Migration from Fibaro HC2 to HAOS

Thanks again for tips.
So, my plan is:

  • purchase Zooz 800 (as it’s the newest compatible stick - to avoid technological debt :laughing:) and plug in to my PVE server for doing pass-through
  • install in HAOS zwave js ui (HAOS already running)
  • do I need to install separate instance of MQTT broker?
  • start device migration.
    The biggest tasks will be:
  • simulation scene (triggered by alarm system when is fully armed), simulates activities/lights on/off at home after sunset - I need to replace Satel comm. module to ETHM-1 Plus
  • garden watering scene - takes into account Netatmo Rain sensor, forecast and controls sprinkles in garden
  • recuperation unit control (StorkAir WHR) scene - controls unit speed in function of CO2 content read by Netatmo.

Above scenes are fully in LUA, do I need to transform then to yaml config or maybe there are some templates or different solutions available?

MQTT: you can use your present matt server, or use the HA mqtt add-on.
I started with my existing mqtt server on a Raspberry Pi, but when moving everything else from my HC2 to HA running on an Intel-NUC, I moved mqtt (and node-red, and alarm sip call-out, maria-dbs, etc) all to the one NUC, freeing up 2 raspberries in the process.

For the various scenes, do some searching on this forum. To replace the alarm system I had on the HC2, I could find good examples on the forum. I have no experience with watering system (still manual labor here :grinning:), but I do have NetAtmo integration running on HA as well, you can use sensor data in your automations.

On Automations, I suggest you start automations in the UI interface, you can switch in the UI to see the underlying yaml code (and switch back again). This gives a good start, and is easier than starting with yaml.

Just to confirm, in order to use HA mqtt I need to keep disabled “Disable MQTT Gateway” in zwave js ui settings? Do I need to change anything there or just use default settings in MQTT, Gateway and HA tabs?

Associations - I’m a bit surprised that I can’t do it directly in HA(??). Is zwave js ui the only way to make associations?

I use Aeotec ZW140 nano switches for handling irrigation valves. In addition I have Fibaro switch as master power control to cut off power completely after watering is done and switch power back on minutes before next watering. BUT… unless I manually do routes rebuild, Aeotec valves remain dead. It wasn’t an issue with Fibaro HC2 controller. Is there anything to tweak in zwavejs to let it rebuild routes much quicker in automatic mode?

EDIT:
Or maybe it has something to do with the fact that Aeotec switches use zwave+ and surrounding Fibaro devices not? I guess with HC2 in control it wasn’t the case…

EDIT2:
Fibaro dimmer FGD211 doesn’t have any lifeline/controller updates association but its status is updated in HA after being locally/manually operated. How come?


Any other Fibaro switch has controller updates group automatically configured

and Aeotec switches have lifeline group…

Keep MQTT Gateway disabled, just use default settings.

The ZWaveJS UI integration is the part of HA controlling all zwave aspects, so it’s logical that you setup associations from here.

The Aeotec switch problem is not clear how this is setup, is the Fibaro switch not switching? Why do you need to rebuild routes? Only reason I did a rebuild was after relocating my HA system including the zwave controller, messed up my network for a day before new routes were stable.

Fibaro switch cuts completely power off to Aeotec switches (and controlled valves) - they become dead nodes. Main reason is an extra measure to avoid uncontrolled flooding.
Like I said, with HC2, after restoring power, Aeotec switches were back online within a less than minute. With HA/zwave js ui, it’s not the case. I need to manually run rebuild routes. How to improve this behaviour?

If I understand correctly, the Fibaro switch works ok, but the Aeotec remain “dead” in HA.

I’m not familiar with the Aeotec switch (if it’s battery operated, you have to wait for the next polling period), but as you switch this with the Fibaro switch I assume it is direct powered, but even a direct powered switch takes a while before it is seen again in the network.

What you can do, after the Fibaro switch has powered the Aeotec, do a soft restart (reload) of zwavejs-ui , much simpler and less time-consuming than a rebuild of the entire network. It will interrogate all devices, the Aeotec should then seen as “alive”.

For an automation how to reload zwavejs-ui, see Automate ZWave JS Soft Restart - #5 by kramttocs

Thanks very much @wimjanse
One more thing every restart of HA or system reboot, Home Assistant Core Integration: Generic turn off kicks in and turns off that Fibaro (master) switch - if left at “on” state.
I have only a small part of my system migrated - still testing possibilities; but as far as I can see, it affects only that particular switch. What the hell is it and why does it?

@listhor

I have 40 Fibaro switches and dimmers, after a HA reboot those remain in the state as before the reboot.

I know that you can program a Fibaro switch to go to off mode after a power failure, so I suggest that you look into the settings of that particular switch and see whether that is the case. Alternatively, remove the switch from the network, do a full reset on the switch and re-add. Hope this helps.

I have all of them: Fibaro master switch and Aeotec switches (all of them are for irrigation valves control) set on purpose to remain at OFF state upon power recovery. I had already power outage during watering period and after having it back, all valves remained open - and garden was flooded.
But restarting HA doesn’t cut the power off and this issue affects 1 out of 10 already migrated Fibaro switches. And I can’t find the reason why…

I understand, strange that only one switch has this issue. Only suggestion remains a full reset of that switch and see if that fixes the problem.

I am having a tough time with exclusions - no luck with the triple press at the relevant momentary wall button. Dreading the idea of getting into the ceiling to find the pucks and press the B button. Any other tips from your project for successful exclusion?

I faced same problem with some of the Fibaro switches.

Solved this by bringing the HC2 close to the device I needed to exclude (extension cord for power and long network cable).

In this manner, device was directly seen by the HC2, no “hops” over other devices, prevented me from removing lamps and opening my ceiling.

Clever strategy. I’ll give that a go, thank you

I had a zwave antenna on long cable…
Older generation switches might also require quick triple click (in 1s) to initiate the process.

I thought Fibaro switches required 5 or 6 clicks to put them in inclusion mode, not 3. Also, if you have switched the Left/Right switches via the configuration, then you may need to click the other one. (I ran into this at some point, but I don’t remember which way it worked.)
-David

It’s 3 clicks, and for the older devices 3 clicks within 1 second. So it can take a few attempts before you succeed with the triple click for insert/exclude mode.

Hi
As far as I know,
ANY Z-wave controller can exclude ANY Z-Wave device.
It doesn’t matter if the device was initially a member of a different master.
Exclusion should work with every Z-wave master with the correct frequency.
Cheers

1 Like

I don’t think this is correct, the controller has to send the unique security keys for the particular network.

Otherwise it would be quite simple to destroy my neighbors’ ZWave network.

No, @juergen3 is correct. You can’t destroy your neighbor’s Z-wave network because his device has to participate in the exclusion process (usually by pressing a button) so physical access is required. Also, encryption keys are only used for S0 or S2 included devices. You can include devices unencrypted.

Correct :slight_smile:

1 Like