Earth as ground in single wire swith outlet, what is the code?
tonyp12
Posts: 1,951
I would say around 50% there is no Neutral inside the switch outlet as it's a drop down for the hot wire only.
But as switches are getting smarter they need constant small amount of power, does the North American code allow a steady current of 5mA etc to Earth?
If you do, you have to for example use a SR10 type device The code says?:
http://www.supertex.com/pdf/datasheets/SR10.pdf
Inherent short circuit protection is afforded by the reactance of a series capacitor, which limits current even with a dead short on the output
But as switches are getting smarter they need constant small amount of power, does the North American code allow a steady current of 5mA etc to Earth?
If you do, you have to for example use a SR10 type device The code says?:
http://www.supertex.com/pdf/datasheets/SR10.pdf
Inherent short circuit protection is afforded by the reactance of a series capacitor, which limits current even with a dead short on the output
481 x 243 - 10K
Comments
-Phil
Leviton and Intermatic have some type of single wire devices where neutral is not required.
And they should be that incandescent blub is not needed where you feed some mA through to use that path as ground but not enough to light it up.
ftp://ftp.leviton.com/LevitonFTP-Public/LES/Library/TechnicalArticles/Technical%20Article%20-%20OSSMD%20Neutral%20versus%20Non-Neutral.PDF
I'm not sure if draining 5 mA through the safety ground would affect any GFI devices in various parts of the house , but more and more distribution panels come equipped with GFI circuit breakers and these may get affected.
In houses with ground-fault interrupters, a 20mA current is guarranteed to trip the interrupter.
It may trigger on lower values, too...
A 5mA current 'on the loose' in a house with a faulty ground may kill somebody.
But if I'm creating a commercial product where I want it to work with all types of wiring, how do they do it?
Earth is getting spurts of power now and then from surge protectors in your devices etc
Or desktop power supply that use earth as reference with use of capacitor and a 1meg bleeder resistor.
So there must some rule on the length and amount of current that is acceptable?
The problem is that if the main electrical system ground and neutral are lost/disconnected (floating ground), then ALL the grounds in the house will become energized.
And that is the safety built in to the design of modern electrical systems. If there is a malfunction somewhere (wire breaks or comes loose), there is backup safety built in. Note it is not uncommon for the main system ground to become disconnected. Lawnmower cuts wire, homeowner replaces metal water pipe with plastic, etc. And with a loose neutral situation, those were not properly installed with anti-oxidation goop nor the connection torqued with a torque wrench to the specified inch pounds of the panel. They go through a heating/cooling cycle as more/less power is used, then the wire works itself loose.
High amperage connections are in a class by themselves so far as connections are concerned.
Search google.com for...
electrical loose neutral
Tony asks: "But if I'm creating a commercial product where I want it to work with all types of wiring, how do they do it?"
This is just my opinion since I'm not a licensed Electrician...
The product would have to include in the installation instructions the requirements for a safe, proper installation per Code. If the device requires a connection to the system neutral it must be stated without offering a "workaround" which could violate Code.
Also, it would probably have to be certified by Underwriters Laboratories. I'm not 100% positive NEC requires UL approval, but every professional electrician I know will not install non-UL approved devices.
I'm certain that UL will not approve a device if there are inadequate warnings against improper connections or installation. A UL certification is only valid when the device is properly installed and connected according to Codes applicable at the time of installation.
Certification would be needed UL or the less costly ETL, but for now my curiosity is what to do with non-neutral.
1: Use a rechargeable battery that will charge from the power passing thru when engaged, but if switch is not used for 3months maybe it will never turn back on.
though could design it with a low leakage PNP, so when battery is criticaly low it automatically turns switch back on to allow it to recharge.
2: Find out if using SR10 with it's series cap + another series cap to Earth-gnd is allowed by code.
if not, stickers save you once again:
Warning: High leakage current. Connect to earth first.
