Looks like parameter51 was introduced in firmware 1.36 and I was running 1.34. When I use the Basic Z-Wave Tool, the final log line it prints for a configuration report query across all parameters ends at parameter31, as I would expect from 1.34:
2020-08-16 08:01:18.446 pm info ConfigurationReport- parameterNumber:31, size:1, value:1
It seems that either the driver or the Hubitat device interface is clever enough to not show me a parameter not yet supported by the device.
So, next I decided to try upgrading the firmware to 1.36 to see if that works, and since I will have to take it off of S2 in order to do so, I will run some more tests first using 1.34 and no security.
First, using the Hubitat driver, I tested whether I could now see events in the event log, now that it had no security. And indeed, I can:
pushed 1 DEVICE 2020-08-16 08:21:34.334 PM EDT
pushed 2 DEVICE 2020-08-16 08:21:26.023 PM EDT
And with the Inovelli driver:
held 1 DEVICE 2020-08-16 08:26:07.330 PM EDT
lastEvent Tap ▼ DEVICE 2020-08-16 08:26:07.321 PM EDT
pushed 1 DEVICE 2020-08-16 08:26:04.781 PM EDT
lastEvent Tap ▲ DEVICE 2020-08-16 08:26:04.771 PM EDT
Note the difference to the above logs from when the switch was on S2 - in these logs I am getting the pushed/held events. In the logs with the device on S2, I only got the lastEvent lines, not pushed/held.
Note as well that while the Inovelli driver treats a fan button single press as a held 1 event (as my button controller expects), the Hubitat driver treats it as pushed 2 instead. The button numbering assignments are not consistent, as perhaps I should have surmised from Hubitat's double-click driver feature.
Going into the button controller app and editing the existing app just to reassign the same device and same buttons to my testing action:
And boom, everything works when using the Inovelli driver. All elements of both actions. Nothing at all works when I press either button when using the Hubitat driver, even though now we're connected without security and the events do appear in the device logs, and even though this driver reports pushed 1 events in the same way as the Inovelli driver.
So, new theory, there are actually 2 issues:
- Some problem with incomplete event delivery on S2-connected Inovelli switches.
- Some problem with the Hubitat LZW36 driver preventing it from triggering the events.
Next, I upgraded the switch to 1.36, leaving it on the unsecure join. This yielded basically no change. I was able to now see and set parameter 51:
parameter51value : 1
The event log using the Hubitat drivers was the same as in 1.34. And still, none of the Button Controller automations triggered. Using the Inovelli drivers, logs are the same and everything works.
If anyone is interested I can put it back on S2 and verify that stuff breaks again.
