You have 6 devices using the oldest driver. If it were me, I'd probably go into each and set them to use the Device Driver. It does nothing and just retains your device for later in this process.
Then, in HPM, UnMatch any 'Aqara FP300...' packages you find.
In the Drivers Code folder, delete all the 'Aqara FP300...'
Now you can more or less start over, Install using HPM, then go back to those 6 devices and swap them from the Device driver to the one, only driver you want to use. Giving each a click on its Configure button won't hurt.
@TArman advice will work, but won't fix HPM's db on the hub like my method does.
I guess I was going on the assumption that the latest version was installed by HPM, and everything would be OK.
Your method certainly is the definitive fix.
I think that's true but if there are multiple copies in the code folder(s) then the HPM db is probably going to have multiple entries. The devices using the driver were still 'attached' to the oldest code. That suggested to me that there were multiple HPM db entries and that technique of UnMatching all of them is the only way I know to get the HPM db back into a usable form.
I can imagine how @tray_e method worked, but I do wonder if there's an extra HPM db entry still. The problem is that HPM is thinking it's managing the code. Ok, we think it's gotten lost on that journey, but it doesn't change HPM's opinion that it thinks there are more than one similarly named code files to manage. The ONLY way to stop HPM is to UnMatch. That puts code management back into our hands. Following that, with either an Install or a MatchUp and potentially a Repair, puts code management back into HPM where it belongs
There was only one entry in HPM for some reason (My first thought was just to unmatch them all like when HPM had duplicate copies, and then run a repair). Just three copies in the driver code menu and available in the device type drop down.
Once I got all of them on the latest version, unmatched, and deleted all but the latest in the driver code, it matched up and "updated" just fine this time. (That is why I left the toggle off for asssuming the latest version so I could force an update just to test)
I have Mr. Trouble on the new driver, since last night, with the Aqara Detection enabled and so far I have had only 2 false events I can probably blame one of them on me.
It's now in dual mode as I have given up on the PIR only in that room.
Plus the empirical evidence that it's only really there to keep the mmW in check.
Just noticed that when Hubitat is rebooted (did so to install the latest update), all the FP300 turned to active (occupied), and had to do Spatial Learning again to "reset". Is there a way to get around this odd behaviour, instead of having to reset all FP300s everytime hubitat reboots?
After a reboot, it might take a little bit of time for the FP300 status within Hubitat to get sync’d back up, especially if the sensor status changed while the hub was rebooting.
I have multiple FP300 sensors connected to my HE hub, and have no issues whatsoever. My hub gets rebooted all of the time for beta release updates.
Wierd ... I have been struggling with one of mine - I did the update right after you released it ... went back to both PIR and mmWAVE ... still spotty. Then about a week ago, we noticed it behaving much better. More reliable triggering and remaining on when we want and off when we r gone. Hmmmnn ... Dan I think you are getting it if not already - many thanks.
Since the last change 17 days ago my 2 PIR started behaving and the "impossible" en-suite one, since converted to mm/PIR has been flawless. Boy that thing hated being a small bathroom.
All - just a heads up that some users are reporting issues with the latest FP300 firmware that was just recently released. I have not tried it, as my sensors are working well. Please be aware that there is no way to revert the firmware on these sensors. I recommend you hold off on performing any firmware updates on any sensors that are critical to your home automations.
If anyone has a spare FP300 sensor, and wants to give the new firmware a try, I'd love to hear your feedback. To perform a firmware upgrade on the FP300, I would factory reset (do not REMOVE it from your HE hub) the sensor by pressing the button on it ten times fairly quickly. Then use the Aqara mobile phone app, to change the firmware to Thread and then back to Zigbee. This will update the firmware. Finally, re-pair the sensor to your Hubitat Hub and it will jump back into the existing HE device on your hub. You will need to wake up the sensor to CONFIGURE it, and then to run SPATIAL LEARNING again.
Personally, I am going leave well-enough alone for right now. My FP300 sensors are working well and I'd rather not mess things up. YMMV, of course!
I just got a new FP300 and was able to add it with no real issues - it downloaded the current firmware as part of the setup process.
The only issue I have was the blue light never turned on when I pressed the button which was a little confusing during the save settings and learning process and I had to press refresh to get it to detect anything, but since then it seems to be working normally (it’s only 15 mins in tho).
Looks like @kkossev has an early release that fixes the Zigbee issues caused by this new Aqara FP300 firmware version. Great work @kkossev! I will look to merge his changes into this dedicated driver over the next couple of days and will then do some testing. I am currently busy with some other more urgent projects around the house, so it'll be a couple of days before I can really look into this.
For now, please DO NOT update the firmware on your FP300 sensors if you're using this dedicated FP300 driver. I'll let you all know when it is safe to do so.
I have been an enthusiastic user of the Aqara FP300s since released. While not perfect, I find them overall very reliable and have 21 installed in my home. I was attempting to install two more recently and have been frustrated by my efforts as they work fine when paired in Aqara, but lose motion detection in HE and only provide temperature/humidity/lux levels. I was originally using the Aqara P1 Motion Sensor/Aqara FP300 Presence Sensor driver, but also tried the dedicated Aqara FP300 Presence Multi-Sensor driver. Same result with both drivers.
Reading about the recent firmware update, now it all makes sense as these two have the newer firmware 6542. So, if anyone needs a test case for an updated driver, let me know as I have a mix of the old and new firmware.
I appreciate the offer. I am working with @kkossev on testing his changes to this dedicated driver to support the new FP300 firmware. I hope to have something ready for community testing within a few days.
However, if you're already using @kkossev's Aqara P1 Motion Sensor/Aqara FP300 Presence Sensor driver, you should be able to give his updated driver a try to see if it resolves the issue you're experiencing. Please see this post in his thread for details on how to test his revised driver - [RELEASE] Aqara Multi-Sensor FP300 (PS-S04D) - #273 by kkossev
Just a quick status report. Aqara has changed some significant things in their latest FP300 firmware that make the device eventually disconnect from the Hubitat platform when paired directly via Zigbee.
PLEASE DO NOT UPDATE YOUR FIRMWARE ON THE FP300 sensor! Doing so will result in a pretty unhappy experience. The FP300 firmware version to avoid is 0.0.0_6542.
@kkossev understands much more of what is going on at the Zigbee level than I do. I wish I had the necessary skills to set up a Zigbee sniffer to try and help debug this issue, but alas, I do not. @kkossev was able to work with the Hubitat Engineering team to make a change to the HE Zigbee stack to help resolve one of the issues that he had identified. Unfortunately, this did not fully resolve the issue with this new FP300 firmware version. I want to thank @kkossev for all of his hard work digging into this issue thus far. Hopefully a solution will be found in the near future. I would also like to thank @mike.maxwell for the recent Zigbee change to support Zigbee Time Cluster (0x000A) read attribute requests. Hopefully, between @kkossev and @mike.maxwell, a solution can be found.
Thanks - these are some of the exact same changes that have been added during testing. While helpful, they do not resolve the issues fully on the Hubitat platform.
