[Bug] LZW31-SN parameter 50 invalid value

[Hanerd]

Hey Inovelli folks. I wasn’t quite sure where to put this, but figured you’d find it here.

I have about 20 LZW31-SN that are working great, but I realized after getting a bunch of LZW42s that I needed to update my switches for smart bulb mode as they were all still on 1.48

However, when upgrading, I think I found a bug as Parameter 50 is set to 0 which “disables” Central Scene notifications.

1.48 → 1.50 Parameter 50 = 1
1.48 → 1.52 Parameter 50 = 1
1.48 → 1.54 Parameter 50 = 0
1.48 → 1.57 Parameter 50 = 0
1.48 → 1.61 Parameter 50 = 0

Even more strangely, if I update from 1.48 to one of the versions that provides a value to that parameter, a subsequent update to latest production resets parameter 50 to 0 and sets that value to parameter 51.

1.48 → 1.52 (P50 set to 5) → 1.57 Parameter 50 = 0, Parameter 51 = 5

It appears that somewhere between 1.52 and 1.54 the default value went from 1 → 0 for Parameter 50. This, unfortunately, creates a really confusing user experience where the scene controller notifications don’t actually fire even though you have parameter 51 set properly.

Additionally, when doing a two-step upgrade, the values for P50 and P51 become swapped.

I think this is the root cause for this issue and some that I’ve seen pop up on the Homeassistant Discord where central scene notifications are inexplicably not making it to Homeassistant.

It’s simple enough to reset the value, but it appears stock that was produced with a firmware < 1.52 will behave erratically when upgraded. I should also note that the note says default is 7, but I’ve only ever seen a value of 1 returned from the device when upgrading.

[BertABCD1234]

I’ve seen this mentioned before, so I did some searching, and here is the best explanation I could find:

@EricM_Inovelli

My guess is that the storage space for that parameter already has data in it. When you do the firmware upgrade the storage space is not written over (not sure if this should be part of the z-wave firmware process or not). When I have programmed firmware I have had to write the data upon first boot after a firmware update. Otherwise it retains whatever information was there before.

Another thing that makes me think this is that I queried the parameter after a firmware update and it returned a value of 254 or something really strange.

This might have been a different parameter number — firmwre upgrades have added a couple that I can remember — but the outcome in all cases is that the value does not seem to be “initialized,” so might be random or even invalid, causing unexpected results. The best course of action as an end user is to explicitly configure new parameter values to something if the release notes indicate that one was added.

Whether this is something the firmware itself can do is another story; theoretically yes, given what Eric said, but whether they have space on already-jam-packed 500-series they’re using is probably a consideration.