Alexa Rooms Intermittent

I just switched from SmartThings to HE. I have a room with 4 Z-wave Plus switches (GE/Jasco Enbrighten in-wall). I also have the Alexa Hubitat skill enabled. I'm able to individually control each of the 4 switches through Alexa and/or Hubitat. Individually, they work reliably regardless of how I operate them (HE or Alexa). The 4 switches are put into a single Alexa group and when I trigger that group, I get 2 or 3 of the switches activated -- and it's a different 2 or 3 each time. There is a substantial delay between them (1-2 seconds is typical). But almost never are all 4 switches activated. I did have to turn on the "respond immediately" toggle in the Alexa skill otherwise I would get "device did not respond" messages.

Any ideas on how to fix this?

I don’t know. This could be a symptom of mesh inadequacy. But before we go there:

  1. Have you updated the zwave radio firmware on your Hubitat? If you go to Settings -> Z-Wave Settings, and you see a button called “Update Firmware”, it means you have not, and should do so.

  2. Does the same thing happen if you try to control the four switches as a Hubitat group, i.e. a group created on Hubitat using the Groups and Scenes app?

@user4371 Can you also post your z-wave details page in it's entirety?

I tried to upload a screenshot of the Z-wave page, but this website didn't allow that. . One thing that I note which seems significant to me is that the two devices that routinely have problems both have the S2 level of security. Those two devices are unreliable through Alexa, both when commanded as part of a group (Room) and when commanded individually. In the later case I've checked the Alexa log and it does correctly identify the device -- so there wasn't some voice transcription error involved, the command was simply ignored or failed by Hubitat. I've poked around Hubitat and I haven't found a log of commands received through the Alexa skills app. This would be a good addition for debugging problems like this.

If somebody could tell me how to upload the actual web-page that would be great!

Join the owners group Hub owners - Hubitat to allow posting and uploading of pics and then simply use windows built in snip tool to snap shot then paste

It looks like a lot of your switches are of the older non plus type and they're down at 9kbs and at best some are 40k. This is a serious bottleneck. You could throw some repeaters in but they still won't be that good. Ultimately the bottlenecks are causing the delay.

I would use groups to group the lights together with a single activator. That should stop the popcorn effect and the odd one not turning on.

Take a look at this article... Look at the z-wave section. Are you new to home automation or Hubitat or coming from a different platform such as Wink, SmartThings or Vera? This may help!

@user4371

@rlithgow1 may be correct, although I'm uncertain how to determine plus/non-plus from the data you've provided, especially because there are S2 Authenticated devices (which have to be z-wave+) that connect at 9.6 kbps and 40 kbps.

This uncertainty aside, your z-wave table is revealing in one major manner. It appears that your hub is not positioned well relative to your devices. The only node to which your hub is connected directly is 0x0A (Office). All other nodes are either routed through 0x0A or 0x0E (Garage). Further, 0x0E itself uses 0x0A as a router. This situation is exacerbated by the RSSI to 0x0A being rather poor. It is only 4 dB.

This is one of those rare instances where it may be beneficial to "Repair Z-Wave" after you re-position your hub. And then wait several hours to let the z-wave mesh sort itself out.

I'll also tag Hubitat's chief z-wave engineer (@bcopeland) and a few community members who are well-versed in z-wave (@csteele, @dennypage, @JasonJoel) for their input.

Your z-wave table indicates that you have updated the hub's z-wave firmware. Could you answer my second question from earlier, quoted below:

2 Likes

Looks like you have a lot of nodes bouncing around a lot. Lots of route changes. Nodes 17, 19, 20 and 23 all have 10+ route changes. Node 19 has 25.

My first inclination would be to run a repair followed by a power pull reboot of the hub (shutdown hub, unplug it for 30 seconds, replug).

What happened with node 13? Was it explicitly excluded?

3 Likes

I used the Hubitat scenes to group the four lights and triggered them manually through Hubitat -- no Alexa involved. No difference.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.