But signs like this make you look more dangerous from a marketing perspective.
In that regard, you would just want to have a proper connection diagram.
edit:
forgot another sticker if > 3.5 mA:
Warning: Not for use in residential areas
might want to reconsider, and only show the proper wiring diagram with the neutral:
With that being said, your product should be designed with the new code in mind (which will be required by UL anyhow) and end-users should understand that they may need to retrofit to meet code.
("Neutral wire required")...
http://www.lutron.com/TechnicalDocumentLibrary/032442.pdf
With the use of a large capacitor and a 5% duty for actual work, you could use spurts of 10mA for mosfets/triacs etc that needs it.
Some ground wires are not connected to anything! This can happen with new room additions to an old house. They use new wire and connect the grounds at the outlets/switches, but don't connect the grounds at the other end.
Also there may be an electric motor or appliance "leaking electricity" to ground. Or some other malfunction which places a voltage on the ground wiring. Those voltages can be high enough to shock people.
So basically you can't rely on a ground wire being connected or being at ground potential. With "do-it-yourself electrical work", no telling what they might have done. The following picture is my all time favorite...
http://img2-1.timeinc.net/toh/images/galleries/0607_homenightmares/cigar.jpg
Yikes!
This thread sent me off doing research on this topic of using earth wires to carry current. Kind of scary people are selling commercial products that do this. Maybe it isn't so bad in the US as the voltage is 110V but in all the parts of the world where volts are 220 to 240V this would not be an option. Mind you, 110V can still kill. I gather this is to save pulling a neutral wire through the wall.
Maybe you can get away with this where there is no earth leakage circuit breaker http://en.wikipedia.org/wiki/Earth_leakage_circuit_breaker, ie older houses. But then there are posts on discussion threads where people have rewired their house and added a leakage breaker, which then trips all the time, and they have traced it back and found the source of the leak. So that is a good thing. I've discarded appliances when they constantly tripped on the earth leakage - kettles, bench ovens, usually things with heating elements in them which degrade over time. I have no desire to get zapped or worse.
So, 10mA briefly on a ELCB would be getting close to tripping it. And what if someone who liked the wall switch so much decided to install them on every light switch? Perhaps 20 in a house? 500uA on 20 switches is 10mA - enough to kill.
The earth wire is there for safety. It is not supposed to carry any current, and if it is carrying current then every earthed device in the house is now live. The metal case of the stove, the fridge, the washing machine. Ok, maybe not at full mains volts, but consider a room like the laundry 20 metres from the switchboard, and the lightswitch is 2 metres from the washing machine. The earth wire is a low value resistor and this is a simple voltage divider - 2/(20+2) so you will have 100V on the outside of the washing machine. And that also assumes the earth is a perfect earth. It may not be - in reality it is a metal stake in the ground and if it is summer and the ground is drying out, that earth may have a resistance of a few ohms. So this contributes also to voltage rise on the earth http://www.weschler.com/_upload/sitepdfs/techref/gettingdowntoearth.pdf
Thinking of the time needed to replace lightswitches with smart swtiches, surely it is worth a few more minutes to pull down a neutral wire?
NEC probably have rules on that to follow.
And with a Supercap, would its ground not be the return path before the Earth?
A GFCI does not measure Earth current but the difference between Hot and Neutral, so it would not trip from spurts of 10mA on Earth coming from the cap.
As the idea is to use a constant 0.5mA and storing it for other (up to 10mA) capacitive loads in the circuit.
I've done some electrical work in the last few years to help friends. Enough time passes that I always get the book, work through it, update old habits, then do the work, and I never do work that won't be inspected, so that I know it's gonna pass and I didn't kill somebody.
The things I've seen in older "jalopy" style fix-er-upper type houses! Nothing quite that bad, but sometimes pretty bad. Unconnected grounds are chronic in my limited experience. They have been a factor in every single job I've done. Another one is loose wire nuts, or where somebody used tape, decay of the tape, due to it being cheap, or the wrong tape (duct tape anyone? Yes, I've seen it and cleaned it up.) and or poor mechanical bonding of the wires is chronic too.
