3 way switches

[Jpflex]

You keep saying "code says", please show us the code you are referring to! I'm very sure you won't find such a code.

Perhaps not being able to use stranded would ONLY be code if the manufacturer specifically diss allowed stranded wire for their product under their installation instructions

 

[gar]

230612-1145 EDT

This thread is about 3-way switching. What I am trying to make clear is --- when you run wiring it is essential that current carrying wires in normal operation must have close proximity to each other so that the forward current and return currents for the circuit cancel their magnetic fields.

..

 

[Jpflex]

230612-1145 EDT

This thread is about 3-way switching. What I am trying to make clear is --- when you run wiring it is essential that current carrying wires in normal operation must have close proximity to each other so that the forward current and return currents for the circuit cancel their magnetic fields.

..

However if neutral is not used or hooked up inside box, then how would you have canceling current magnetic fields for low impedance on EGC?

Perhaps this is achieved from line input to switch going one direction and when switched line is routed to returns in opposite direction to cancel magnetic fields?

 

[winnie]

However if neutral is not used or hooked up inside box, then how would you have canceling current magnetic fields for low impedance on EGC?

Perhaps this is achieved from line input to switch going one direction and when switched line is routed to returns in opposite direction to cancel magnetic fields?

Exactly this.

The term of art is a 'switch loop'. Power is brought to a light fixture with a 2 conductor cable having grounded and ungrounded conductor. A cable runs from the fixture location to the switch location. The unswitched, ungrounded conductor ('hot') is brought from the fixture location to the switch location. The _switched_ conductor ('switched hot') returns from switch to fixture. The current in these two wires balances perfectly (unless there is a fault).

There is no need for the grounded conductor to be brought to the switch for proper current balance.

You can have a three-way 'switch loop', again where the only conductors needed _ungrounded_ conductors (either constant on, or switched). There is no need for an ungrounded conductor in a 'switch loop' for proper current balance.

NEC requires a grounded conductor be brought to the switch location because of the potential future need for power at the switch location.

(NB: @gar is quite correct about the need to balance currents, and there is a history of old style 3-way light installs that don't maintain current balance. Look up 'carter 3 way' for an example of the wrong way to do things!)

-Jon

 

[Frank DuVal]

I do not understand what most of you have been saying.

Don't use the words hot, neutral, and .EGC.

Funny, I use a hot, a neutral and an EGC on 3 way circuits all the time. How else does the lamp light? Sure there are also travelers between switches.

 

[Jpflex]

Funny, I use a hot, a neutral and an EGC on 3 way circuits all the time. How else does the lamp light? Sure there are also travelers between switches.

Do you prefer to put traveler wires on the silver screws on 3 way switch on the same side for each traveler wire? Such as right side to right side and left side screw of one switch to left silver screw on other switch?

 

[Jpflex]

I was never taught or shown how to install these but by doing a continuity check I figured line goes to black screw on first switch and black from second switch to load power

Travelers being silver screws with two jumpers to screws on same side of two switches

 

[Joethemechanic]

So do you install the switches so either both handles up or both handles down is on? Or do you connect them so one handle up and one handle down is on?

 

[roger]

So do you install the switches so either both handles up or both handles down is on? Or do you connect them so one handle up and one handle down is on?

Either way and then I cover them with locking covers to make sure nobody ever screws them up.

 

[LarryFine]

I came across a code that wires serving no other purpose must be bonded to box. Therefore if white wire for future neutral ran with other circuits is not connected to anything then wouldn’t current code require it to be bonded to metal box?

This is a bad idea for at least two reasons, if the box grounded or is not grounded:

Possibly energizing the boxes if someone uses the white without checking it first.

Possibly believing it to be properly connected if someone does check the white.

 

[LarryFine]

So do you install the switches so either both handles up or both handles down is on? Or do you connect them so one handle up and one handle down is on?

I don't bother checking, unless it's brought up. There are more important things to worry about.

 

[Jpflex]

Either way and then I cover them with locking covers to make sure nobody ever screws them up.

If you do it how I do, you’ll get both handles up for on and both handles down for off.

 

[Jpflex]

So do you install the switches so either both handles up or both handles down is on? Or do you connect them so one handle up and one handle down is on?

If you do it how I do, you’ll get both handles up for on and both handles down for off.

Both travelers on same side of switch. I sent this message to someone else by mistake.

Also are you a mechanic, auto mechanic?

 

[Little Bill]

Perhaps not being able to use stranded would ONLY be code if the manufacturer specifically diss allowed stranded wire for their product under their installation instructions

Nice dodge!
You were saying it was code to bond a spare wire to the box. I/we asked for a code reference.

 

[tortuga]

So you had devices which _intentionally_ used the EGC as a current carrying conductor, but for an extremely small current in the range of what the UL previously permitted as unintentional leakage.

It seems like there should be code rule that prohibits using a EGC also as a circuit conductor (current carrying)

 

[roger]

If you do it how I do, you’ll get both handles up for on and both handles down for off.

Both travelers on same side of switch. I sent this message to someone else by mistake.

Also are you a mechanic, auto mechanic?

That would mean you run to both locations every time you use the switches.

 

[synchro]

If you do it how I do, you’ll get both handles up for on and both handles down for off.

That would mean you run to both locations every time you use the switches.

Yes, he would need to be one of the travelers between the 3-way switches.

 

[Frank DuVal]

I figured line goes to black screw on first switch and black from second switch to load power

And these screws are marked COMMON. You might need a magnifying glass to read them though.

If you do it how I do, you’ll get both handles up for on and both handles down for off.

No No No! If you wire the two 3 way switches for both to be down and the light off, when both handles are up the light is also off! The light will be ON when one switch is up and the other switch is down, does not matter which switch is up or down.

 

[Frank DuVal]

Both travelers on same side of switch.

Hard to do when the traveler screws are on opposite sides of the switch. Most every 3 way switch I've wired has them this way.

 

[winnie]

It seems like there should be code rule that prohibits using a EGC also as a circuit conductor (current carrying)

AFAIK, the NEC _never_ allowed the EGC as an intentional circuit conductor. The closest was to permit the neutral of certain circuits to be used for bonding equipment on those circuits (range and dryer circuits in the past).

But it is an unavoidable aspect of the physics of the situation that there will be _some_ current on the EGC because of leakage and capacitive coupling. No insulation is perfect and even if it were there is still going to be capacitive coupling. Part of the UL standard for switches was a maximum allowed leakage current.

NEC permits listed switches with their maximum allowed leakage current, which means the NEC permits a certain amount of leakage current on the EGC.

The question is how to treat _intentional_ pseudo-leakage. I say 'pseudo' because it is such low level current that it is lower then the maximum allowed leakage, but it isn't really leakage because it is intentional.

The CMP decided nope, we are going to require a grounded conductor be available or easily brought to each switch location, so that intentional current on the EGC isn't needed.

-Jon