I’ve dug all through this forum and I see this topic come up periodically, but it gets lost in the noise. I’d love to see a direct answer. As I understand it, Z-Wave secondary controllers (e.g. handheld remotes and wall-mounted keypads) can send out a message to activate a scene, which can trigger the primary controller to do things. If I’ve got that right, it’s the solution to many woes: an inexpensive Z-Wave keypad can be a hello/goodbye button by the front door, or you can freely mix and match Z-Wave, Zigbee, and IP switches, or if you’re using those damnable Jasco keypads (by far the least expensive, battery-operated, wall-mounted keypad), not be troubled by the “node IDs above 32” limitation… and best of all, you don’t need to walk all of your controllers around to every device in the house to associate groups.
I’m not a z-wave engineer, so this may not be gospel truth, but it’s to the best of my knowledge after playing around with SmartThings for over a year now and lots of learning from others here.
Secondary controllers, like the one you listed above as well as others, don’t operate quite the way you indicated in your first paragraph. Specifically:
The important part here is that the secondary controllers to NOT send a message for the primary controller to do this. Rather, the secondary controllers send the message directly to the devices in question. The primary controller doesn’t ever really see these commands. (Well, it probably does, but like in a conversation with a group of people, if someone isn’t directly talking to you, you may hear what’s said but not really register it.)
This is (a very crude) map of how things interact:
The bright red are Z-wave devices, the blue are Zigbee devices, and the dark red is the secondary controller. As you can see there is VERY limited communication between the primary (ST Hub) and secondary controllers. Basically this is just initial setup with the primary recognizing that the secondary is on the network, and then passing to it what devices are on the network.
After this setup there is pretty much zero communication between primary and secondary. This is just how the Z-wave protocol was created.
So, long story short, the secondary controllers can’t send a command to the primary and because they can’t do that, the primary can’t act on button presses from the secondary and therefore can’t send commands to other non-zwave devices.
** EXCEPT ** for one device… namely the AEON remotes. These devices can actually be setup as “things” rather than secondary controllers. Because there are things they will communicate directly to the Hub when buttons are pressed and actions can be taken based on those button presses. Unfortunately these are not cheap, but will work for what you’d like them to do.
@chrisb That is what I thought until yesterday. Now I believe this fundamental limitation only applies when a secondary controller is talking directly to a device or group. Scenes (which I have never bothered with in my setup) appear to, at least in some cases, overcome this problem. This is why I would like to hear definitively from someone working on the ST Z-Wave implementation.
I think there’s only one Z-Wave chip, so if these other products can implement this scene detection, it should be possible for SmartThings as well. I do take back what I said about using cheap GE/Jasco keypads, though, as further research indicates they are probably not scene capable (they have a button labeled “scene” but don’t send scene commands). Still, I’d gladly replace all of my keypads if this could work!
I’ll jump in quick here… Depending on the device, scene commands are communicated directly to the primary hub. The Leviton VRCS2-MRZ has two switches which, when paired with a hub, send Scene on and Scene off commands. I was able to create a device type that responds to those commands as if they were standard switch on/off commands, with the benefit that they are sent immediately when the button is pressed and don’t need to be polled for.
Other controllers, like the GE remotes, as you said handle scenes directly, sending out a series of on/off commands to each light switch instead of communicating to the primary controller. With SmartThings, those aren’t very helpful unless all you want is lights on and off. The hub won’t know the devices have changed state until it polls them.
Coming in late here, but it occurs to me that there is a second device category that sends messages to the ST hub: motion detectors.
So it’s pretty klugey, but there are two possibilities:
(may not work) the secondary controller triggers a device state change in the motion detector
(will work, but klugey) the secondary controller turns on a device that moves which is positioned in front of a motion detector. However much movement is needed to trigger the sensor. This could be a rotating fan, a remote control car in a box with the sensor, etc.
So from ST’s point of view, we’re into “When Things Happen.” From the secondary controller’s point of view, we just turned on the device the sensor will detect. But it could work.
I’m not an expert in Z-Wave or Smartthings. I’m just learning both. I do want to set up some scenes and I have purchased 4 Cooper RFWC5 controllers only to find that SmartThings does not really support them. In reality I don’t care if SmartThings can control them or hear them. I would like to be able to set up the scenes in the RFWC5 units. I’m very disappointed that the Z-wave implementation in the SmartThings hub is not complete enough to do this. Has anyone taken a shot a loading scenes from smartthings into a scene controller? Could we write an app to do this?
If I want Smartthings to take action on a button push I believe Smartthings could just monitor the condition of the switches and dimmers affected by the scene. If Smartthings sees the correct combination then it could take further action such as active a Zigbee device.