Wireless switches: 1 slower than another to toggle Zigbee lights

First off, here is my setup

  • Coordinator: Conbee 2 stick
  • Integration: Deconz (Phoscon)
  • Lights: 2 Zigbee bulbs (LIDL Livarno E14)
  • Switches:
    1. Ikea 5-button Zigbee remote E1524 (the round one)
    2. Aqara H1, the double rocker wireless Zigbee switch WXKG15LM (WRS-R02)

1 - with the Ikea 5-button Zigbee remote, toggling those 2 lights is blazing fast :zap: :+1:

alias: Toggle Study lights with Ikea Tradfri remote
description: ""
trigger:
  - device_id: e731dfd279799fd6d208e604e727e0a7
    domain: deconz
    platform: device
    type: remote_button_short_press
    subtype: turn_on
condition: []
action:
  - service: light.toggle
    metadata: {}
    data: {}
    target:
      entity_id: light.study
mode: single

2 - with the Aqara H1, it’s not as snappy as with the Ikea.
There is a noticeable delay.
Nothing crazy but clearly slower than with the Ikea button :neutral_face:
I first thought that could come from the fact that I use the platform: event (deconz_event here) instead of the platform: device like with the Ikea button.
But with the Aqara H1 I don’t have a choice because all is exposed for that device (with my Deconz integration at least) as a trigger is the battery level change :frowning:

alias: Aqara H1 Toggle study lights
description: Aqara H1 Toggle study lights
trigger:
  - platform: event
    event_type: deconz_event
    event_data:
      id: aqara_switch_dining_room
      unique_id: 54:ef:44:10:00:06:5f:e8
      event: 1002
      device_id: 77d17f9d2a63cc777023d3f5f60aad1a
condition: []
action:
  - service: light.toggle
    metadata: {}
    data: {}
    target:
      entity_id: light.study
mode: single

So I told myself the relative slowliness of the Aqara might come from the platform: event so I created an automation with the Ikea button using the platform: event and it was quick nonetheless.

Conclusion: could there be a setting related to the Aqara H1 switch itself that would make it a tad bit “slow”? If so, can it be modified? :thinking:
I deep-dived into the Phoscon settings but they are absolutely not user-friendly :sweat: :laughing:

Can’t help with deconz-specific issues, but check the trace for your automations, once using the Ikea button, and once again using the H1.

If there’s no noticeable difference in the timings for both individually, try pressing each device at the exact same time. The trace should let you narrow down where the lag is coming from.

Thank you for the suggestion.
I’ve just tried that (clicking on both buttons at the same time) but cannot come to any interesting conclusion :neutral_face:

Trace Timeline tab in your traces :wink:

We start with the lights off.
On the left, we see that the Ikea button is faster and turns on the lights.
The Aqara button comes second and, since it’s a toggle in both automations, it then turns off the lights.
Nothing I didn’t know already ^^

Hmm. Automations are both taking equally as long to complete (0.04s). That should really not be noticeable to the naked eye. Yes, Ikea technically DID get there first by turning on the light, but welcome to the world of race conditions.

Anything in the deconz logs? ~Not sure how to turn on enhanced logging in there, so can’t help much.

Late reply as I stumbled over your thread looking for something else. I have three Zigbee bulbs in a ceiling fitting in my bathroom under z2m. Using HA to control the “room” or the group helper with them in always resulted in them switching sequentially. I guess HA has to serialize the control messages it publishes to MQTT or z2M does the same as it reads them off the topic but I could be wrong there.

Anyway, as I use them with z2m and I was able to add them to a Zigbee group there and, when I control the lights through that, they operate, as far as I can tell, simultaneously.

Grouping them at the Zigbee level doesn’t stop them being addressed individually or via HA groupings if you still need that for scenes etc.