Not just with OZW 1.6, and not just with the new MQTT integration. but even with OZW 1.4, and with Vera.
They do not return their state changes like the previous generation Qubino flush devices.
if you use the switch enumeration (instead of level) and the device is off, you click it on and the toggle moves to on, and then back to off. Meanwhile the device turns on just as expected.
now the U.I does not match the device and you want to turn it off, so you click it on again and this time the toggle moves to on and the devices stays on.
but you want to turn it off, so you click it again, and then the toggle flaps to off and then to on, meanwhile the devices turns off just as requested.
now the U.I again does not match the device state. So you be nice to your future self and click it one more time, or just leave it and let that other guy deal with when they want to change it’s state.
the level enumeration is better, but only because i think home assistant notices that the level is infact greater than 0, and so updates the state of the U.I when it notices. so there is a significant lag, which makes using the dimmer/slider hard to use.
I’ve sorta had this explained to me in the past, I think I must have been chatting in discord. it has something to do the dimmer returning the current level before it has had a chance to adjust the level in response to the requested demand.
I just figured surely this would get resolved somewhere between Qubino and OZW. I fear though it’s Qubino who have done something a bit too quirky for anyone to deal with easily, and I fear more that Qubino’s are not updatable even if there is a firmware resolution.
Does anyone else have other experiences? resolutions even?
I have a fair few of these dimmers (10), and they all behave exactly the same.
Pfew, hard time reading and understanding what you mean. I have ordered a mini and will soon implement with OZW1.6 (QT openZwave) I’ll report back behavior.
So I’m not really talking about home assistant, it’s just one U.I.
And the OZWAdmin is another U.I.
The behaviour I’m describing transcends all U.I, and as I saw the same behaviour in Vera, it may not be z-wave implementation either as I don’t believe Vera uses ozw.
Any of the U.I I’ve listed do not initially reflect the state of the device upon sending a state change.
So the U.I says the state is off (and it’s correct to start with). The U.I forwards the request to change state and however the U.I determines the success of that request (I don’t know) sadly for the mini dimmer it’s always a failed request. So the U.I flaps back to original state because the U.I believes the device is still off.
Unfortunately, the U.I is misinformed as the request was successful and now the state is on.
On second request to change the state to on, the u.i is better informed and thus updates the u.i to reflect that the device is in fact on.
The same is then repeated with further state changes
Hope you can get yours working better, let me know!
They are certain that they would not have been certified by z-wave if they had not implemented these interactions correctly.
However, they couldn’t really comment why their “Flush Dimmers” behave differently.
Instead they were insisting that the gateway was mixing MULTILEVEL commands with BINARY_STATE commands.
The “Mini DImmer” can be switched between Dimmer mode and Switch mode. Where in switch mode there is no MULTILEVEL command.
The confusion could well be in the implementation of on/off over MULTILEVEL.
But i’m well beyond what I know.
Lucky for me, the consumer, it seems to fall to me to connect the dots between Qubino and the gateways.
Although I was impressed with their support (very quick replies, quick to replicate my issue, helpful and friendly) really it is in their own interests to ensure gateways (particularly ones they list as supported) work correctly with their devices, otherwise why would anyone else buy them over the competition.
I guess at this point i’m now a stake holder having a reasonable percentage share of their sold stock. Reasonable to me it seems is 10 of them.