Note, I don't think this belongs in DIY; it's a bit more complicated than that.

I bought a Lutron Maestro occupancy sensor switch, and was perplexed to learn that it doesn't need a neutral in the switch box (they brag about this as a feature, and the instructions don't mention it).


Note that in addition to the line and load (black) wires, the switch has both a bare wire and a green wire (drawing on top of 2nd page). I'm thinking maybe the bare wire (which is connected to the yoke) is the "real" ground and the green one carries the return current from the sensor and logic circuitry. But the instructions say to tie both to the switch box's ground

This seems most odd to me. I thought the whole point of NEC rule 404.2C, which requires a neutral in every switch box (with a few exceptions), is that when "smart" devices such as this are installed, that no currents are introduced into the grounding system.

I wonder what's going on. Is it:

  1. Does the switch somehow not return any current to ground ? I'm not sure this is possible, unless there's some sort of charge pump or something working off the hot wire(s).

  2. Does the switch's logic sensor and logic circuitry use so little current that it is deemed (by UL and/or NEC) acceptable for those currents to flow through the EGC ? I did not know there was a threshold.

  3. This is only meant for retrofits into older construction not compliant with 404.2(C) ? Would this even be legal ? And if so, why don't the instructions say to tie one of the wires (presumably the green one) to a neutral if it is present ?


3 Answers 3


There is more than one way to skin this cat

Manufacturers over the years have employed several solutions to the problem of how to power switch "smarts" (in occupancy sensors and the like) without using the neutral wire (as the mandate for having the neutral at the switch location first appeared in the 2011 NEC, and installations prior to it often omitted it in order to reduce cost and conserve truck-space, aka the electrician's version of BoM space, when switch loops were used):

  • Some devices (like the ones in the OP's post) use the equipment ground as a return path. This means they require a grounding conductor in the box in order to install successfully, and are largely limited to retrofit applications in homes and other small buildings to avoid excessive current on grounding conductors. (It also means that you shouldn't put one downstream of a GFCI to minimize the risk of nuisance tripping.)

  • Other devices "trickle" their operating power through the load, relying on circuitry inside to make sure that current flow through the load is not excessive and to maintain a powered state whether the internal switching elements are on or off. This means that they require a minimum load to guarantee functionality and often cause sensitive loads like LED lights to light dimly if such a minimum load (such as an incandescent bulb, or a dedicated return capacitor such as a Lutron LUT-MLC) is not present. However, they can be used in older installations where ground wires are not present.

  • Last but not least, there are a few wallbox devices that rely on a long-life primary battery (such as a CR123) for power. (Some of Intermatic's timers do this, most notably, and so do some wall-mounted fan remotes.) This, of course, means you have the downside of having to change a battery every few years in a device that you normally wouldn't expect to have a battery to change.

There is one more point to discuss with the Lutron Maestro occupancy sensor the OP asked about, and that is the distinction between the green and bare wires. Many devices that use the grounding conductor for return have a single connection to the grounding system, shared between the yoke ground connection and the return path from the internal circuits. However, the Lutron devices separate these, with the bare wire connected to the yoke and the green wire connected to the internal circuits. This allows them to be converted to neutral-based operation simply by slipping a sleeve of white spaghetti tubing over the green wire and nutting it in with the neutrals instead of the grounds (some higher-end Lutron Maestro devices already come with the tubing piece and instructions for its use, even).

  • \$\begingroup\$ Nice explanation, thanks. I amazed the instructions don't talk about the possible neutral connection for the green wire. \$\endgroup\$ Nov 6, 2018 at 20:37
  • \$\begingroup\$ How do you know this Lutron Maestro is the 1st of the 3 types you describe ? This means it doesn't suffer from the issues of the 2nd type. But connecting the green wire to a neutral is still a good idea to eliminate EGC currents. \$\endgroup\$ Nov 6, 2018 at 20:39
  • 2
    \$\begingroup\$ @RustyShackleford -- the fact the instructions say it requires a ground wire is the tipoff \$\endgroup\$ Nov 6, 2018 at 23:12
  • \$\begingroup\$ Hah ! I should read the text more and not just rely on the wiring drawing. Maybe I'll put a meter on it, but sure sounds like some current flows. Still seems odd that a component that defeats the intent of NEC 404.2(C) - preventing EGC from carrying switch operating current - would be legal while doing just that. Especially as there are the other approaches you describe. \$\endgroup\$ Nov 7, 2018 at 3:20
  • \$\begingroup\$ @RustyShackleford -- the Code folks are working on getting new installs migrated over to using the neutral, but it's going to take a while \$\endgroup\$ Nov 7, 2018 at 3:38

These devices work by pulling power from the switch loop they occupy. When “off”, they trickle a little current through the load to get power. When “on”, they actually dim just a little bit, delaying turning on just a little so that they can trickle enough power to operate. When dimmed between, there is plenty of off-time to trickle current through.

So the advantage is not needing a neutral, which is great for older houses that do not have neutrals in switch boxes. The disadvantages are that when off, enough current may trickle through to make CFL or LED light glow dimly or flash and when on, the light is not full bright.


Dimmers and sensors that do not require neutral use the load to power themselves.
This is a problem when the load lights up at the slightest current, such as LED lights.

Which is why they have a minimum load and are often not capable of switching LED lights.

  • \$\begingroup\$ So a tiny current is allowed to flow between the line and load terminals, and somehow the logic is in series with that path, so that tiny current flows through the logic and powers it ? With maybe some sort of current-to-voltage level shifter or transformer or something to give the logic what it needs ? \$\endgroup\$ Nov 6, 2018 at 7:41

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.