Odd 3-way Wiring

Put a GFCI in an outlet in a wet bar and decided to put the rest of the circuit (largely lighting) on the load side of the circuit interrupter. A light that was 3-way switched on the load side of the GFCI would trip every time either switch was turned on.

It's a 15 A circuit and runs 14/2 throughout. There is no evidence of the use of 14/3 in the 3-way lighting circuit. I'd never seen a 3-way lighting circuit wired without three conductor wiring (14/3). The load appears to be at the end of the cable run with the two switches preceding it in the run.

Anybody know what I'm looking at here and how it might have been wired? Also, would this alone cause the GFCI to trip when the circuit was energized.

Another odd thing about this circuit is that the neutral wire of the

3-way lighting circuit on the second switch has been "gathered" together with neutral wires of another circuit serving a single-pole switch in the same two-gang box as the second switch. Could this be causing the GFCI to trip too?

Is this a dangerous setup and, if so, why is it dangerous?

I'd read some of the other groups and ran across mention of something called a California or coast/coastal/coasted wiring setup for 3-ways. Anybody know what that is? Some of what I read suggested that a "coasted" wiring of a 3-way switch would be the way that two conductor wire (14/2) could be used in a 3-way circuit. But what I read suggested that this was illegal (forbidden by the code) as it switched the neutral wire. I also read that this was "a good way to get someone killed" but I don't know why.

Can anyone point me in the right direction?

Thanks,

LongFisher

Reply to
David W. Walters
Loading thread data ...
040514 1819 - David W. Walters posted:

You don't really need 3 wires between the two 3-way switches. All you do need is 2 wires for the "travellers". At one end, one 3-way switch is fed with the circuit, and on the other end, the other 3-way single wire goes to the light. Of course, the setup must be on the same circuit and use the same neutral.

Reply to
indago

I understand about the travelers. These are the two wires (generally speaking, black and red) of the three conductor 14/3 which are hot if switched into the circuit by the 3-ways. But if 14/2 is used vs. 14/3 then there is no neutral free to return the current from the 3-way nearest the load to the 3-way furthest from the load. So, there's no way to return the current to the service box. The light would likely not work at all when it was switched on as the circuit would not be completed.

The only way this would work is if there was some way to return the current to the box, i.e., if the neutral from the light was gathered with other neutrals from a second circuit. Then when the circuit was switched on the current would have a path back to the box. I think that's what's happening here. And, because the current returns to the box outside the original circuit the GFCI trips because the current in the hot is not the same as that returning via the neutral.

Reply to
David W. Walters

|> You don't really need 3 wires between the two 3-way switches. All you do |> need is 2 wires for the "travellers". At one end, one 3-way switch is fed |> with the circuit, and on the other end, the other 3-way single wire goes to |> the light. Of course, the setup must be on the same circuit and use the |> same neutral. | | I understand about the travelers. These are the two wires (generally | speaking, black and red) of the three conductor 14/3 which are hot if | switched into the circuit by the 3-ways. But if 14/2 is used vs. 14/3 | then there is no neutral free to return the current from the 3-way | nearest the load to the 3-way furthest from the load. So, there's no | way to return the current to the service box. The light would likely | not work at all when it was switched on as the circuit would not be | completed. | | The only way this would work is if there was some way to return the | current to the box, i.e., if the neutral from the light was gathered | with other neutrals from a second circuit. Then when the circuit was | switched on the current would have a path back to the box. I think | that's what's happening here. And, because the current returns to the | box outside the original circuit the GFCI trips because the current in | the hot is not the same as that returning via the neutral.

DO NOT DO THIS AT HOME! Or anywhere else.

This information is for historical curiosity only.

There once was (and sometimes people can still find, perhaps with knob and tube wiring) a kind of three way circuit often known as a "French three way". It was wired by bringing hot and neutral to a three way switch, which then selected between them, feeding one or the other to the common. The common ran single to the lamp. Then from the lamp another single wire ran to a similarly wired switch (hot must be on the same side if powered by an Edison style split phase circuit).

Switch A Switch B light

  1. neutral neutral off
  2. neutral hot on
  3. hot neutral on
  4. hot hot off

Note that in case 4, the lamp terminals were always hot, even though the lamp would not light, hence the hazard. Case 4 is also when you get double voltage at the lamp if the hot wires are from different split phase poles.

