I have 3 FYRTUR blinds (all have the latest 24.4.11 firmware). I noticed one of them stopped reporting its battery state via REST API. So I opened up die deCONZ and checked there.
I looked at the “Power Configuration” cluster. The 2 which report value via REST API have the same value at “Battery Percentage Remaining”. The other one has “Battery Percentage Remaining” 166. Does this mean 166 %? Is this why REST API stopped reporting because its above 100 %?
There are devices like the Phoscon Hive in the screenshot that report a value in deCONZ that has to be divided by 2 for the percentage value in the API. Maybe it is the same with the roller blind.
But i still don’t understand why on his capture he need to have a device with a /2 and not the second one …
The battery voltage maximum is 7.5 V, so the second capture can be fine 7.2V for 96 %, and in this situation the value don’t need to be divided.
For the first one, even the voltage value is strange, 7.8 V ???
And on C++ code (for native support), the value is not divided by 2.
So for me the problem is more the first device itself.
I updated deconz to 2.22 (DDF for Fyrur) just a few days after it came out. So like end of June. No changes.
I did not charge the batteries in the time range. But I updated to firmware at the time the battery values doubled. ID 47 is mising since then. That’s the device that’s missing the reading.
So maybe the / 2 need to be done with the latest firmware only?
Yep, for me too, problem, IDK if it’s possible with DDF ?
Not possible using a firmware check on selected DDF, perhaps used direclty on the value field ?
everything dropped to half as expected. So I fully charged one of them (ID 53). After that it is on 200 in deCONZ and 100 in the REST-API. So I guess its save to day the latest firmware needs that.
But the problem is you will use the same DDF for all your device, even the one that don’t need to divide the value by 2. So you will have some device with value 0 > 50.
I have all my 3 devices on version 24.4.11. So for me / 2 is fine for all devices. All data is correct now (just fully charged another one - 200 in deconz, 100 on the API)
Not sure this is the wisest thing to do, but anyways. Note that there’s also some of the older IKEA controllers that now report battery in 0.5% with firmware 24.4.x.