Sub Panel circuits without grounds or neutrals

I’m getting conflicting opinions from electricians. Sub panel has separate ground and neutral bar but the individual circuits do not. They are in the main panel. Is this acceptable to code?

That’s not surprising. There is much debate out there about this scenario.

Search the forum for “300.3(B)” and you can also read the thread at the link below for more insight…

I would think that’s it’s not okay, being how circuits fed from a sub-panel should have isolated neutrals.

But it’s an interesting scenario to say the least and I’m curious to see what Robert @rmeier2 and others have to say about it.

Actually, the isolation aspect doesn’t really come into play in this scenario. With the neutrals and EGCs’ for these circuits bonded together only at the main panel, there are not multiple paths involved.

Electrically I don’t see an issue with it (it’s similar to a switch loop) but the NEC makes no provision to leave the neutrals only in the main panel due to 300.3(B).

Ryan posted a lengthy thread from Mike Holt that runs the gamut of varying opinions.


So technically not a problem.

Thanks for the quick reply Robert.

So each of these circuits has their own neutral and ground, they just terminate in the service panel.

In a 2-wire switch loop you have only two conductors run to the switch. One gets wired as the hot leg the other gets wired as a switch leg. Electrically this is okay because the magnetic fields are both in the same raceway or cable and the fields cancel each other out. If you look at the photo in the OP you’ll see that the same scenario exists where the feeder conductors and the branch circuit hot legs are both in the same raceway and the magnetic fields of each will cancel out. So although electrically this is safe there is no wording in the NEC that permits such an installation. Wall switches have been wired this way for over a hundred years for no issues. If you look at this graphic below you can see the 2-wire switch loop.


This is quickly getting deep into the rabbit hole, but what happens when the magnetic fields do not cancel each other out? What scenario would the magnetic fields not cancel each other out, and what are the consequences?

You have heard of “cheater” circuit breakers? This is a cheater subpanel.

I’ll try to keep it simple. By running the ungrounded and neutral conductors in the same raceway or cable there is no induction in the surrounding ferrous metal panel enclosure (same would apply to a switch loop). If you ran all of the ungrounded condcutors in one raceway and then ran the neutrals in a separate raceway there would be inductive heating in the surrounding ferrous metal between the raceways which is something that the NEC tells us we need to avoid which is why the code has 300.3(B).


Morning, Robert.
It has been some time that I have read someone mentioning induction in electrical, would it be safe to say; By running the ungrounded and neutral conductors in the same raceway, or cable, there is no ‘magnetic induction’ in the surrounding ferrous metals like the panel enclosure. Magnetic induction could heat ferrous metals, metal containing iron, in close proximity almost instantly.

That’s correct, unwanted induction can heat the surrounding metal.

1 Like

This is exactly how induction bolt heaters work.


Not to go down another rabbit hole, but @mparks3 posted a good video about this but I can’t seem to find it.

Edit: this may be the one I was thinking of. Tracing Magnetic Fields EMF by Karl Riley - YouTube

1 Like

Oh boy, did you just create an invitation? :scream:


I presume that when you say “Acceptable to code”, you mean the National Electrical Code, in which case the answer is no, it is not acceptable.