I've turned on debug logging to see if I can get any errors logged. Is there any other common cause? I don't think RF distance is maybe likely since opens seem to always work. I thought maybe the ball is sticking, but not sure how to really confirm if thats happening (I don't think so since I hear it moving around fine)
What's annoying is some devices appear to have a "refresh" option you can manually click but this doesn't. Here is a picture of it closed, but when the failure happens it's left open.
If this is mounted how I think it should be, there is a potential for it to be roughly 10-12ft further away when close vs opened. Might want to provide a snip from the Zwave details page showing its route and signal info in case anything jumps out there.
Battery powered sleepy devices can’t be manually refreshed as they are not actively listening for commands.. You can only send commands when the devices has initiated a wakeup notification.
I had a tilt sensor long ago but it was unreliable. I think it was the ball, and yours may be as well, yours sometimes failing on close events could be potentially be caused by the movement of the ball differing from open->close vs close->open. I changed to a contact sensor, which can be a challenge to figure out how to mount on a garage door but I solved it with 3 angle brackets (see photo). It has never missed an event in the ~year I've used it.
Thanks all for the suggestions. I actually do have reed switches also on the door(s) but long story with using those (see my other posts here about my garagedoor app). To be clear though, that's not related to this issue since they're not connected to anything. I put my tilt sensors at the tops of the doors also just FYI.
One common (and effective) solution is to de-solder and replace the metal ball switch with a mercury tilt switch. There are several of us here who have done that.
Here is logging from this morning. No failure today when I closed the door just a bit ago. At 7:45 it looks like it sends out a but of useless battery messages. I wonder if sometimes those collide with a close event? Here it didn't.
So I am trying moving the sensors a bit. Like this. So far so good for a couple days. You see the sticker up to the right where it used to be. (you also see that I have a reed switch there but it's not hooked to anything) My hope is that the issue was RF interference from the rail.
