sensor16 is working as expected,
sensor23 not, unicast, any idea what i can do?
tried to delete the group that was created for the sensor23 (sensor16 does not have one) but still triggers too much…
@Mimiix Thanks,
i tried this (unbind from group 0xFFF0) but i get the message:
failed - no entry
Update: I can bind and then unbind the switch, but it seems the switch is not bound to this group without me binding it first, but it still broadcasts to group 0 somehow…
Short update: battery of the switches looks good so far after upgrading to the pre release.
Still have the issue with the one switch broadcasting to all light, without being assigned to group 0xFFF0…
As I have quite some automation build in ha on top of the deconz stuff, I would like to avoid a complete re-setup.
Also, how can it be a feature that a switch gets assigned to all light during paring, especially if it is not possible to undo this easily?
Thanks,
Already tried this, but no luck with my switches…
What i figured out is that it seems only the state in phoscon is changing(group and light).
I have some automation in home assistant and only the lights i control via those turn on for real…
what else can i check?
Hi,
I am using phoscon groups and the issue was that all new switches triggert ALL groups in Phoscon (and ha as well), BUT in phoscon only the groups showed active, no lights turned on (only those i control via aturomation based on phoscon group in HA).
I finally found a workaround that works for me:
Following the guide from @Mimiix, i had to apply the sqlite change and set the 0xFFF0 group to deleted.
Than i had to remove the switch from deconz and make sure no assignments to a group still exist (manually delete in UI the switch assignements).
if i would keep an assigment, let’s say to my kitchen, the switch after re-paring still triggert all groups, if i removed the assigment it works fine.
Now I can add the switches without them triggering every group.
This looks like a very Buggy behaviour to me: Groups in Phoscon show active but it is not true…
I am still using Phoscon version 2.25.3 (Via HA deconz plugin).
The battery situations seems way better now, I still burned 10+ CR2032 batteries while testing.
In my case it seems that every new tradfri switch triggered all groups in phoscon without really changing the lights in the group (nothing todo with HA at all!).
I checked in the rest api (groups, sensors, rules, resourcelinks,…) but could not see a reason for this.
In Deconz the swicthes have NOT been assigned to group 0xFFF0, so it looks like there is a diff between the zigbee network (groups on, light not) and the phoscon data in the rest api.
For the little story (not an anwser for your issue) but it’s interesting to know (It’s a remark from ebaauw)
Source routing only affects unicast messages (which need to be routed). Groupcast messages are broadcasted through all routers, without any route discovery nor source route.
Hello.
I updated yesterday the conbee3 to the last firmware (0x26510900) because i have a aqara smoke sensor that lost the conection each 24 hours (im using zigbee2mqtt).
And i want to try if this firmware solve this issue.
not pass 24 hours and i dont know if solve this issue, i will report as soon as posible.
BUT I HAVE OTHER PROBLEM.
My 4 “SONOFF ZBMINI” lost the conection. If I push the reconfig button in the zigbee2mqtt, it will conect but after some time, they lost the conection again.
This never ocurs with the previous firmware version.
other stupid thing: in the firmware changelog the date for this firmware is february, not april
In case of switching the coordinators a device may stuck in error state which is indicated by slowly blinking light (about 1 flash per second). To switch a device to a pairing mode power it off for 10 sec. After power on it enters pairing mode automatically
And thoses devices have issue for pairing, I don’t remember exactly but from my memory the inclusion need to be done on the coordinator, don’t use routers, there is a post about that on Z2M github.