned here, so if no one touches it, it doesn't become illegal. But if I remember right, you're mostly doing safety inspections, not code inspections. The more-general question becomes this: is there a safety hazard to having a non-grounded metal box? If this were the only issue, I'd flag it as a "not in keeping with today's practices." I'd recommend adding GFCI protection unless it were a lighting outlet in the ceiling, or maybe a clock-hanger. However, this particular panel can be made legal through the radical change to the downstream grounding rules I pointed to, sponsored by safety advocate (and IBEW mouthpiece) Jim Dollard.

Now, (not very) funny thing--and here I point back to Jerry's warning--if the circuits downstream of this subpanel had grounding conductors, all good. If, OTOH, you were referring not just to the feeder but to the branch circuit conductors when you mentioned nongrounding circuitry--and you meant cable, not conduit--then you could have a serious case of the Gramps. Anywhere you see a three-prong receptacle that's plausibly fed from the subpanel, even if the wiring is in good shape you have reason aplenty to advise getting in an electrician to investigate and remove bootlegs, add GFCIs, all like that.

Parenthetically, the change in box fill requirements from one conductor per device to two came about in response to proposals (modestly muttered: mine was one of them) citing the crowding caused by replacing vanilla 5-15Rs with GFCI devices. Hence if some of these boxes were full up as installed, that swap could prove to be an impracticable solution.