Zigbee light switch with decoupled mode?

[mikaelhertzman]

@andrazk

I have two of these modules. While they support decoupled mode, they lack direct Zigbee binding, which is a drawback.

Thank you for verifying! I will add it to the list since it meets all requirements from my original post. I agree that the lack of binding support is a drawback but it’s the same for the Aqara T2. When it comes to that option the only hope I think we might have is the Ubisys devices. Thank you again for your contribution!

@manoswerts

How can you see that direct binding is supported? If I look at documentation of those switches, I don’t seem to find n exact reference to the term.

The best source of information I have found is the device pages on https://zigbee2mqtt.io/devices/. As a alternate option you can also try it out for yourself (using HA, Zigbee2MQTTT and the device in question). There might be a better way to do this, but those have been my two methods of choice.

@manoswerts

The Ubysis C4 module seems to work very different from the others (even from their own brand like the S2) as it only exposes “action”. Is it something categorically different and why does it exist compared to the other modules that seem to operate very similarly?

The only bits of information I can find explaining this is at Ubisys C4 control via MQTT | Zigbee2MQTT and Control unit C4 – ubisys. If none of those sources answer your questions I recommend getting in touch with Ubisys support. They have been really helpful in our previous e-mail conversations.

@manoswerts

Does that mean that even without support for those actions, you can work around them with HA scripts/automations to support such things like dimming the light by holding the switch or cycling through scenes if subsequent clicks follow each other fast enough?

That is correct. I have created modified scripts from Trigger different actions on a single, double or double click on a binary sensor (if I recall correctly). The reason why I’d prefer if the device had native support for hold and double tap/click actions are:

1. More responsive - the switch sends the correct action to the gateway instantly. This means that any action (turning off lights, turning on your lawn mower etc.) will be carried out as quick as your Zigbee network allows. Say that you, for example, configure a 2 second delay to allow HA to “listen” for a second click. In this case you will not be able to have the single click action occur before you have “waited out” the 2 second window. This means that after every interaction you will have to wait 2 seconds before anything is carried out. I hope that makes sense.

2. More reliable - if you press a button twice or hold it and the subsequential clicks (or releases) are not received by the gateway for any reason, things will get weird. A package lost due to transmission issues, or the timing of your interaction being off, will in my experience result in frustration from the end user. (I know this from personal experience after hearing the sounds of my partner double-clicking switches again and again and again without it having the desired effect.) With finetuning it should/could/might work but your mileage may vary and it is in no way as effective as built-in actions in my experience.