[fowlkesfamily]

I installed an occupancy sensor in my laundry room. The light stays on continuously. After verifying that I had all wires correctly spliced and insuring that the ground from the sensor was run to the metal outlet box, I called the technical support (for the product). The technician verified my wiring connections. Then he told me that the reason the sensor would not shut off is because my house ground is not strong enough. I've been reading about house grounding problems, but have not found a good solution. I think I need to improve the ground connection which is located outside near the fuse box, but I do not know how to do that.

 

[oh'mike]

This sounds like you don't know a lot about household wiring.

has it occurred to you that the house ground might be just fine and the outlet you are using is simply a non grounded box?

Have you tested these connections?

What unit are you trying to wire? I know of no motion sensor that relies on a ground wire as a conductor---post a link,--Mike----

 

[Dave632]

I think I'd go looking for a poor/intermittent connection in the neutral wire. I agree with Oh'Mike, except for GFCI (and maybe AFCI), no device can use the ground connection... that's for safety only.

 

[TTW]

Oops, I re-read the OP, it's one of those things that replaces the light switch, not a plug in thing. My bad. Completely useless advice removed.


They would rather blame your wiring then their unit.

These are still handy, get one with a button to test your GFCI's


Oh, and these are handy, cheap, and will check for ground connection.

 

[bobelectric]

You aren't confusing neutral with ground,are you?

 

[tylernt]

You aren't confusing neutral with ground,are you?

My thought exactly. Those things often require a neutral, but a neutral often isn't present in switch loop light switches.

 

[silversport]

I have a leviton that requires a ground but not a neutral.

http://www.leviton.com/OA_HTML/ibcG...XGKtbWDA&label=IBE&appName=IBE&minisite=10251

 

[tylernt]

I have a leviton that requires a ground but not a neutral.

Interesting. That's a neat engineering solution to the problem, though of course not compatible with GFCI (and I wonder if it might even a trip an AFCI -- depends on the mA the relay coil draws).

Well then, the OP may indeed have a grounding issue. Just because the device is in a metal box and the device is bonded to the box, doesn't mean there is conduit all the way back to the panel or that the conduit has good electrical conductivity.

fowlkesfamily, what year was your house built? Can you take a picture of the metal box with the sensor hanging out of it so we can see inside the box and all the wires?

 

[rjniles]

Sounds like "technical support" is using this device on the OP:

Tobacco Smoke Enemas (1750s – 1810s)

The tobacco enema was used to infuse tobacco smoke into a patient’s rectum for various medical purposes, primarily the resuscitation of drowning victims. A rectal tube inserted into the anus was connected to a fumigator and bellows that forced the smoke towards the rectum. The warmth of the smoke was thought to promote respiration, but doubts about the credibility of tobacco enemas led to the popular phrase “blow smoke up one’s ass.”

This has been reintroduced in Washington D.C.

 

[Oso954]

The device operates as described by Technical Support. A high resistance ground might keep it from operating properly. A missing ground would definately effect it. But, it could also be a defective unit.

"except for GFCI (and maybe AFCI), no device can use the ground connection.."
You would be surprised how many devices leak small amounts of current to ground. Most do it for RFI suppression. Many electronic switching devices (wall warts, voltage converters, CFL ballasts, etc) use "a minimal amount of ground leakage".

 

[electures]

Occupancy sensors impose a small amount of current on the equipment grounding conductor. It is how they work and is the reason for the code change requiring a grounded conductor at every switch location. The current imposed by multiple sensors on the grounding systems of large buildings was causing problems with GFI breakers. Most, if not all, sensors will not function without an equipment grounding conductor.

Check you grounding connections.

 

[AllanJ]

"except for GFCI (and maybe AFCI), no device can use the ground connection.."
You would be surprised how many devices leak small amounts of current to ground. Most do it for RFI suppression. Many electronic switching devices (wall warts, voltage converters, CFL ballasts, etc) use "a minimal amount of ground leakage".

A ground fault circuit interrupter does not (should not) use the equipment grounding conductor. It senses the difference between the hot and neutral currents and trips if the difference is more than a few milliamperes. It does not matter where the difference current went if there was a difference.

It is not unusual for electronic equipment to have one or more capacitors between various parts of the circuitry and the chassis (for RFI suppression). The intent is not to put current on the EGC (back to the panel) but some 60 Hz leakage could occur. In addition, capacitance or inductance induced phantom voltage resulting in a minute current flow on the EGC can occur anywhere in the circuit.

 

[fowlkesfamily]

Occupancy sensors impose a small amount of current on the equipment grounding conductor. It is how they work and is the reason for the code change requiring a grounded conductor at every switch location. The current imposed by multiple sensors on the grounding systems of large buildings was causing problems with GFI breakers. Most, if not all, sensors will not function without an equipment grounding conductor.

Check you grounding connections.

---------------------------------------------------------
Thanks electures (and everyone) for your responses so far. "electures" describes the occupancy sensor exactly like the tech support rep did.

