It is puzzling why your devices are updating humidity at a greater frequency than temperature. My experience with units has always been temperature and humidity at same frequency or temperature rate greater than humidity.
Have you tried putting a heater or hair dryer at one of the units and see the rate of change? This, I think, is a good gauge of getting the maximum sampling rate.
With the exception of one, unfortunately with the sensors I have seen, whether BLE, Zigbee or Wifi, you have zero control over the frequency of samples or frequency of broadcasts of data. All of the devices are ‘push’ model, they go to sleep for a period, wake up and sample then broadcast. There is no way to ‘wake’ the unit from the hub when the unit is a asleep. As I said, with the exception of the Xiaomi LYWSD03MMC-ATC, the hack of this allows you to set the broadcast frequency, still does not (as yet) allow you to change the sensor reading period.
https://github.com/atc1441/ATC_MiThermometer
Maybe there is a configuration property that your Sonoff sensors have that might be set-able for sampling and broadcast. But, if there is such a setting, I doubt if it is made public by Sonoff.
The third factor that you have to consider is the signal quality issue. Even if the unit is sending updates at one rate, if the unit’s packets are not being received at the hub due to distance or other interference, you will see a lower update number. You can monitor this by the RSSI and LQI values you are seeing for the unit, and in a low tech test, by putting a test unit right next to your hub and using it’s update rate as your ‘gold standard’. Also, I am pretty sure that none of these ‘end devices’ in Zigbee parlance can have a ‘hop’ between the unit and the hub. So no amount of repeaters will improve the unit’s performance.
Ok looks like I’m an idiot
When there’s a significant change they update more frequently. I guess with the humidity sensors the noise on the sensor is greater than the threshold. Not great as it will eat up battery life.
I blew on one sensors and got a quick update. Then the other 4 and also got almost immediate updates
I have 10 TH01 devices in my ZHA setup and none, absolutely none of them is reporting at 2 mins interval nor to a 0.1°C delta. I’m also using Tasmotised Sonoff Zbridge. And is not about the distance or obstacles between the coordinator and the devices. I have the zbbridge in one room with a TH01 ( 2 meters apart - no obstacles between ) and the update frequency goes up to 2 hours ( while the temperature variation can be as much as 3 degrees ). So for sure, something is wrong. I thought I got a defective batch of 10 sensors ( although is very very strange that the humidity values of same sensors are being updated very nice and frequent ). But I see others have the same issues. So… where is the problem, I’m asking myself?
I switched from ZHA to zigbee2mqtt in a effort to try and resolve this and the configuration in there let’s you change the reporting intervals easily. I have it reporting at a 10 second interval with no issues. I appreciate that’s not tremendously helpful if you want to stick with ZHA but depending on your use case might be worth considering.
@demig, the reporting interval, you’re sure is the one the sensor will use to send data to the zigbee bridge, or just the interval that the bridge is reporting to HASS? Because if the the second, nothing will be fixed in my case. Is zigbee2mqtt capable of questioning the sensors as opposed to just waiting for their report?
I’m currently looking at a trace of an SNZB-02 placed into my freezer. It’s a beautiful smooth curve recording the temp drop from 21 deg down to -22 at 10second intervals so yes I’m convinced! I’ve changed the minimum report interval and minimum reported change using the zigbee2mqqt front end, pressed the button to activate the sensor and have confirmation that it has updated. I didn’t know you could do this. You can see a flat plateau when the battery was dead and then the curve when I started it again and reconfigured
It is interesting to see others with the same concerns I had. I never found a solution to this problem/issue/concern that I had. However I am not sure if it is actually a problem unless you want smooth regular graphs.
My final solution was Tasmotised Sonoff Zbridge to ZHA. 12 SNZB-02 sensors to cover the rooms in the house and an Ikea TRÅDFRI plug to act as a bridge for two of the rooms that were just too far from the bridge. All sensors are paired with the bridge the Tradfri is only a repeater and controls a lamp on the landing.
I have now been running my heating system since August using the room temps from these devices and it works. Most sensors are still at over 90% battery with the kitchen (which has the most temp and humidity changes) at 64%. So it looks like I will get a year or more out of a battery.
I do have temperature graphs with all sorts of reporting intervals. But to put my mind at rest I have two LYWSDCGQ 01ZM sensors that I have moved around and on the whole agree (although with more data points) with the sonoff sensors. It is strange how depending on people in rooms, wind outside and many other things room temperatures do not change much. I think I would prefer the Mijia sensors (screen is nice and AAA battery is better) but the overhead of getting good BLE coverage in my house is not worth it.
I was also tempted by zigbee2mqqt but the Tasmotised Sonoff Zbridge can be placed in the middle of the house, is cheap and prebuilt hardware. I am also not sure what a decreased/regular reporting interval for temperature would do to the battery life. It would be nice to have a max interval of 15mins though.
Not sure this will help anyone but thought I would share my current experience. Maybe in the summer when I do not need the heating I will look at this again.
@AndySM 15 mins interval for reporting the temperature is indeed reasonable, but when that interval goes up to 2, even 2.5 hours, there’s a problem. A thermostat can no longer rely on that sensor.
I have the same problem, I need more frequent update on temp. I am using ZHA with ziGate.
I saw that it’s possible to change the report interval using Zigbee2MQTT, so I thought it might be possible with ZHA.
In the device property I found some attributes to set, but even with the sensor awake, it doesn’t seem to update anything. Is it expected ?
I came across this thread because I have the same question, though mine stems from devices behaving differently from one another. I bought 6xSNZB-02 in 2 batches of 3, from the ITEAD store, a couple of months apart. The humidity graph speaks for itself, you can tell which ones were bought together. I can’t work out why they would be different unless they have different hardware inside.
I managed to fix the issue with sensor updating interval on Zigbee2MQTT integration.
Here is what I’ve done:
-Enable ‘Permit join’
-Put sensor in pairing mode
-Using Zigbee2MQTT interface select desired sensor and manage to the tab: ‘Reporting’
Your screen should look like this:
I can’t seem to find the things you’re referring to here. I’m using the same sensor and zigbee2mqtt in HA. I see no place where I can adjust the minimum report interval or a button to activate the device. I feel like a noob now. Can you be more specific as to the place where these settings can be adjusted?
I can never get to that screen that has Reporting as a tab. It seems strange to me that one has to have the device in pairing mode to set this. Wouldn’t these things be set on an already-paired device?
If you don’t have the Reporting tab, you probably have a very very old Z2M version since it has been available for months now with the new embedded Z2M GUI.
Thanks for that. I have version 1.15.0 of Zigbee2MQTT which I think is the most recent. I have Auto-Update switched on. Oh wait. I see on the Github page that the version is now 1.19.0. Not sure why I am not auto-updating. Must correct that!
