3

I'm fixing some wiring mistakes in my house and installing some GE Z-Wave Plus switches at the same time. Where there were 3- and 4-way switches I'm using the GE Add-On switch to add additional control points.

Because the Dimmer switches have a nice blue LED to show when they are on/off, and the Add-On switches do NOT have this LED, I want to place the dimmers in certain locations, not always in the box where line enters or load leaves.

Are the following diagrams all 'legal' with respect to code and safety? The diagram on the bottom right is the one I'm most concerned with. Is it legal to have neutral enter the box from a different wiring path as line does, if they are on the same circuit?

Diagram showing the four combinations of line entering and load exiting  either the box where the main switch or add-on switch is.


UPDATED diagram:

enter image description here

  • 1
    The bottom right example is a example of an incorrect circuit. – Dan D. Apr 11 '18 at 20:23
  • 1
    Current supplied by a cable must be returned by that cable. The total current in every cable must be zero. One can test this with an AC clamp meter. – Dan D. Apr 11 '18 at 20:25
  • 1
    Also, @DanD. I'm happy to accept your comment as an answer if you post it as an answer. – Phrogz Apr 11 '18 at 21:26
  • 1
    By " cable" we mean what UL and NEC mean. A single sheath with markings indicating UL listing, type,size, stranding, number of conductors, brand, milepost, etc. You can modify your drawing to deploy /4 cable for instance. – Harper Apr 11 '18 at 21:31
  • 1
    /4 is black white red blue. It may be easier to find /2/2 cable, black white red white-red-stripe. It has become a popular necessity owing to 404.2. – Harper Apr 11 '18 at 21:50
2

Imagine ants climbing up a tree to harvest its fruit. It is a proper "tree topology" - None of the branches ever touch each other. Ants are allowed to scurry up and down any branch in any order, but cannot leap or fall from branch to branch.

At any point on the tree, if we clamped an "ant current meter" around the branch, it would count exactly as many ants going up the branch as down. Once past that point, ants might zig-zag up any number of sub-branches, but the ant must come back past this point at the end of the day. 144 ants go up past the point , 144 ants come back. Ant currents are equal. If there was a Residual Ant Fault Current Device, it would not trip.

If any of the branches touched each other, it would be possible for ants to go up branch 1 and come back on branch 2, making ant currents unequal on those branches.

So a "tree topology" is a pretty strong defense against the unequal currents problem. How do you detect whether your diagram is a tree? The "paint bucket" fill tool makes easy work of it. Trees don't enclose areas, so the paint should be able to reach all points (except between wires inside a cable.)

enter image description here

Nope, in #4, the wires enclose a space, so it's not a tree diagram, and currents are not equal because currents can encircle the space, imbalancing currents in every cable/raceway on each side of the space.

1

In your original diagram, the bottom right configuration violates NEC 300.3(B)/310.10(H) due to the neutral current back from the switch taking a different route (cable or raceway) to get back to the panel than it got to you via the hot powering the switch. Or in other words, if you clamp a clamp-on ammeter around the cable between the two boxes, it would read something other than 0, which points back to 300.3(B)/310.10(H) as another way of stating that part of Code is that "all currents in a cable must add up to 0, algebraically speaking (i.e. one direction is + and the other is -)". Oops!

Your updated diagram fixes the bottom-right configuration, using a /4 cable to move the neutral into the same cable as the hots.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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