Or systematically work your way from the dongle outward with your fuses, test, enable, test, enable. You might have to rebuild routes in between but generally not.
1 Like
Whatâs different about this one? Brand? Type? Itâs odd the furthest away is the best one.
You mean just one of them could be the cause of the whole problem?
When I was first starting this journey, I almost went with the Lutron Caseta. The cost steered me to Z-Wave.
Yes. Itâs happened to me a couple of times.
Z-Wave is pretty solid.
Itâs the one I just posted that got 0/10 on all the health checks. Yet it works the best. Itâs a Zooz ZEN15
I mean, it works great but ZWJS says it should suck. ZWJS may be wrong or you are really lucky.
So just remove one at a time and look for improvement? How did you discover which ones were âchattyâ?
I sometimes wonder if the materials my home is constructed of is part of my problem.
Look at the Z-wave log in the same UI. You can open it in a separate window.
- Check the tx,rx, timeout stats for each device in zwavejsui or enable the sensors in HA. Chatty device will have high RX.
- How much RAM is on that NUC?
You mean just by hovering over the far right? The most chatty device is the light switch in the room the hub is in, and itâs a repeater for several other devices. The TX & RX numbers are about equal. The rest donât seem bad, but then again, I donât know what bad is.
But the one device that is WAY off from the rest, ironically, is the most reliable device. It has 162TX & 1400+RX! Itâs the Zooz ZEN15.
the nuc has 32gb of ram.
You can start then by adjusting the configuration on the zen15 to send power readings much more infrequently.
All those counters reset to zero when you restart zwavejs. So restarting, waiting an hour, will make it easier to understand. For example, 1400 in an hour is way to many, 1400 is a week may be fine.
I enable those counter in HA in the integration device panel.
Then have them all in a dashboard
Where exactly is that stick located? Ideally Iâd like it 5 feet off the floor, not anywhere near the brick wall, microwaves, WiFi base stations, etc.
Presently, the stick is in the middle of a window. However, the window is in the middle of a brick wall, though. It is only a few feet from my router, so moving it away from the router will not be easy. But it is 5 feet off the ground.
I do have an idea of a place to move it, but that wonât happen soon. Itâs somewhat part of a remodel planâŚ
Re: the brick wall: does the brick wall cause problems? This is part of my question about the building materials my home is constructed of; lapboard and plaster walls is also pretty dense and Iâve wondered if this causes problems for the zwave?
I just disabled the power settings in the ZEN15 to reduce its chattiness. I donât need them.
If the window is metal coated I would not be surprised if that gives nasty reflections.
But a whole diffrent question: do you have devices with S0 security? That is to be avoided - you can better have insecure than S0. S2 security is fine.
I had tons of trouble with S0 because S0 is very chatty in itself, with very big messages that are easily corrupted or overloading other devices. Once I stopped using S0 I had rock solid experience - before it was as bad as you describe.
The windows are all composite and glass.
All of my devices are S2 except the ZEN15, it has no security. And again, that is the device that works the most reliably.
I have a big mix of older S0âs with my Sxâs and it works pretty good, but Iâm sure it depends on the manufacturer too.
I would definitely try relocating the NUC / stick even temporarily to see what affect it has. Then heal/repair each line powered device one at a time, starting with those closest to the stick.
Brick is not good, itâll absorb lots of signal.
At this point my thinking is.
- you have lots of devices, so it should be possible to get a good mesh
- you have replaced the stick and that didnât improve it, so itâs not a bad stick.
- you are tracking down and reducing traffic, keep at this, and look at the zwavejs debug logs, you should have minutes of no activity.
If you move the stick and it makes no difference. Then at least you know that is not the problem.
1 Like