Lost Fan Control migrating from Wink

Google Assistant Relay works quite well to have Google Assistant do Alexa’s bidding :stuck_out_tongue_winking_eye:

I have to admit to using my Google Nest Hub more than Alexa lately, but still dislike having to say “Hey Corporation” instead of a name.

1 Like

I HATE "Hey, CORPORATION." but I am in the exact same boat. Alexa is so obtuse. Google is miles ahead of Amazon as far as understanding dynamic sentence structure. And then even if it does understand, even with all of the millions of billions of skills, it still never has as much information available to it as Google does with the power of that ridiculous search engine behind it.

That is why... I never call Google "Google" but instead I use some of these alternatives from this delightful article.

They came from the reddit post here:

"Cocaine poodle" is one of my favorites.

:rofl: But what if you already have one of those too? Very difficult to integrate with HSM, but still a very effective intrusion alarm nonetheless.

1 Like

OK, thanks.... I got them working... sort of.

Some of the On/Off controls work fine, while others consistently don't work at all. Yet I did the exact same thing for each one (I think).

Here is one example of a program that is not working. I selected the Game Room Fan because it is the closest to my Hubitat hub - so I'm fairly certain it's not a Z-wave reception issue.

By contrast, here is an example of a Rule where the Fan On/Off Control is working:

The only obvious difference is the [FALSE] tag in the Game Room Fan rule... but I'm not sure how that got there and how to fix that... assuming that is the problem. Does anything else strike you as an obvious programming error? Note that Master Bedroom Fan is further away from my Hubitat yet works consistently.

OK, I have a bit more information about my problem. It's starting to look like the responsiveness to Alexa is intermittent for specific fans, not always working or always not working.

I'm still not sure what the problem is. The latest non-functioning fan is my Study Room Fan. Here is the Rule in Rule Machine:

I think it's set up correctly, though I don't understand the (f) [FALSE} and what that means there.

The false means the logic is false. In other words, the IF statement is off, so it is false.

If it was ON, the false would become true, or it should at least.

Maybe the fan doesn't know it is on or off? What does the device information say if you look in the Devices tab. Does it match what the fan is actually doing?

Okay - so I'm just seeing this. Please do a z-wave repair and don't run any automations or use any devices for a few hours after the repair is complete.

Thanks. So I ran the Z-wave repair overnight. I'm not sure if it was supposed to give me any particular message when it was done, but next morning it appeared to be done. I can now turn on the study room fan and most of the time I can adjust the speed but... oh dang, I think I just figured out why I'm not able to turn it off! Stupid me.

OK, maybe I spoke too soon. I had Sun Room Fan Real instead of Study Room Fan Real in the Off position of the rule, so I fixed that but.... i'm still not able to turn off the Study Room Fan using Alexa.

Hmmm, OK, never mind. Now the Study Room Fan On/Off control seems to be working again. I guess maybe it takes a few more minutes than I expected for Alexa to be able to work through the new Rule?

Maybe. I rarely use Alexa (or GH).

I have a suggestion that will help debugging similar issues in the future. When I create/test a new automation, I turn on logging (descriptionText) for the devices involved (trigger & actions), as well as logging for the automation.

So it looks like I spoke too soon. I'm still having problems with controlling the fan speed in my sun room and also with turning my master bedroom fan on/off. Neither of these was an issue before. At least in the case of the Sun Room, I know that Alexa is hearing my command correctly because it graphically displays my command, but for some reason it's not translating that into action via Hubitat.

When I get home from the office, I'll turn on logging as you suggested and try to troubleshoot further.

I suspect that the issues are z-wave mesh related. I'm not sure if you caught this video or not, but it is a must-see for establishing a solid z-wave network with HE:

Let me explain the issue to you. ST and Wink did all their processing in the cloud, and therefore constantly polled z-wave devices for their status (thus masking inadequacies in the z-wave mesh). Hubitat doesn't do that by default; it relies on devices using instant status to report and a strong stable z-wave mesh (or zigbee mesh for that matter). Hubitat does have a z-wave poller app that you can install. If you do so, I would recommend restricting polling to a subset of your devices. Polling can be chatty and takes CPU time. Too much polling can really slow your z-wave network and HE.

