In case someone need to perform factory reset. Here are the steps.
remove all power (battery and dc)
press the button on the back of the board.
plug the dc power and count to 4 to 5 seconds (not more).
release the button. At this point, the module is in pairing mode.
The module should be in paring mode out of the box. I am still testing each module before I ship at this stage. When I remove the module from my hub, one or two of them may not cleanly leave my network. You can do the factory reset by performing the above steps.
Do you have it plugged in to either a usb or cigarette lighter plug with 5v converter?
The sensor should switch the power source to dc.
While on the DC, the timeout will be 1 minute. While on the battery, the timeout is 7 minutes. You could be 5 miles away before it changed status in this case.
Lets make sure your sensor change status to DC when your car started. Otherwise, the scenario will not work as expected.
The car was at that address for over an hour. I reset the timers to their lowest setting (1 and 2 minutes if I remember correctly).
I will verify that it switches from battery to DC when plugged in when I get home later. It is on battery currently. I haven't tried it while plugged in.
Hello @iharyadi and everyone else interested in this sensor.
Unfortunately, due to ongoing construction, I have been unable to test this sensor in my car; however, I have tested it at home and have the following comments:
After charging the battery, I have unplugged the sensor, and it has been communicating fine with the hub for >72 hours. The battery reading is still 100% at this time.
While it is on battery, it is probably a good idea to configure the Presence timeout (minutes) to be 7 minutes (default) or longer.
I had configured the presence timeout to be 5 minutes and started observing sporadic not present/present events in the logs, even though the sensor hadn't moved.
These sporadic events stopped after I changed the Presence Timeout to 7 minutes.
The temp sensor is pretty accurate after correcting for the offset. It is currently reading 74F, and three other sensors in the room are at 74.2F, 74.3F, and 73.8F.
Thank you for sharing the information. It is interesting that you got so many present/not present cycle. What this mean is that the sensor lost connection to its parent. In addition, the recovery process could take longer than 5 minutes. This time is enough to change the present. My Zigbee network in my home does not have this many lost connection during its operation. I may have a couple reconnection in my worst day.
Please turn on the log if it is possible. You have set the timeout to 7 minutes. The state of presence is not going to tell you about "lost connection to parent" any more. I do log it when you have it enable. One impact of loosing the parent is battery consumption during re-connection process.
Regarding this not sure what driver you are using but the.one i have for the samsung arrival.sensor has a setting for the minimum time it has to be in the away state before it will trigger .
This cuts down on false alarms. Even if your driver doesn't have this you could simulate the same behavior in your rule.
We (I and @aaiyar ) are using a driver that I wrote for my Arrival sensor.
The longer timeout will cut down the false alarm. This is correct. While the car is at home and powered off, the sensor will run on battery. "YES", we can relax the timeout interval in this state. I provide a very generous timeout while the sensor is battery powered exactly for cutting down the false alarm.
Once, the car is started( assuming it is most likely leave the house 90% of the time). The sensor is powered from the DC. Using my DTH, we will be using a different timeout setting. In this case, we want to be as aggressive as possible so that the "not present" state is detected as soon as possible.
The above is the difference between my Arrival sensor vs ST arrival sensor. Using a usb port (or cigarette lighter to usb) in the car, we could use different set of timeout.
seeing the below log entry frequently in my logs. Is it anything to be concerned with?
[dev:1106](http://10.0.0.42/logs#dev1106)2022-04-06 12:59:24.061 pm [info](http://10.0.0.42/device/edit/1106)device recovered from lost of parent at 2022-04-06 12:59:24
This is actually too much. Wireless protocol is not perfect. But, if you loose connection to ita parent at this rate, there is something is not quite right.
Yeah I am definitely seeing issues. The device seems to have just dropped off my network. Strangely enough it is the only zigbee device that seems to be having connection issues.
@cwwilson08, this sensor has more aggressive DC activity during DC power. The activity will be significantly reduced while it is in battery mode. The Zigbee is intended to operate in DC and connected to the network reasonably in short period. Only when your car is approaching and leaving home. Once the car is powered off, it will be operating in battery.
My guess is during the DC powered and the combination of probably less than ideal Zigbee network cause it to drop from the network frequently.
I personally tested both DC powered and battery powered for very log period of time. Zigbee device does lost its parent and reconnect from time to time. There is no perfect connection. However, we are talking once or twice in a day in a bad day.
I have no doubt of it. I firmly believe what is happening is something on my end at this point. I think I may have tracked down the issue (fingers crossed). I had some hubconnect devices spamming the hub with excessive events, even though the source hub wasnt affected. Until now I had not noticed any issues. I disabled the whole set of virtual devices and the sensor really seems to have calmed down (no more missing parents lol). Its been humming along good for a good awhile now. I am optimistic.
So while I have had some frustration in the process, I believe it ultimately led me to correct a problem I did not know was there.
@leeonestop, yes... The project is still alive. I am making small batch as we speak. The manufacture is still collecting parts. I will update everyone here when I got better estimate for the availability from the manufacture.
