In regards to recent update
Generic Zigbee Outlet, enable power reporting for Salus SP600 plug
There seams to be something wrong with the power reporting conversation.
Using a previous driver when my tv was on the power reporting was around 55. Using the generic zigbee outlet is only reporting 5.
Reporting has been set to 10,5 and 1 but all the same
@bobbles, @andydvsn ave you tried using the generic zigbee outlet on the latest version? @mike.maxwell do you have one to confirm power reporting levels?
The issue is with the Salus devices ..... some reported with a different precision. I have some that report in watts (so the driver receives 55) and some that report in tenths of a watt (so the driver receives 550).
The generic driver probably uses a divisor based on the model name / number in the fingerprint of the device, so for the Salus it likely divides by 10.
Not sure it can work out what divisor to use unless there's a discernible difference in the device info (product number, name, etc).
SmartThings sockets have a similar issue, but they change model name / number so the driver should behave correctly there.
Do I have one of these devices?, I believe so, the generic zigbee outlet driver uses a divisor of 10 as the default.
I wasnt aware that the sp600s reported in different precisions.
Yeah PITA, early ones reported raw watts, later ones tenths of watts (so the /10 works OK).
IIRC technically the divisor / multiplier "should" be readable from the device for the Metering Cluster, but not all devices implement it ..... I've never tried with my bunch of Salus as I wrote my own driver for them and never bothered changing away from it (at the time I just cheated and had one driver with a /10 divisor and one without).
I've turned off either my hub or just the Zigbee several times without any issues. The LED on the Salus flashes red when it can't connect to the Zigbee network, but then flashes green when it is reconnecting and once it connects just returns to whatever state it was in previously.
I don't think you can correlate what happens on deconz to what happens on HE.
Edit. In any case this doesn't have anything to do with the power consumption reading.
Ok from the fingerprints posted, I see some with application 14 and others with application 09
It is an accurate statement that 09 reports correctly and 14 doesnt?