The problem:
I am updating a 4-way switching circuit that has 14-2 (black, white, and copper) run to the 4-way switch in the middle. The boxes containing the switch for the line power and load contain a neural, but there is no neutral for the auxiliary switch replacing the 4-way switch. This is a likely scenario in older homes where electricians saved on copper by using 2-wire instead of 3-wire romex to a 4-way switch as long as the neutral completed the circuit elsewhere.
The circuit with a Master (GE 12724) and Auxiliary (GE 12723) switch works if the old 4-way switch is bypassed by tying the load (black) and traveler (white) wires together in place of the 4-way switch. But, this solution changes eliminates a switch in a prime location.
How can I functionally replace the switch in the middle of the circuit without fishing a new wire?
Prior topics:
I tried the forums and found a number of discussions that I thought would be relevant, but in the end didn’t solve my problem:
-
Old wiring and Z-wave switch: No neutral: very long discussion going back to 2013, couldn’t find situation similar to what I encountered.
-
4-way switch with aux missing the neutral - This entry seemed like it would be the answer, but it was actually a case of ms-identification of where the line power entered in the middle of the circuit where the 4-way switch was located.
##Potential solutions:
These discussions lead me to 2 alternative solutions that I thought may work without pulling a new wire to the location of the 4-way switch without a neutral.
###Relocate Master switch and add a physical switch with programmatic control
Connecting the 2-wire from the 4-way switch to the line voltage and neutral where the Master is currently located (where power enters the circuit) would get power and a neutral to the location of the 4-way switch. However, this eliminates the traveler, so secondary control from an un-wired switch (i.e. a switch that is not wired into the same circuit) or scene controller would be needed to replace the current master switch. This would require the un-wired switch/control to programmatically control the lights through software running on the SmartThings hub.
This seems like the optimal solution, but brings up a few different questions:
A) What z-wave scene controllers are compatible with SmartThings that could take the place of a switch on the wall? Scene controllers like the GE 45631 or [Enerwave ZWN-SC7] (http://www.thesmartesthouse.com/products/z-wave-7-button-scene-controller-zwn-sc-7-w) seem like they should work, but are there others?
- (+) A controller switch would allow programmatic control of multiple connected lights and/or scene toggling.
- (-) Not widely available, and not on list of officially supported devices
B) Do connected switches work without a load? If they do then any connected switch could provide the toggling command signal to SmartThings.
- (+) Switches are widely available and prices have come down significantly
- (-) Lights are likely to be slow to respond because of indirect control via app.
###Replace all switches in circuit with switches that do not require a neural.
Lutron Caseta switches do not appear to need a neutral wire, and they have small battery-powered remotes that can be mounted in place of a standard switch. Although Wink compatible, there is no clear indication if the current version is SmartThings compatible with use of the Zigbee standard, or if a Caseta hub and IFTTT are necessary to make Caseta and SmartThings communicate.
- (+) Wired control of lights would leave least delay to controlling lights after button push.
- (-) Added cost and potential inter-platform incompatibilities.
Does anyone in the community have experience with one of these solutions?
Is there another solution?
I’m sure @JDRoberts will have some insights.