Thanks for the pointer Julian, What Ewan is trying to achieve is pretty much exactly what I'm wanting to do too. However my main focus is on being advised (within HC2) of input/zone triggers to my Comfort system (rather than mode) so I am not sure that the level of an intermediary dimmer could dependably reflect those if they are changing frequently or several at the same instant.
I will look into this further and that thread is bang on topic rather than hijacking this one anymore. Maybe I should just return the Z-Wave UCM though and pursue a different solution e.g. a Fibaro plugin , although that's another story.
Cytech... it's a shame we can't have a more useful bi-directional integration to a Z-Wave master controller. I understand the issues are complex but I too would be happy with a non approved firmware - to talk to my non approved Fibaro system.
I've responded in the other thread too but just wanted to reiterate that even if polling was working it wouldn't be a useful solution to me as that would swamp Z-Wave bandwidth impacting overall performance and also be prone to missing short zone trip triggers.
Polling is bad
I was looking for a near 100% assurance that any and all virtual node changes would be reported to the master controller, never missed, and pretty much instantly. I accept the best that I could expect is as Z-Wave security sensors operate already so Z-Wave may be wholly unsuitable for this need. I expected Comfort Z-Wave virtual zones to appear like security sensors to the primary controller and not need polling. I effectively wanted to use my Comfort system as a multi channel Z-Wave I/O device with lots of inputs and outputs.
I tried the Z-Wave UCM path because it was a quick 'out of the box' solution rather than using an Ethernet or serial UCM and some coding to implement the associated protocol.
I am not asking Cytech to put in a fix for a Fibaro bug - but implementing status in a realtime event way as a sensor must be a feature many people would find useful.
Alternatively and if my understanding is correct, if some third party manufacturer offered an affordable multi endpoint (say 16+ relay) Z-Wave device that could be associated with these virtual nodes that might work too ? I'm not aware of such a device though and it's really how I wanted my Comfort system to appear to my primary controller.
Last edited on Friday Jul 3rd, 2015 02:24 pm by xAPPO
Yes agreed that polling is bad. This is the big problem with Zwave, that the devices do not report their changes.
Zwave sensors like PIRs normally work by association to lights so when the PIR triggers the lights turn on. Zwave works best for such simple applications.
If HC2 can create virtual nodes like the UCM then zone or counter changes would update these nodes, without needing physical zwave switches or nodes. If not and polling is not working then it is a problem.
What you are asking for, ie comfort zones to report to zwave controller is more complex that it seems.
if the controller is not to poll the node, then assume that each zone response can be programmed to send a zwave command to the controller node ID, but from diffferent virtual node ids to represent Zone numbers (this is not available now)
We do not know if a zwave controller is able to accept such a command/status - in fact most would not especially the handheld remotes. Then assume Fibaro does accept such commands, because it has been designed to work with its own range of sensors in a nonstandard way. Does it want a Status command or a SET command in Zwave terminology? Or perhaps it accepts such special messages only from Fibaro devices?
Even if we were to attempt to do this, we would probably end up trying to solve Fibaro problems which is not an entertaining prospect.