2

I'm looking to replace one of the two switches (wired in a standard 3-way configuration) which control an overhead light in a utility room, with an occupancy sensor switch. The room is small enough that the one sensor should provide adequate coverage.

I've purchased an Eaton Model #OS306U, and the wiring instructions look most peculiar to me. The switch has 3 wires, black, red, and blue (plus ground); I'd expect one of the wires to be the "common", and the switch makes a connection between common and either the red or the blue wire, much like a standard mechanical 3-way switch. But it's not wired like that at all. Here is an image from the instructions, for my scenario:

enter image description here

Neutral is wired as normal. The occupancy sensor turns the light on or off by making or breaking the connection between its black and red wires (and the black, being connected to hot, probably powers the logic as well). Meanwhile, the blue wire goes back to the other mechanical switch; this switch either leaves the blue wire open, or shorts it to ground. Apparently the mechanical switch needn't even be a 3-way switch, a single-pole one would be fine (despite the labeling in the drawing).

I can't quite figure out what's going on.

Improve this question
14
  • 1
    But if you think about it, you wouldn't want things wired as a normal 3-way. Suppose you walk into the room and turn the light on with the mechanical switch; then once the sensor sees you, it turns the light off.
    – RustyShackleford
    Feb 10 '19 at 23:20
  • I actually laughed out loud. So true!
    – Harper - Reinstate Monica
    Feb 10 '19 at 23:52
  • 1
    "sensor sees you and turns the light off"... I see so many attempts to combine different devices with little thought to how they'll interact, and that just sums it up!
    – Harper - Reinstate Monica
    Feb 11 '19 at 0:10
  • 1
    Is your ground reliable and genuine?
    – Harper - Reinstate Monica
    Feb 12 '19 at 0:55
  • 1
    So I connected a floor lamp to hot and ground with jumper clips. Lights up fine. Current is about 1/4 amp (had a CFL in it). No noticeable difference in measured voltage with the lamp turned on or off (123.0 on Fluke DVM). Looks like a solid ground and an utterly faulty sensor switch. I exchanged it for a single-pole Lutron I've used in my shipping container that works like a champ; this room is 10x11ft so I think I'll be fine with the occupancy sensor and no second switching location.
    – RustyShackleford
    Feb 12 '19 at 23:43
2

Yup, I agree with your read. That is exactly what's going on.

I am fairly impressed that UL even allowed this... I seem to recall seeing a bulletin from NFPA to the effect of "alright UL, it's time to stop approving things that use ground for more than ground". Although it may be still allowed since it is surely a millivolt current.

Regardless, it's rather ingenious. It liberates two wires completely, so for instance if the other switch is a remote spur, you can now use black and white for onward power.

Still won't let the put the sensor at the remote spur, that would require 4 wires (line, load, neutral and sense) plus ground.

Improve this answer
11
  • I still don't understand exactly what's going on. Flipping the mechanical switch changes the blue wire from "open" to "shorted to ground", or vice versa. So if the light is off, flipping the mechanical switch turns it on, and the sensor goes into its "wait for no motion sensed for specified period of time" state. If the light is on, flipping the mechanical switch turns it off; so the sensor needs to ignore sensed motion for some time after that (otherwise we have the amusing situation I meant to describe in my 1st comment on OP).
    – RustyShackleford
    Feb 11 '19 at 19:16
  • So there's a kind of state machine in the smart switch, that knows when the blue wire's connectivity has changed, in either direction.
    – RustyShackleford
    Feb 11 '19 at 19:18
  • Haven't installed it yet. Trying to understand theory of operation; that was meant to be my original question.
    – RustyShackleford
    Feb 11 '19 at 19:24
  • 1
    @RustyShackleford I would bet the engineers thought seriously about UX and chose that which would be least likely to cause WTF returns from customers. At least, if the device has a manual button, I would expect the remote switch to do the same thing as the button.
    – Harper - Reinstate Monica
    Feb 11 '19 at 20:04
  • 1
    @RustyShackleford After much observation and contemplation of this question over the years, I conclude the likeliest answer is 42.
    – Harper - Reinstate Monica
    Feb 11 '19 at 20:12
2

I think that you are getting confused by previous experiences with 3-way switches.

The OS306U is an electronic device, much like a light dimmer.

It has a controller that has several inputs, and the output of the controller causes the HOT (black) to be connected to OUTPUT (red); this turns on the light

The controller accepts inputs from three sources.

  1. the push button on the OS306U
  2. the motion sensor on the OS306U
  3. an external switch

All three of these inputs probably connect the associated input pin to ground (same as the remote switch).

The controller is programmed to operate the output in a specific way, dependent on the state of the inputs.

Improve this answer
1
  • Yep. There's a finite-state machine (FSM) in the controller. Hopefully it'll do what I want it too.
    – RustyShackleford
    Feb 11 '19 at 20:04

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.