Association Groups won't save for zwave light panel

Hi friends,
I started organizing my ZWAVE light panels into Association Groups so that a few light switches on different panels will turn the same lights on/off.
I managed to set five of my panels to work in Association Groups, but one panel fails to save the Association Groups info I’m trying to calibrate.
Any idea why this would happen for this particular panel? I have similar panels from this brand and they saved properly.
Any idea how to fix this?
This is the panel I’m trying to calibrate:
Lumi LM-S4ZW (C-L) Z-Wave Switch with 2 on/off light switches and 2 up/down blinds switches
Thanks for your help.
Hope you are all well and healthy.

I think I found a reason why the particular zwave panel is not saving the association groups info.
For documentation purposes and to help other users I’ll share my theory.
Using openhab’s info data for my panels, I found the problematic panel is running zwave_version: 2.40, whereas my other 3 properly working panels are running zwave_version: 2.57.
Could it be that zwave_version 2.40 does not support association groups?

No - association groups are not linked to the zwave version. This is the protocol level, where association groups are a command class (so at the application layer). If the device supports associations, then the protocol version is not relevant.

2 Likes

Thank for the info, Chris.
Do you think the particular panel is faulty?
OpenHAB’s Thing settings is showing the option to add association groups for this panel. The problem is the association groups I add are being reset after I press save.

I would be surprised if it was faulty. It seems like the device is working fine in general, and associations are a software function of the device.

I would suggest to reinitialise the device in the first instance (there is an option to do this in the device configuration). If that doesn’t work, then possibly getting a debug log during initialisation might help understand the issue.

1 Like

I’ll try. Thanks.

Here is a similar issue that was recently resolved by excluding and reincluding after cleaning up ghost nodes. Do you see any of them?

1 Like