I plan on removing it from the door when I get home and waiting for it to go to sleep, since it does eventually after several hours. That way, hopefully, the zwave spam won't be an issue. Out of 6 sensors, only 1 is still behaving REALLY bad. But, 3 more still act erratically. After an event, they'll send battery updates for several minutes. At least it's not hours, but opening a closet door shouldn't cause 30 - 50 battery updates to occur.
The one that's behaving really bad will send updates every 7 seconds for about 3 - 5 hours or so. That's about 1,500 - 2,500 events for a single door opening.
@agnes.zooz I have a suggestion for a slight redesign of the door sensor. Make the part that sticks to the door the empty part and the top part that comes off have the zwave pcb and battery. That way I wouldn't have to rip the whole thing off the door to take it closer to my hub and would make replacing the battery be a little easier too. If the back didn't have writing all over it, I'd just mount it backwards.
Also FYI that the next version of my driver should have the firmware version report in it, too. It will update on the next wake-up interval (every 12h) any time after you save Preferences.
I made the change this morning, just need to wait for a device to wakeup to verify it works as expected before publishing the update.
If it works as expected, I'll add it to the leak sensor driver as well.
Got the firmware updated (1.20). Immediately after updating I was still experiencing the battery issue so I excluded and re-included. I think the problem got worse. Now it's reporting the battery status and the sensor status every 7 seconds.
It crippled my entire zwave network. Other automations were delayed by at least a full minute. I factory reset it and now I'm having a hard time including it again. So far I've had 2 different results. Either the hub never sees it and the inclusion times out. Or, the hub sees it and starts to initialize it but then nothing ever happens. So far I would not recommend firmware 1.20.
EDIT: Finally got the device to include. Problem persists even after factory resetting the ZSE41. I've pulled the battery from the device to prevent it from killing my zwave network.
The person in me that likes data is considering flashing one of the sensors that never had a problem to see how it goes.
If I could actually get one to misbehave reliably I would at least fire up the sniffer and see if the reports are really unsolicited, or if the hub isn't sending an ACK when it should (even then though device is supposed to stop trying after a few attempts), or other.
I'll keep watching it though.. Maybe I'll catch one at the right time.
I setup a contact sensor and a water sensor today.
The contact sensor didn't have the battery reporting issue but the water sensor started doing it. I would wake it up and it would stop but eventually start again.
So I'm watching this carefully. Maybe I should test using their driver that they are testing with.
I have no idea what Zooz is testing it with. They don't usually write drivers of their own (they didn't write the Smart Things drivers for these devices either - @krlaframboise did), and the Hubitat made one isn't out until 2.2.9...
So my GUESS is that they aren't testing it on Hubitat at all, or if they are testing on Hubitat then they may be using my driver?
If not testing on Hubitat at all, then it will probably take someone capturing some logs w/a sniffer while it is happening to prove whether it is the device (Zooz's problem) or a hub issue (Hubitat's problem).
Or maybe it is a driver issue and all these issues will magically go away when 2.2.9 comes out. Dunno? I can't see any evidence that it is a driver problem though - but I could be wrong, of course!
If someone wanted to they could ask Kevin for permission and port the ST driver over and see if it works differently.
I agree. The driver is pretty simple. I also wrote my own driver and we both ended up with the same issue. So I don't think its a driver issue at this point.
Noted! I hope we'll be able to shuffle things around with a simple modification rather than general retooling. I think it makes sense though I haven't seen a Z-Wave sensor with a similar design yet. We'll have to be the first
We have both but they apply the same type of fix which doesn't seem to help in this particular case so we're back to square one and trying to replicate... I'll keep you posted when we have a new firmware file to share.