Reply to
phil-news-nospam

Hello David, What the fellow did was use the 14/2 w/g between the two three-way switches as travellers as stated. Yes, the neutrals should be the same circuit at the "line" side of the three-way switches, as the "load" side. In other words, the hot and neutral coming into one of the three-way switches, is the same hot and neutral at the switch leg three-way switch. I wish I could draw it out for you, it would be much easier! The reason the GFCI is tripping, is that he must have the white traveller connected at one of the ends to all the neutrals. What would have helped, would to have had the white traveller taped red, but doing residential wiring, he wouldn't use red tape often. Good luck...Blue Crown

Reply to
Blue Crown

There is a method called a California 3 way. Not legal. The wires could switch the neutral and then you would not need the travelers. Not legal. If the fixture is hot all of the time and you become between hot and ground say changing a light bulb on a wet surface, there could be disastrous results. If there were ever an accident and the insurance and or police found out you would be liable.

Reply to
SQLit
040515 1413 - David W. Walters posted:

I had already mentioned that the circuit neutral would have to be used for the lighting. That is what makes the lighting work.

Reply to
indago

Yes, you do, to provide cancellation of the magnetic field around whichever traveller is carrying current. The NEC requires this to prevent inductive heating that the magnetic field might cause.

You can make the circuit work, but it's not safe, and it is what is causing the op's GFCI to trip.

Reply to
ehsjr

You are exactly right. Here's your existing circuit:

GFCI--hot--switch========switch--lite---neutral

where the neutral is wired to a different circuit, or a different point on the same circuit. The neutral to the lite is not connected to the load side of that GFCI, so it trips. The circuit is unsafe - you can get inductive heating due to uncancelled magnetic fields around whichever traveler is carrying current.

Reply to
ehsjr

OK, so what's the fix for this (by the way, there are two more 3-way circuits in the house wired exactly the same way in two hollywood baths)? I guess the only fix would be to replace the 14/2 w/g with

14/3 w/g in all three circuits and wire the switches appropriately for a standard 3-way? In all three switch-legs the load is at the end of the circuit.

Oh, thanks a bunch all of you who responded. You guys respond more quickly than an electrician would and you cost a lot less too.

LongFisher (Dave)

Reply to
David W. Walters

There are some X10 type solutions that do not require a 3 wire branch circuit. You only need to bring the hot and neutral to each location.

Reply to
Greg

Hmmm...could you provide me with a link to such solutions. Although two of the odd 3-way installations can be fixed relatively easily by running 14/3 between the switches, the last is going to be a bear for which to run wire (one switch downstairs in a foyer and the other upstairs on a stair landing requiring the wallboard to be opened in several place in the 25 ft. tall foyer...oh, my aching feet from standing on a ladder for days). So, an alternative which would not require re-wiring would be great.

Reply to
David W. Walters

formatting link
This will get you started but if you back up one page to the smarthome home page you can see what else you can do.

***caution*** X10 can become habit forming!

DO put in some panel based surge protection tho.

Reply to
Greg

I had a similar problem with an existing circuit operating lights on a stairway.

The problem was the switch at the top of the stairs where there was a dual box: one (three way) switch for the stairs, and another for the 2nd floor walkway. Just as your example, it turned out that there were two different circuits in that box. It was here that the neutrals of both circuits were mixed together, and the grounds were secured together too. Of course, the circuits were on opposite sides of neutral (ie: 220 volts from hot to hot).

I first had to separate the neutrals, and then went back to separate the grounds. Since the breaker box was in the garage, I had a great amount of exercise running back & forth between the top of the stairs and the back of the garage. Damn thing took a Saturday afternoon to figure out.

Reply to
Jtiggr

or try

formatting link

radio shack also sells some x.10 stuff.

Reply to
Jtiggr

It does sound like you have a setup like mine, JTIGGR.

I hope you don't mind if I ask a couple of dumb questions of you. I apologize in advance for not knowing some of the terminology you use and beg your forbearance.

What do you mean about the circuits being on opposite sides of neutral?

Also, although I can see that the neutrals need to separated out so the return current via neutral on the light circuit matches the hot current and so the GFCI would then work, why must the grounds be separated out too? I thought the grounds just went back to the panel where they were gathered and connected to a rod sunk into the ground near the panel. There must be something I don't understand about the grounds, huh?

Lastly, you said you took a Saturday to figure it all out. I'll bet. But may I ask what solution you imposed? Did you run 14/3?

Thanks,

Longfisher

Reply to
David W. Walters

The part you are missing is the "main bonding jumper" that connects all of those grounds to the neutral of the utility transformer. THAT is the real path for fault current. The stake in the ground is only to assure that the case of your drill is at the same potential as the basement floor you are sitting on when you use it.

Reply to
Greg

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.