Rick, honestly I am not sure at this time what could cause this problem - the driver, or the Zigbee platform instability in the last month (both C-8 and C-7). ... I am now afraid to pair the sensor from my C-8 hub to the developments C-7 hub, because I am not sure it will pair to the production hub later again...
When did this problem with the dropping of the network start?
C-7 or C-8 hub?
The warning 'unprocessed water event unknown' is logged when the driver determines that the devices is offline (nothing received in the last 12 or 13 hours). So the problem is that the battery percentage remaining reports are not sent automatically by the device anymore ...
This driver does not explicitly configure the battery reporting periods. It wasn't needed before, all the tested water leak sensors had the 3 or 4 hours reporting period set by default.
rxCounter = 0 means: nothing has ever been received from this device!
Have you recently removed it or re-paired it?
@rlithgow1 I paired the NEO Coolcam to my dev. hub anyway (this 'unprocessed event unknown' is not a correct log message and must be fixed), but very strangely, although the previously paired device was found, the water sensor didn't work (didn't send any events to the C-7 hub) until I clicked on the Configure button.
Can you test with your sensor?
Update: you can also update manually to ver. 1.0.8 2023-05-13 (dev. branch) - I have implemented the fixes for the 'unprocessed water event unknown' , but this is not much essential for the going offline issue.
Please let me know whether pressing the Configure button worked.
Pressing the configure button did not work for either sensor. Pressing the wakup button on the sensor didn't bring it back into the mesh. Going to try your fork and re pair the devices and leave them on my desk (closer to a zigbee repeater) and see if that helps.
EDIT: these are not actually battery reports, but time synchronization requests from the Tuya device.
(digital) in the logs means, that the battery report was not physically sent by the device, but generated digitally from the driver (based on the last battery report actually received), and the purpose of this workaround is to keep the 'LastActivityAt' device property updated with the last time something was received from the sensor. This means that the device is alive and paired to the hub.
In the latest HE platform update 2.3.5.136 the method for updating the "Last Activity At" was changed:
Bug Fixes
Fixed Last Update Time to reflect the last time a message is received from a device (as opposed to the last time device attribute changed).
So probably this 'digital' battery reporting is not needed anymore, I will test further and will remove it from the driver if no longer neccesery to keep the Device Watch Dog status correct.
This is a new one (actually 2 of them). My others that work fine were migrated from my c7 to the c8. They both drop off. I'm sending @mike.maxwell one of them so that he can experiment with it.
An update to version 1.1.0 was pushed via HPM, where the Presence was replaced with healthCheck attribute.
The change will happen automatically after the first check-in message is received from the leak sensor.
I have several reporting type=physical every 4 hours. Battery values vary, one is 100%. I will press the button and see if type changes to digital. I have one I had to rejoin recently, it is reporting hourly and type=digital.
No, I do not. This physical battery event posts every 4 hours or so.
This is with your Tuya NEO Coolcam Zigbee Water Leak Sensor, but in looking into it, I may have a down-level (1.0.7). I'll upgrade to 1.1.1 to see what happens. I do use HPM, but somehow missed intervening revisions.
I only have 100% reports currently, but in the past I have gotten low battery reports leading to me changing the batteries on the sensor. I don't remember if I got gradually lower battery reports or if the reporting just dropped from 100% to a low level suddenly.
False alarm. This is one is a Aqara sensor that I had wrong driver assigned.
So my Neo Coolcam water leak sensors are reporting:
40% physical
33% physical
28% physical
100% physical
3% physical, needs a new battery.
I have tried pressing the button in some cases and rejoined in other cases. 24 hours later, no change, battery reports are 4 hours apart and type=physical.
model: TS0207
thanks to your advice above I paired near the hub and had to do multiple times
In 2 cases I had to do a 3rd time....
Is there any explanation for the need to do multiple times ?
also once when it didnt pair I got the following error message (not really concerned since they eventually paired but just in case it helps at all
I have an issue with the Tuya TS0207 series and C8 where they pair but do not report wet/dry.
Bought 6 different 0207s' from different manu's with same result. Works fine on C7.
Ended up installing a mechanical float switch connected to a door sensor in place of it's original reed switch to Mc'Giver a working C8 leak detector.
you may have to repair several times.... don't remove, just press the button for 10 sec to start pairing and then under devices search for new devices ....it will show up as 1 that was already paired, click on it and check if working...If not repeat ...in some cases I had to do 3 times and then they worked.
I converted a _TZ3000_upgcbody to hard wired power DC power. It seems these things go to sleep on me. Do you know of a way to keep these things awake? I have power to spare.