I sent this in to Zooz support, but I figure I'll post it here too in case anyone sees a pattern or has suggestions. This is a map of my house, provided by my Roborock S4. I drew in the laundry room and garage, so that part of the house is likely not to scale. I recently purchased a C7 hub and was making this map anyway to plan my migration from C5 to C7. But that's irrelevant for now because the C7 has no devices on it currently. The 6 ZSE41s I have are the brown squares. The one I flashed with the 1.20 firmware is number 2. For most of the testing (and all of it after flashing) it was not in the indicated spot. I had it near my laptop, which was sitting on that white space pretty much right in the center between the living room and kitchen. Unfortunately, since I am on a C5 hub, I can't see what route these devices are taking to communicate with the hub.
Last night I went and opened and closed all the doors that have the ZSE41 on them. This was just before 11pm. Here's what I saw in the Hubitat logs.
Garage - outside door. It is still sending battery status updates every 7 seconds as I type this 2 hours later at 1am.
Garage - laundry room door. This is the one I flashed with 1.20. It completely kills my zwave network now so the battery is disconnected.
Front entry door*. This reported a single open, a single close, and then nothing. So 100% as expected. Though, I have seen it send multiple reports in the past. I'll keep an eye on it.
Joel's Room - Closet Door. This reported a single open, a single close, and then nothing. So 100% as expected. I think this one has been behaving fine the whole time.
Office - Closet Door. Same as number 4.
Master Bathroom - Closet Door. Same as number 1.
After I typed the above, and I opened and closed the front entry door, number 3. The open was at 2021-09-13 01:00:33.760 am and it reported the battery status every 7 seconds for about 3 minutes. The last report in the log was at 2021-09-13 01:03:20.229 am. This is the behavior I've been seeing from at least this sensor, but I'm not sure about the others.
So out of 6 sensors remaining, 2 of them are operating as expected,1 is operating sorta as expected, 2 are misbehaving but eventually go to sleep after several hours, and 1 sends so many reports it kills my entire mesh. Even if the hub didn't send an ack, I wouldn't expect the sensor to keep trying until the battery dies.
The above is basically what I sent to support last night. Here are some observations.
Nodes 4 and 5 are working with no issues and 3 only has slight issues. These are the closest to the hub. Though, node 6 is having serious issues and it's about the same distance. Nodes 1 and 2 are having serious issues and they are the furthest from the hub. However, there is no ZSE41 that is not really close to a mains powered switch that can act as a repeater. I really wish the C5 could see the route the sensors are using.
I am just setting up my Zooz ZSE41. I am seeing the same results as you with the repeat battery events. The statistics show 676 events for a window i never opened since i installed the sensor. Opening and closing the window has generated 40 events so far.
The Sensors that do not route through another device seem to not be doing this at the moment.
I have a ZSE41 paired, not sure where I want to stick it for testing (before I possibly get more). Anyone have anything that works for semi-permanent mounting? I tried some basic double sided tape but not sticky enough. The included mounting strips look like heavy duty 3M that will take paint off stuff. Thinking maybe some command strips might work, just would have the little pull tab thing hanging out? Anyone else get something to work?
Also, I am NOT having the battery reporting issue. Looks like maybe a little when I first paired, then it stopped. I do have a C7 and a separate zniffer setup so I could get some good info if it were to happen. Mine is making a hop through the light switch in the room it is located in correctly.
I use command strips on all my sensors now. The 3M permanent tape all the manufacturers sends with sensors holds fine, except you ruin your paint if you ever have to move or remove one for replacement.
If the placement can handle a slightly bigger gap/bump off the wall without looking too weird or being vulnerable to getting knocked around, the velcro-like Command picture strips work nice...
They grip like strong velcro, but permit easy removal for battery swaps etc. With the added strength of a velcro-ish connection, you then may be able to hide the pull tab "behind" the sensor show it doesn't show.
I just took the sensor that I flashed with the 1.20 firmware into the same room as the hub and put the battery back in. It immediately connected to the hub and is sending open/close events and nothing else. Of course, it could still be routing through a repeater; I can't see that. So my best guess is it happens with a weak signal.