Here is some more detail (and a recommendation I'll try tonight at home). The incoming house wire has 3 wires (red, black, and copper). I connected the red and black wires to the occupancy switch. The copper wire I connected to a ground wire from the occupancy switch, then connected both of these wires to the metal box using a pigtail (and a screw). The technician verified that I had the wires connected properly before telling me that my ground wasn't strong enough.

Friend's recomendation: A work friend recommended the following. Connect the ground wire (from the incoming current) to the ground wire on the outlet. Run a (temperary) copper wire, which connects a copper water pipe to the ground connection (from the switch and the incoming current). If the switch turns off (after the room is vacated), we have identified that the current ground is either not present or too weak for the occupancy switch to work.

BTW, I'll post pictures tomorrow if this doesn't work.

 

[electures]

Sounds like a plan. Keep us posted. You could also use an extention cord and use the equipment grounding terminal to trouble shoot your problem.

 

[Speedy Petey]

Occupancy sensors impose a small amount of current on the equipment grounding conductor. It is how they work and is the reason for the code change requiring a grounded conductor at every switch location. The current imposed by multiple sensors on the grounding systems of large buildings was causing problems with GFI breakers. Most, if not all, sensors will not function without an equipment grounding conductor.

Check you grounding connections.

I'm sorry, but I don't buy it.
There should be NO current on the equipment ground, period, and I'm not sure what the EG has to do with the grounded conductor (neutral) being required in switch boxes. I can see your point only in that dumb people were using the EG as a neutral in places where a switch loop existed and an OC switch was install that required a neutral. This creates a tiny amount of circuit current on the EG systems which is NOT good.

The equipment ground and the neutral are two different things. I have never seen an OC switch that required an equipment ground to work properly.

 

[Speedy Petey]

fowlkesfamily, please post the name and model of the switch you have. Also, please tell us exactly what wires and splices exist in the box.

 

[Dave632]

Not necessarily the sensor that the OP has, but follow the link in Post #7 for an example of a neutral-less occupancy sensor.

In fact, it comes in two versions, one where a neutral is present and one which uses the ground wire.

 

[Speedy Petey]

Not necessarily the sensor that the OP has, but follow the link in Post #7 for an example of a neutral-less occupancy sensor.

In fact, it comes in two versions, one where a neutral is present and one which uses the ground wire.

I realize there are two versions, but they cannot use the ground as a neutral. As stated in those instructions: "These versions use the ground connection to
complete the relay circuit while producing a minimal amount of ground leakage."
They are calling it "ground leakage" as opposed to ground current. How they do this is beyond my knowledge, but I would think they'd be taking far too much of a risk imposing actual current on the ground.

I guess this is what electures was referring to.
Thanks.

 

[tylernt]

They are calling it "ground leakage" as opposed to ground current. How they do this is beyond my knowledge, but I would think they'd be taking far too much of a risk imposing actual current on the ground.

Yeah, I'm surprised the UL listed it.

 

[electures]

I'm sorry, but I don't buy it.
There should be NO current on the equipment ground, period, and I'm not sure what the EG has to do with the grounded conductor (neutral) being required in switch boxes. I can see your point only in that dumb people were using the EG as a neutral in places where a switch loop existed and an OC switch was install that required a neutral. This creates a tiny amount of circuit current on the EG systems which is NOT good.

The equipment ground and the neutral are two different things. I have never seen an OC switch that required an equipment ground to work properly.

Copied and pasted from an IAEI article found here;

Grounded Circuit Conductors at Switch Locations

New requirements in the latest edition of the Code requires that where switches control lighting loads such as luminaires, ceiling paddle fans, etc., and are supplied by a general-purpose branch circuit that is grounded, the grounded circuit conductor (generally neutral conductor) for the controlled lighting circuit is required to be provided at the switch location. An exception to this requirement permits the grounded (neutral) conductor to be omitted from the switch enclosure where one of the following conditions applies:
(1) Conductors for switches controlling lighting loads enter the box through a raceway, with the raceway having sufficient cross-sectional area to accommodate adding a grounded conductor at a later date or (2) where cable assemblies such as Type NM cable enter the box through a framing cavity that is open at the top or bottom on the same floor level, or through a wall, floor, or ceiling that is unfinished on one side [see NEC 404.2(C)].
Many electronic lighting control devices, such as occupancy sensors require a standby current to maintain a ready state of detection for the function of these devices. This also allows immediate switching of the load to the “on” condition. These types of devices require standby current when they are in the “off” state as well. These devices typically utilize the grounded conductor for the standby current flow. In some situations, such as where the grounded conductor is provided at the lighting outlet location, a grounded conductor is typically not provided in the switch box for switches controlling lighting loads. When a snap switch is replaced with an occupancy sensor at a later date, installers have typically employed the equipment grounding conductor for the standby current of these control devices. This is not a good practice due to the introduction of circulating current onto the equipment grounding conductor. Occupancy sensors are permitted by UL 773A to have a current of up to 0.5 mA on the equipment grounding conductor. Current flowing at the load when the device is presumed to be off could create a potential risk to installers.