I have PI4/2GB running HA 7.6, I have Sonoff USB 3.0 stick plugged directly into the PI’s USB port, with the latest versions of Zigbee2mqtt and Mosquitto broker installed.
I have four Sengled smart wall energy monitors and two Sonoff wall power switches. Each of them act as a router as well. See attached screenshot of the map.
According to Zigbee2mqtt, here is device info (they are identical):
Firmware build date 20191205 Firmware version 3000-0001
In the Exposes section I see the following info of one of the deice directly connected to the Coordinator:
voltage
Voltage of the battery in millivolts 3085 mV
linkquality
Link quality (signal strength) 105 lqi
The thing is that the first time it measured battery capacity, it was showing 80%, now, after 12 hours it is showing 67%. The similar trend is with the second sensor as well.
Are these devices really draining batteries so fast that you have to replace the coin cell every several days? it would be no starter for me… Is there a way to change the firmware to fix the problem?
Or there is a problem with reporting the actual usage? How do I correct/get real info?
In my experience I’ve had more issues with the aqara temperature & humidity device batteries than any other device in either zigbee or zwave.
the batteries seem to drop quickly then stay at the same level for a while the drop again, etc, etc until they suddenly die. I’ve found battery life fairly short as well compared to other devices - some devices last a couple of months and others a really long time as expected.
As an example, I just replaced a battery in a device yesterday that I replaced 3 months ago and the battery right now says it’s only 67%.
One I replaced two weeks ago reports 62%.
I guess it’s possible that I got a bunch of bad batteries but I find that unlikely.
Well this sucks. I literally just ordered one of these this morning to test. Currently I use Tuya temp/humidity sensors. I only get a couple of months per battery. These Aqara sensors are advertised to get 2 years so I was hoping to maybe get 6 months… I guess it comes down to how often this sensor checks in. The Tuya sensor checks in about every 30 minutes. One review I read on the Aqara sensor says it checks in every minute. If that is true that may be why the battery drains so quickly. Another possibility could be the battery that came with it. Most of the time they are no name brands. The Tuya sensor I have all came with no name and died quick. I get far better battery life with a name brand batteries. I should have my sensor today so will see what happens.
I do not believe you can really learn a lot from the battery percentage. In Z2M they also just changed the model, giving new percentages, as part of latest version.
I have 5 of these. 4 have been running for 15 month and one have had the battery changed after app 9 month, however it sits outside and it was below 0 when the battery died, so not a normal usage. The 4 have voltage that are super stable at just below 3V, in Z2M they show 25%-37%, one is in a refrigerator, at a stable 10 degree.
The one I changed the battery in started at 3.2 and dropped quickly to 3.05 and have stayed there for the last 2 month,
You can not change the reporting, it reports when the change at big enough. I do not know the numbers.
One comment I forgot, hence a edit:
I did not use the original batteries as I read they were bad. I replaced batteries before installing the devices, with standard good batteries.
Not in my experience. I have six+ of the things, all going 8+ months on the stock battery.
The “heartbeat” check in is once an hour, otherwise they only report when there is a change. They seem to have some sort of threshold built in with regard to frequency. Seems like a 0.1 degree change may be delayed a little in reporting, and a significant change is reported more immediately.
It hasn’t been worth my time to try and figure it out in any more detail. They work, response is good when I need them to turn on the bath fan, and they have gone 8+ months so far.
Looking at my shower this morning, I had six reports inside of 2 minutes as the humidity/temp ramped up quickly, and much less often as humidity/temp floated down to normal levels.
Got mine setup. The OE battery is a Panasonic so leaving it. Both Aqara temp sensor and Aqara Door/window contact I got both show 59% battery. Strange they would both be that low out of the box. Not going to worry about since its sounds like you all are getting good runtime out of the sensors.
I recently got a batch of four of the newer MCCGQ14LM Door Sensors from Aliexpress that had dead batteries out of the box. They seem fine after replacing the batteries.
Quite fun we are just discussing this, I have one dropping out today. After +15 month in operation.
It dropped out at 2.955V, Z2M showing 30% battery left.
The problem with lithium batteries is they tend to fall off a cliff when they get near the end. That is then compounded by the fact z2m/zha doesn’t really know what voltage should be considered “0%” for any specific device.
Devices that self report battery percentages have a theoretical advantage - their engineers should be able to specify device appropriate bottom end value for the calculations.
For the devices that only report voltage like Aqara, z2m/zha can only use a generic range they think should work in most instances.
The z2m github battery thread is talking about changing the calculation again and lowering the upper voltage limit used in the calculation for 100%, but I wonder if they should also increase the lower limit assumed for 0%.
Agree, it is not a exact science. I have one Aqara weather sensor with even lower voltage, working stable. It is in the range of 2.935 - 2.965, changing a little up and down for the last month.
