[vgb88]

Hi,
I have a home (circa 1940) that has some kludgy wiring. The basement stairs has a 3-way switch; 1 switch at bottom of stairs, 1 switch at top of stairs and two lights in between. I know that the wiring must not be up to current code, because whenever I operate the lights, a big magnetic field is created, which shouldn't happen if wired to any of the currently approved configurations. Anyway, both switches have 3 wires going to them. In testing the light sockets, I found that one light has hot in the center and neutral on the shell. However, the other light is hot on the shell and neutral in the center, and this is true for all on/off switch combinations. This socket is housed in a box that has two wires running to it. I turned off the power and i could clearly see the black wire tied to the neutral screw of the socket, and the white on the hot screw. So, it seemed logical enough for me to reverse the wires, which I did. However, when I turned the power back on and flipped the switch, sparks were flying :huh:, so i immediately turned the switch off. I then reversed the wires back to how they were and the light is working again. I can't understand how flipping 2 wires going to an A/C bulb would cause this to happen and am hoping someone can shed some light on this for me.

 

[RAUL HIDALGO]

if all you did was just flipping the wires nothing like that should happen however you might have nicked the wires when you put them back in place and if the box is grounded you would get something like that.

 

[vgb88]

Thanks for the reply, Raul, but if I nicked the wires, then I would think that the burn mark would reflect that, however, the burn mark is right where the hot wire fastens to the screw and it's apparent that only the end of the wire (under the screw) was overheated.

 

[Koldhammer]

I'm wondering if maybe the wires themselves are reversed? That is, maybe the black wire is nuetral and white wire is hot? What are these wires connected to? Where are these wires connected at the other end of the circuit? Maybe they are connected backwards there.

 

[joed]

Sparks were flying? That indicates a loose connection. If it was a direct short the breaker should have instantly tripped. Where were the sparks? In the light? in the switch?
Can you post a picture of the fixture and connections.

 

[frenchelectrican]

Sound like you ran into Carter system they are tricky to deal and per NEC and CEC code they are not legit for decades ( years ) and if you have Tube et Knob the best answer is remove them and run the proper NM cable.

Merci,
Marc

 

[AllanJ]

Sound like you ran into Carter system they are tricky to deal and per NEC and CEC code they are not legit for decades ( years ) and if you have Tube et Knob the best answer is remove them and run the proper NM cable.

Merci,
Marc

What is the Cartier system?

There is one no-longer-legal setup where raw hot and neutral are connected to the traveler terminals of each of the two switches and the light is connected to the common terminals. In this setup the socket shell will be hot (line voltage hot to ground) some of the time, depending on the switches' positions.

Simply exchanging hot and neutral, the only two wires from a light fixture, will not cause a short circuit unless there was a ground fault in the fixture, for example the metal socket shell touches some other metal part that has a connection to (is bonded to) the metal outlet box that in turn is grounded.

 

[gregzoll]

Carter system

The Carter system is now prohibited.




The Carter system was a method of wiring 3-way switches in the era of early knob-and-tube wiring. This obsolete wiring method has been prohibited by the National Electrical Code since 1923 (actual: see NEC Article 404.2 [3]), even in new knob-and-tube installations which are still permitted under certain circumstances. This wiring system may still be encountered in older "grandfathered" electrical installations.


In the Carter system, the incoming live (energized) and neutral wires were connected to the traveler screws of both 3-way switches, and the lamp was connected between the common screws of the two switches. If both switches were flipped to hot or both were flipped to neutral, the light would remain off; but if they were switched to opposite positions, the light would illuminate. The advantage of this method was that it used just one wire to the light from each switch, having a hot and neutral in both switches.


The major problem with this method is that two of the four switch combinations apply the energized wire to the outer shell of the light socket, presenting a shock hazard. The shell may be energized even with the light off, which poses a risk when changing a bulb. Because of this problem, this method is prohibited in modern building wiring.

http://en.wikipedia.org/wiki/Carter_system#Carter_system

 

[joed]

Even a Carter system should not produce the results of sparks by swapping the wires at the fixture. Something else is going on with this one.

