I setup my hubitat a couple of years ago. One application I wanted was to monitor the open/closed state of my freezer door.
Everything was going well until recently. This morning I went into the freezer to find that it was room temperature. I don't know if the door was left slightly open or not, but when I went to hubitat, all the devices are not responding and at the bottom of each device page, it says last activity at some date in the past. The freezer multi-purpose sensor is saying july 13th. Some others are saying last week.
If I look in the device events, the events tend to correspond with the last activity.
How can I get the hub to try to initiate contact with the device again?
I don't think the batteries are dead, each device says it had around 45-ish percent battery.
Well... it looks like something went badly awry with your mesh. Are these devices Zigbee, ZWave, or a combination of both? And did anything change? Have you added any new devices, moved your hub, etc? Have you checked the hub log? And have you confirmed that the Zigbee and zwave radios are still on?
BTW I use @bptworld's device watchdog app to keep an eye on devices that disappear. Works great, but obviously won't help you now.
I'm not saying the battery is dead but battery reporting can be notoriously inaccurate.
Just something to consider.
@bobbles yea, I got suspicious of that too, so I broke out the multi-meter. The button had 2v left in the battery of a 3v battery. I assume that's not good enough. The two freezer batteries were 0v, so that's definitely not good.
So not it is off to the store to get some new 2450s.
These were all smart things devices on zigbee.
@brad5 thanks for the suggestion of the device watchdog. I will get that setup once I get the batteries replaced.
Right now I am watching an ecobee sensor in the freezer to see if the freezer died or if the door was just left ajar last night. fingers crossed it was just the door. And that the poor little ecobee sensor survives the cold.
Inaccurate battery reporting is the bane of using z-wave/zigbee sensors. My practice is to preemptively replace batteries when they report ~50%.
If that’s the v5 SmartThings multi sensor, they are not very good on battery. At least that’s my experience with the one I had on my chest freezer. I now only use them converted to powered.
You probably will have seen this elsewhere, but battery reporting may be the single worst problem with devices in a network - at least in mine.
Overtime, I've come to recognize that 52% is about the lowest I can safely go. While not all devices are poor about battery levels (some I could let go down to 30 and be reliant) its so hit or miss as to be a really stupid problem. Everyone sneaks solutions of one sort or another - replacing battery with external packs, USB charger integrations, adapting to Mains power ...
Like one poster before me, I use a combination of Device Watchdog and Battery Monitor apps to keep my eye on it.
2 more factors to make note of to help you gain time and reliance - There's some devices that allow configurability on the reporting schedules - ie; a device that reports every second is gonna take a lot more juice than a device that reports every hour. Finding sensors that allow being configured can be really helpful vs. ones that are hard set.
The other point, you mentioned freezer - batteries of devices inside of a freezer won't last as long as room temp batteries. " When temperatures drop the internal resistance of the battery is increased."
Hope this is useful experience I'm sharing!
That seems excessive
The batteries we use, lithium, have an end-of-life "curve" that looks more like a cliff. at 99% of life, the battery is producing greater than 80% of it's voltage. Clearly at 100% it produces 0%. That final 1% drop is precipitous. If the device is sampling the battery voltage, its final report might be at the very moment that there's not enough power left to get the message out. Thus it's the penultimate report that is the final received report. In the best of cases, the final received report is going to be greater than 45% of the battery voltage. The more usual report is in the 60% region. Worst case, with high drain devices, the final report might be 100%.
The quote being clipped tainted the statement. I repeat "... in a network - at least in mine." YEMV - shu
the grinning emoji was my hint that I was trying to be lighthearted..
I just remember the titles or positions you've espoused in the past couple months and got a smile on my face from what I perceived as one more "worst thing"
HE’s biggest weakness :crossout: opportunity
I'm not suggesting you alter what you're doing. I simply found myself with a grin plastered on my face.
You have the memory of a steel... um... what's it called?