I have been working on wiring a Z-Wave 3 way dimmer. In my current three way configuration, I only have line, ground, and 2 travelers in the remote box.

In the primary box, I have 2 travelers, ground, load, and neutral.

It seems that in nearly every configuration (that supports LEDs and Fluorescents), both boxes need a neutral wire.

I could achieve this if I could use one of the traveler wires to get neutral to the remote box. To do this, I'd need to take power from another hot wire in the same box as my dimmer.

The dimmer is in a 2 gang box with the switch for my outdoor lights. Both the outdoor lights and the indoor light (that's on the dimmer) are on the same breaker.

My concern was that if I took too much power on the hot wire for the outdoor lights that it could be a fire hazard; however, it seems like if they're on the same breaker, that both of those loads are going into the breaker box on one wire at some point anyhow so they must have a high enough ampacity for both loads.

Is it safe to use the hot line from another switch like this?


Forgive me; these are my first wiring diagrams

Here are the current switch configurations Master Dimmer / Multi-gang Box Secondary Switch

And what I think the current diagram would look like Current Wiring Diagram

This is what I'm proposing as an interim until I understand more about 3-way Z-wave switches Proposed configuration

What I'm hoping to accomplish someday if I can get a better understanding of inductive coupling: Ideal case for Z-wave slave

  • My tiny brain can't comprehend your wiring without a diagram, could you include photos or a diagram of the wiring? I can't understand how one box has an ungrounded (hot) conductor ("line"), without also having a grounded (neutral) conductor. Based on your description, the circuit looks like this to me.
    – Tester101
    Apr 13, 2016 at 13:47
  • @Tester101, that's what it looks like, yes. I'll draw a diagram with the other switch later this afternoon.
    – D. Patrick
    Apr 13, 2016 at 17:03
  • @Tester101, I added diagrams and photos from the switch boxes.
    – D. Patrick
    Apr 14, 2016 at 1:41
  • 1
    For reference, this is how your circuit is wired. Not that it helps much, but simply so you can have a more accurate picture of what's going on. Also, it looks like one of the black wires is falling out of the twist-on wire connector, which is not good.
    – Tester101
    Apr 14, 2016 at 2:13
  • 1
    In your first image, I can see a black wire going from the twist-on wire connector to the same cable as the red and white wires. See. I also know that NM cables that have red wires, also have black wires in them. That twist-on wire connector is where the incoming "hot" wire branches.
    – Tester101
    Apr 14, 2016 at 3:49

2 Answers 2


Yes, as long as you return power on the same neutral that's partnered with the "hot". (Why is that? Because if hot wires are protected by breakers, but neutrals are not - so if a neutral carries more than its load, that will not be detected.)

If you are converting to smart-switches, you should be able to convert the hot+2 messengers into hot, neutral, signal for the smart switches.

  • I just realized here that you actually answered a question I had that I hadn't asked. I am trying to convert to smart switches. I imagine in my actual wiring that there's a junction box somewhere in which a 12/3 gets split and hot runs to the slave box and neutral runs to the load box. Then, the travelers are left intact. If I convert it as you said for the smart switch, then everything seems to pair up. For the slave span, I have hot in and neutral out. For the load span, I have neutral (traveler) in and neutral out (to breaker). The signal wire is probably negligible voltage then?
    – D. Patrick
    Apr 14, 2016 at 1:48
  • 1
    Are you guessing here? That thing, where hot goes to one junction box, and neutral to the other, should not happen ever. It makes sense in DC electronics, but in AC there is a serious issue with magnetic fields. The upshot is if you digram out the current flow, every cable should have current moving in opposite directions. Apr 14, 2016 at 2:09
  • I'm definitely guessing. Tester101 gave me the impression that the hot wire in the slave switch j-box probably came from the master switch j-box. That probably makes more sense than my previous hunch. I was thinking that because the red/white wire doesn't appear to have a black wire coming out of it in the master j-box but it does in the slave j-box. That's why I was thinking they split somewhere out of sight. But, I also don't see any single white wires in the master j-box so that makes sense.
    – D. Patrick
    Apr 14, 2016 at 3:09
  • 1
    The normal way that would be done is to take the "hot" to the main switch box (it already goes there) and then have it double back along the 12/3 to the remote box. That way each cable has power flowing both directions down it, and it cancels magnetic fields. Notice your drawing shows 12/3 but you have only drawn two wires, so the hot would be the third wire coming all the way back to the big switchbox. (12/3 is red-black-white-ground, the ground is not counted as one of the /3.) Apr 14, 2016 at 6:57

If all the loads are on the same circuit then the neutral can't be overloaded without the hot wire also being overloaded and consequently tripping the breaker. You have a good grasp of this concept. You never want to take power from one hot wire and return it on a neutral from another circuit. That could overload the neutral, and as Harper points out, neutrals are not protected so this would be a dangerous situation.

The requirement to use the same neutral that is cabled with the hot wire is in the National Electrical Code in article 300.3(B).

The major reason for this is inductive heating of ferrous materials. This occurs when the electromagnetic field on a conductor is not cancelled by an opposing magnetic field on a conductor with current traveling in the opposite direction.

Sounds like you have a pretty good idea what you are doing.

Good luck!

  • I've been seeing a lot about 300.3(B) and the need to avoid inductive coupling. Part of my confusion is that there seem to be 5 or so neutrals in the primary box that are all spliced together. Is that still up to code? I mean, they're not quite 1 to 1. I think a handful of them are inbound from the lights. I'll bet that's what it is. One of the neutrals is the one that's going out with the load. I'm not sure where the neutral from the secondary switch ended up. I've been assuming it's one of the neutrals in the primary box. I'll draw a diagram.
    – D. Patrick
    Apr 13, 2016 at 17:09
  • 1
    5 neutrals is not that unusual. In fact, the problem condition (2 supply neutrals) would leave only 3-4 load neutrals to split among them, so only 1-2 each. I've seen 10 neutrals together, that is a place I'd look for that problem. Apr 13, 2016 at 20:11
  • OK, so inductive coupling becomes an issue when there's a large load difference between hot and neutral over a longer distance? There's definitely some distance in the existing setup where the hot wire travels without a neutral wire. But, it seems like that has to happen sometimes. Oh, or the inductive coupling is reduced in a three way switch because one of the travelers is conducting load current but in the opposite direction.
    – D. Patrick
    Apr 13, 2016 at 21:32
  • 1
    Yes it doesn't have to be a neutral to carry current in the opposite direction. It can be either of the travelers counteracting a hot wire or switch leg. It only affects ferrous materials but a lot of cables are installed in steel boxes or use steel cable clamps.
    – ArchonOSX
    Apr 14, 2016 at 0:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.