 

[JuzRick]

Sounds as if you can't properly figure it out..if knowing how to troubleshoot the issue currently; it's probably best if you consult with an electrician can go from there..but it does sound as if you have a reverse neutral hot coming in either from the fixture point..or into to the light switch box.

 

[vgb88]

I'm wondering if maybe the wires themselves are reversed? That is, maybe the black wire is nuetral and white wire is hot? What are these wires connected to? Where are these wires connected at the other end of the circuit? Maybe they are connected backwards there.

All I did was swap the wires as I thought they were reversed as well. What I didn't expect was 4th of July - when I get a chance I need to do some detective work and draw a diagram to see what's going on.

Sparks were flying? That indicates a loose connection. If it was a direct short the breaker should have instantly tripped. Where were the sparks? In the light? in the switch?
Can you post a picture of the fixture and connections.

No loose connection. I am concerned that the breaker didn't trip, though I was quick to flip the switch off. When I opened the box, it was apparent that the sparks originated at the screw holding the hot wire.

Sound like you ran into Carter system they are tricky to deal and per NEC and CEC code they are not legit for decades ( years ) and if you have Tube et Knob the best answer is remove them and run the proper NM cable.
No K&T - I need to do more analysis, but I see that the circuit leaves the subpanel via bx and reaches a junction box. The junction box has 3 outputs - two bx cables that go to two other rooms in the house and a Romex that eventually leads to this screwed-up 3-way. However, the romex leads to another junction box which then switches back to bx. The romex tells me that someone was in there at some point.

Merci,
Marc

What is the Cartier system?

There is one no-longer-legal setup where raw hot and neutral are connected to the traveler terminals of each of the two switches and the light is connected to the common terminals. In this setup the socket shell will be hot (line voltage hot to ground) some of the time, depending on the switches' positions.

This is why I don't think it's exactly carter, because the hot and neutral do not vary by the switches' position. The downstairs light socket is always hot on the socket shell regardless of the switches position. The upstairs socket is always hot in the middle and neutral on the shell.

Even a Carter system should not produce the results of sparks by swapping the wires at the fixture. Something else is going on with this one.

I agree - this threw me for a loop.

I appreciate all the replies. When I get some time I am going to trace the wires and draw a diagram that I will share here in the forum.

 

[joed]

Sparks will only fly at a loose connection. If the hot connection was tight it simply would have tipped the breaker on fault. It could be that a two wire that shouldn't have been were slightly touching and started arcing.

 

[fa_f3_20]

As for the magnetic field: how did you detect that? It was pretty common in the K&T days that hot and neutral often were not routed together. But in order to generate a significant field, the circuit would need to be drawing more than what you'd typically get from two light bulbs.

It used to be pretty common in industrial construction that, when light fixtures were installed, they only pulled a hot; the neutral from each fixture simply got attached to the nearest steel framing. Many years ago, I was in charge of setting up a computer lab in a NASA facility. When I positioned a workstation monitor near a particular column in the room, there was significant distortion in the video. The chief engineer came by and I made a remark to him about how much current must be flowing through the wiring in that column. He looked at and said, "There's no wiring in that column!"

 

[vgb88]

That's a pretty amazing story. I used a Bell Model 4080 triaxial gaussmeter. There's no K&T in my house, but some of the wiring is suspect. I own Karl Riley's book on tracing EMF's, which is a great resource IMO. A gaussmeter is actually a good tool to detect wiring problems because modern wiring installed per code should not produce any siginificant magnetic fields as the fields produced by the hot and neutral running side by side cancel each other out, unlike K&T where the conductors are apart from each other (as you pointed out). All of the NEC approved 3-way switch wiring methods you'll find in any wiring book do not produce magnetic fields (I can't recall but there may be one special rare exception). The Chicago and California methods do, of course, as the conductors are separated, and the fields are quite high near the conductors, but fade as you move away from them.