Re: Options
I really like the battery one the best. Charge it with switch on, leave the grounding mess alone. Really safe. Of course, I'm really conservative and somewhat anal about it too, for the same reasons Dr_A mentioned.
How big of a deal is it to provide a warning blinker and or some kind of manual "charge it" hole / function on the switch? A really tiny hole could allow a user to depress a switch to juice the smart switch in the "didn't use it for a long time" failure case, and or a blinky that comes on just prior to that failure case would be very reasonable things people would understand and appreciate. I don't think they would even be exceptions, done nicely.
Now that I think about it, a blinky light placed where it would be seen through some plastic, or light gets out through the cracks, maybe do it as a fade in and out, to not be annoying, would be a very nice value add. People are used to seeing status indicators now. If their new switch was blinking or lighting up in some way some day in the future, they would not be surprised at all, and likely turn it on to check, in which case the blinky goes away. Done, next. Most of them will just realize it's gotta be on once in a while to continue working.
Another idea for the "charge it" case would be a sensor that normally is touch, but one could press in on it to trigger the charge too. Just a thought. The documentation could be as simple as, "when switch fails, press in and hold for 10 seconds...." That provides a simple, mechanical reset on failure option that would be completely unobtrusive.
and instead use Dead man's switch circuit for the 1% chance of dead battery.
Say I want to charge a battery from pass through current, should I use a ?:
Common choke http://www.mouser.com/ds/2/212/KEM_LF0011_SC-G-GS-276376.pdf
or
Coupled Inductor http://www.mouser.com/ds/2/54/SRF0703-39521.pdf
The bonus is that sensitive electronics in LED lamps will get some added protection.
I'm trying to think of a solution to the problem. It is quite a challenge.
The battery might work - charge when the device is on. It would have to be a sealed battery though - you can't have the user pulling a battery out of a live battery holder to replace it with a new one.
But I think I am missing something with this
Where can you get 5V from? How can you get 5V if the only reference you have is earth and you are not allowed to feed current into earth? A voltage divider is still feeding current into earth.
Ok, you are worried about resistors burning out to a direct path. I'm worried about the resistors even when they are working fine.
Simple scenario. I come into your house and I connect your washing machine case to the hot wire with a 150K resistor. Correct, this does not trip the GFCI because it is under the threshold. Correct, the GFCI measures the difference between hot and neutral, so the device does not trip. But you still have current travelling through that resistor into the washing machine case. Through dry skin, standing in shoes on a tiled floor, I suspect you would not feel anything.
But what if you have just finished some hand washing and your skin is wet? And what if you are standing on a wet floor in bare feet? And what if the water is going down the drain into a metal pipe which then goes out into the street into a large metal pipe which leaks slightly, and that metal pipework happens to have a better connection to earth than your earth stake hammered into some dry soil next to the house? The path of least resistance may well be through the heart.
Ok, says the devil's advocate, if all that were true and the earth through a drainhole were of a lower resistance than your house earth, and there happened to be a 150k resistance between earth and hot, yes, you could still get a shock. But it is unlikely to happen - the resistance is likely to be lower, in which case the GFCI trips, or higher, in which case there is no problem, or variable (say some water got into the washing machine electrics and was slowly electrolysing) in which case the GFCI will trip. Then you go investigate why it tripped.
So the challenge is to design a circuit where no current ever flows to earth, and the only wire you have is the hot wire, and nothing is connected to the earth wire in any way.
A battery should work. You could use the new low self discharge NiMH batteries. Charge it at a trickle charge rate when the device is on. And when it is off, use a circuit that draws miniscule current - maybe a picaxe or even a propeller in super low power mode.
Not my preferred circuit though. So much simpler to add a neutral wire.