Agree. the polling interval is controlled by the either the sensor device or by the gateway.
BTW I am pretty sure that the interval is not a fixed 1hr, but will vary depending on if there is significant change.
E.g. after a shower I can see humidity readings rise in shorter window than 1hr.
You can see this in the graph as the data points are closer than 1hr.
E.g. for me this morning I got two readings 1 minute apart as humidity dropped:
I’m looking to use this as a reading for heating adjustments so if readings are based on relative change, rather than set intervals, this might work great. If a new reading is recorded when the temp increases, triggered by heating switching on, it would be ideal.
I can see in my current readings that there are a few close together also (arount 5-8 minutes apart):
At the moment I use a Aeotec ZW100 MultiSensor 6, which I have manually adjusted the Z-Wave settings to read every 10 minutes, and I haven’t noticed much affect on battery life:
I’m going to try the shower test also, I’ll put it in the bathroom, and remove it soon after to see what happens.
I’m not sure, I set mine up basically a week ago and you can see where my battery is.
Still, I don’t think the batteries were definitely 100% when I installed them, and they are probably using cheap batteries (not duracell etc)… So your mileage may vary.
I haven’t been paying any attention to the battery usage this past week.
I think the battery level is not well calibrated with these sensors.
For sure none seem to start @ 100% (due to being factory fitted I suppose)
also some people see rapid drops then no change for a while.
I moved one of the sensors into my bathroom, had a shower, and put it back into another room afterward.
It looks like some of the results were as close together as 3 minutes, and indeed results are reported when a significant change in temperature was recorded.
I also checked humidity, which shows the spike in humidity, a total of 5 recorded results in space between 23:22 and 23:42.
Results Temperature:
Results Humidity:
Lastly, @hijinx, on the other sensor, unmoved since yesterday to today the reading was 43% at about 16:50, and is now 41% today at approx 17:10. At this rate it would be a few weeks or months at most, I think.
Yeah, that sounds about right, the batteries should hopefully last a while.
I’ve had my Xiaomi sensors for only a couple weeks now.
As for the granularity for the heating automation, I’m not sure entirely so I think I’ll just need to try it and see after the weather cools down some more…
At the moment, nothing changes, it’s too warm.
The reactive reports due to changes is encouraging though, as at the moment I have the heating automation report from an Aeotec ZW100 MultiSensor 6, which is manually configured to report in at strict 10 minute intervals.
I guess that if the Xiaomi sensor can react in less than 10 minutes then that would be a win?
The temperature+humidity sensor reports with changes over 6% change on humidiy or 0.5 degrees on temperature. This seemed good enough for me, for my heating system, but in the real life, it has some problems for example:
Let’s say you have your heating desired temperature set to 21º, and the last temperature reported by the sensor is 20.9, so the heater is on. Then, at a particular moment in time due to the current climate conditions, it takes 5 minutes to increase 0.1 degree, so to reach 20 degrees, it only takes 5 minutes, but to reach 21.4 that is the point where the sensor senses a 0.5 change and emits a new value, it takes 25 minutes. So there’s an extra 20 minutes of heating that could be avoided just if the sensor could be polled every 5 minutes.
I’m pretty disgusted with this, as there’s no easy way to get around this, and this situation happens more often than i would. So if anyone know a way to hack the electronics or the gateway to increase the interval it would be nice. I prefer to change the batteries more often that having this bad behaviour.