The day after installing 2025-11, and I'm not sure if it's just a coincidence, both thermostats stopped replying to Z-Wave pings and were marked as dead nodes. I've tried all the troubleshooting steps: moving the USB, rebooting, etc. Then bit the bullet and factory-reset one of them. That didn't help either; neither of the provided methods works for inclusion. QR not an option, this is a pre-Smart device. Default, S0, and unsecure options for manual inclusion don't work. Nothing in the logs (diag level) after "The controller is now ready to add nodes" aside from the GetBackgroundRSSI. Please help, I've been at this all night.
#ZWA-2 stopped communicating with my thermostats (TH6320ZW2003)
1 messages · Page 1 of 1 (latest)
Any other devices that still work?
Can you share Z-Wave logs of the communication attempts on loglevel debug? Ideally including the addon startup.
there are 3 others that still work fine.
Above are the last 5k lines of the logs.
And thank you in advance for taking the time to help.
Channel 1 and 2 not looking good:
2025-11-09 09:58:13.170 DRIVER « [RES] [GetBackgroundRSSI]
channel 0: -113 dBm
channel 1: -76 dBm
channel 2: -76 dBm
channel 3: -109 dBm
Are you using a USB serial extension?
Yes. There's an active USB extender in play
However, I didn't lean towards this being a reception issue, because things worked fine for months, and I did try to include the thermostat right next to the transmitter since it's easily removable.
OK. I'm not sure what and how as I didn't do anything to HA in the last 12 hours, but the second thermostat reconnected and I was able to include the factory-reset one on the first try. ...
May be it was due to the channels though: 2025-11-09 16:00:23.505 DRIVER « [RES] [GetBackgroundRSSI]
channel 0: -113 dBm
channel 1: -110 dBm
channel 2: -110 dBm
channel 3: -91 dBm
this should include whatever went "right"
Channels 2 and 3 are the 40kbps / 100 kbps channels, I think. -78 dBm is too high for normal usage.
Why would the tuner bump up the gain that high?
That's a measurement of basically the noise floor.
It means any messages from devices that are lower signal power cannot be heard.
Thanks for the definition. How does the floor drop 30 db on is own
It's a negative number, so it's actually rising. Often that is because of some RF interference or jamming. Nothing obvious in the logs though.
Near -110 is good/expected for the ZWA-2.
E.g. at the end of the last logs:
2025-11-09 16:07:53.514 DRIVER « [RES] [GetBackgroundRSSI]
channel 0: -113 dBm
channel 1: -109 dBm
channel 2: -109 dBm
channel 3: -96 dBm
Is quite good. The LR channel (3) usually seems to be higher.
I could move it to the center of the attic, away from all the electronics. Once esphome fw is out of beta
Sometimes moving just a few cm is enough for the noise to change significantly. Unfortunately, seeing the noise levels change takes a bit long currently. I've got something in the works to make finding a good spot less time-intensive.
I had a metal vase on the same shelf in the closet, when I removed it background went down instantly -5 dBm. Then I did something else, like inadvertently changing the orientation of the USB stick and it went back up.
Thaaaat's finicky. I'm also curious if this might be due to another project in working on. I've got an rtl-sdr stick that I'm trying to get stable gas meter readings though, but that's on 912Mhz
That would be channel 3
The plot thickens... Will these two play nice together!?
I feel like I need some educating here. The data sent over RF; do the packets behave like broadcast? That's what I'm expecting them to do, but if the packets behave like TCP, then all this is self-explanatory as I've stopped getting readings from rtlamr when thermostats came back online.
Well. Are the thermostats using Z-Wave Long Range?
no, they're practically antiques. Zwave Plus v1