Fibaro FGS223 - Is zwave database entry correct?

Hi all, I have today received an FGS223 and struggling with getting it fully working.
I use 2.5.0 snapshot on OpenHabian.
First my issues:

  1. I get status Updates only, when I add the controller in Association groups!
  2. I do not receive scene number updates (-> Param 28/29 = 15)
  3. I get error **unknown command** when keeping S1 or S2 pressed (dimming feature)
2019-02-05 21:29:58.417 [DEBUG] [WaveSerialHandler$ZWaveReceiveThread] - Receive Message = 01 10 00 04 00 1C 0A 60 0D 02 01 26 04 68 00 FF 00 26 
2019-02-05 21:29:58.420 [DEBUG] [nal.protocol.ZWaveTransactionManager] - NODE 28: Application Command Request (ALIVE:DONE)
2019-02-05 21:29:58.423 [DEBUG] [ng.zwave.internal.protocol.ZWaveNode] - NODE 28: resetResendCount initComplete=true isDead=false
2019-02-05 21:29:58.427 [DEBUG] [ng.zwave.internal.protocol.ZWaveNode] - NODE 28: Decapsulating COMMAND_CLASS_MULTI_CHANNEL
2019-02-05 21:29:58.428 [DEBUG] [nal.protocol.ZWaveTransactionManager] - processReceiveMessage input 0<>128 : Message: class=ApplicationCommandHandler[4], type=Request[0], dest=28, callback=0, payload=00 1C 0A 60 0D 02 01 26 04 68 00 FF 00 
2019-02-05 21:29:58.429 [DEBUG] [ng.zwave.internal.protocol.ZWaveNode] - NODE 28: Incoming command class COMMAND_CLASS_SWITCH_MULTILEVEL, endpoint 2
2019-02-05 21:29:58.433 [DEBUG] [ng.zwave.internal.protocol.ZWaveNode] - NODE 28: SECURITY NOT required on COMMAND_CLASS_SWITCH_MULTILEVEL
2019-02-05 21:29:58.436 [DEBUG] [tocol.commandclass.ZWaveCommandClass] - NODE 28: Received COMMAND_CLASS_SWITCH_MULTILEVEL V0 unknown command 4

My assumption: this has to do with the entry in the zwave database. Why:

  1. When I compare the entry for FGS223 with the entry of the FGS213, which differ only in having 2 resp. 1 output channel, the number of command classes in the “smaller” FGS213 is much higher as in the FGS223. Are they missing and can be added? But:
  2. In my today newly created XML file network_f8886ed0__node_28.xml (37.5 KB) file I see command class MULTI_CHANNEL, but in database I see MULTI_INSTANCE. Does this have a (negative) impact?
  3. I see same version in my XML file (version 4) as in one I found at this topic, which e.g contains the "MULTI_INSTANCE command classes. Why do 2 xml files with same version have different content (keywords, …). How can then a new version be created and correctly assigned?
  4. I wonder why some of you write that FGS223 is working well :slight_smile: Has Fibaro changed the firmware but not the version id?

Is there anybody who can bring light in my night(mare). Have I overseen anything and a new version in the zwave database can be created? Is there anything else I could check?

Many thanks in advance! Siggi.

It is amazingly working well (sorry to say) :grinning:

I just tested to make sure, here is an excerpt from my events.log:

Manual Switch:

2019-02-06 07:34:02.735 [vent.ItemStateChangedEvent] - FibFGS223_1_Sw_1 changed from OFF to ON
2019-02-06 07:34:08.769 [vent.ItemStateChangedEvent] - FibFGS223_1_Sw_1 changed from ON to OFF
2019-02-06 07:34:11.473 [vent.ItemStateChangedEvent] - FibFGS223_1_Sw_1 changed from OFF to ON
2019-02-06 07:34:19.803 [vent.ItemStateChangedEvent] - FibFGS223_1_Sw_1 changed from ON to OFF

Energy:

2019-02-04 20:34:39.292 [vent.ItemStateChangedEvent] - FibFGS223_1_kwh1 changed from 22.97 to 22.98

Scenes:

2019-02-06 07:34:01.343 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 1.0 to 2.0
2019-02-06 07:34:02.906 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 2.0 to 1.0
2019-02-06 07:34:05.953 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 1.0 to 2.2
2019-02-06 07:34:05.982 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 2.2 to 2.1
2019-02-06 07:34:06.744 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 2.1 to 2.2
2019-02-06 07:34:09.332 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 2.1 to 1.3
2019-02-06 07:34:11.529 [vent.ItemStateChangedEvent] - FibFGS223_1_Scenes changed from 1.3 to 1.4

Power:

2019-02-06 07:34:13.879 [vent.ItemStateChangedEvent] - FibFGS223_1_watt1 changed from 0 to 7.5
2019-02-06 07:34:20.045 [vent.ItemStateChangedEvent] - FibFGS223_1_watt1 changed from 7.5 to 0

Same goes for Switch No. 2.

Did you delete the Thing and readd it to pick up all the database changes?

From reading the manual you may have your associations set incorrectly:

The Switch 2 provides the association of five groups:

1st association group – “Lifeline” reports the device status and allows for assigning single device only (main controller by default).
2nd association group – “On/Off (S1)” is assigned to switch connected to the S1 terminal (uses Basic command class).
3rd association group – “Dimmer (S1)” is assigned to switch connected to the S1 terminal (uses Switch Multilevel command class).
4th association group – “On/Off (S2)” is assigned to switch connected to the S2 terminal (uses Basic command class).
5th association group – “Dimmer (S2)” is assigned to switch connected to the S2 terminal (uses Switch Multilevel command class).

You need to set your zwave controller to Lifeline, your first dimmer to Group 3, your second dimmer to Group 5.
Direct associations (Group 2 to 5) are not reflected in openHAB (for example scene numbers) because it is a direct connection between your FGS223 and your dimmer.

Hi @sihui, I first added the thing yesterday, so I assume there were no changes within 3 hours in the database :slight_smile:
But for me the database looks strange …

Hi @sihui, The controller is included in 4 groups (one I deleted to check that then nothing is happening:


I have only one bulb as node 10 added for tests. Nevertheless I see nothing in the logfile when I remove the controller from groups 2-5.
What is you meaning about status of the database entry?

It depends on the date stamp of your zwave binding …

Why? That causes problems. Again:

Sorry, I did not take a look because the device is working fine. :sunglasses:

Just out of curiosity (I normally don’t use associations outside from openHAB) I assigned the following direct associations from the FGS223 to a Fibaro Dimmer FGD212

It is working fine: S2 switches the bulb connected to FGD212 ON and OFF and also is able to dim the bulb.
You need to configure your device in the correct manner.

Hi @sihui, when I remove the controller from ass. groups 2-5 I do get nothing! No status update is then arriving at the controller. First when I added the controller in the groups 2-5 I got events …

Delete the Thing (do not exclude!), readd it, try again.

I removed the FGS223 completely from my network and started inclusion again. And now I see that it is still shown as “Unknown Device” in PaperUI:
Unknown%20Device ).
I viewed the log file with Chris’ log file viewer but there everything looks ok.
Can this “Unknown Device” be the starting point of my issues? Does this mean that the inclusion didn’t find the correct entry in the zwave database and this causes any trouble?

Definitely yes.
Make sure you are on a recent zwave binding:

bundle:list | grep ZW*

on the console

I updated this morning my OH and the version is same for ZWave as it was already yesterday:

openhab> bundle:list | grep ZWave
247 x Active x  80 x 2.5.0.201901280820     x ZWave Binding

BUT: this morning PaperUI showed me no longer “Unknown Device”. I’m not sure why, because terribly 3 things happened together:

  1. I switched my controller to “Full Security” level (instead of “only lock devices”), but was not able to test it directly because I had suddenly to leave
  2. Yesterday I searched for the device, it was shown in the inbox of PaperUI (with “Unknown Device”) and I see in the logfile that this night the nightly heal also made something with this device in the inbox
  3. Today I removed the device in the inbox in PaperUI and used HABmin to discover the device again

–> The device was then also in PaperUI shown with a valid text instead of the “Unknown Device”.
–> After changing the config I now also get scene number changes!

I’m happy that it is now working, but really: I like the whole OpenHAB approach very much, but it costs a lot of time and not each thing starts running without issues. The strange here is: I do not know what really was the important topic to get it run … There is ongoing a bad feeling remaining …

I cannot confirm that, sorry.
You are making some mistakes (see above, for example putting your controller into all association groups), that always will cause problems …

If doing that make sure ALL your devices are supporting the security command class.

@sihui, when you read my first topic you cann read that I didn’t get any reaction from the device without adding the controller into the association groups. It was not a mistake of me, it was a trial to get anything happen. I tested various scenarios and each time I added the controller to the associaion group, I got an update event; removing controller from group → no event. This is why I wrote this strange thing.

When I have done a mistake, I want to identify it, to learn from it. And this is often not possible.

I will do today a new test and remove my device again, set controller back to lower security level (only lock devices). I want to see if this change is the one or not … :slight_smile:

I close this thread, because it seems it is based on the issue of this thread: Cause for not updating an item, even channel reports a status change and is more generic as this one thing FGS223.