For me, the selling point of Hubitat was being fully local. The compromise for that has been ensure that I have a really strong z-wave mesh, with mostly z-wave+ devices.

I currently have 57 z-wave devices, broken up as:

Repeaters:
12 line-powered z-wave devices
24 line-powered z-wave+ devices

Non-repeaters:
18 battery-powered z-wave devices
3 battery-powered z-wave+ devices

When I use z-wave toolbox to look at the routes taken by packets to/from HE to z-wave devices, the majority of routes go through the z-wave+ devices. I don't need to poll. For the few power-reporting z-wave devices I have, I've set power reporting to occur not too frequently.

Off course, none of this happens overnight. I remember going through all manner of frustration when I moved from Wink to Hubitat, with my first reaction to most z-wave related issues being - but it worked with Wink, so why doesn't it work with Hubitat?

The turning point was making a decision to understand the reason for the difference. And having done that, I have z-wave and zigbee meshes that work far better than they had with Wink. I've posted times for some of my automations in the past, but as an example, my motion lighting automations take about 120-200 ms from the trigger to the light turning on. And they work consistently - every single time.

The most effective way for me to wake up in the morning is to have my night stand lamp gradually ramp up in Kelvin value from 2000 (reddish/breaking dawn) to 5000 (daylight) while simultaneously increasing in level from 1 to 100. That's a lot of zigbee data flying back and forth from the nightstand bulb to the Hubitat. Yet, it proceeds flawlessly without issue every morning.

Something similar was possible by integrating Wink into Home Assistant - and it failed frequently - or would have abrupt jumps in level or Kelvin value (meaning intermediary changes were lost/dropped).

So, for me, it has been worthwhile sticking it out with Hubitat (despite my initial concerns) and changing my zigbee and z-wave meshes to work properly.

Ugh. So I set up most of my device (60+) already and now I'm seeing this. Still, thanks, that gave some insights. I actually set up a few devices at a time but I think I added a bunch after the first few days as I got more confident.

Something that occurred to me - my Wink Hub is still on, mostly because I have not removed it yet. I wonder if that might be causing issues. I'll unplug it first thing.

I did try to exclude devices from Wink but was not always successful - I often had to just factory reset the device for lack of a better way to do it. Would this be sufficient to make it ready for use with Hubitat?

I have one Z-wave extender, but in retrospect, all of my Leviton Z-wave switches, which constitute the majority of my devices, should be repeaters themselves.... I mean I literally have every switch in the house replaced with one of these Leviton Z-wave decora switches. This is a 4400 sq foot house with about 50-60 of such switches.

My Hubitat is located on the second floor on a landing that partly overlooks the first floor. It's the most central location I could find and worked ok for Wink. Should I still plan on using extra Z-wave repeaters? I have the Aeotech 6 range extender - do I need more than one? Do I need to have on close to the Hubitat as he was suggesting?

Last but not least, and this is the question that makes me worried about the answer - since I did not strictly follow the "add in batches of 20 per day" rule, do I need to uninstall a bunch of devices and start the whole thing over?

I have to say that, despite the frustration, there are some positive things I already noticed about Hubitat. First, the devices added to Hubitat's network more reliably than to Wink - most of the time on the first try as opposed to Wink often taking 2-3 tries. The second is that, yes, Hubitat is way faster, even when I use with Alexa (which still uses the internet).

What exactly is Z-wave polling? Is did install that App but I haven't used it. Does it do the polling automatically or only when I tell it to?

Some people do recommend that. However, I have two z-wave networks running in my house, and three zigbee networks. And there's no issues.

You need to configure it. And I recommend you configure it to do the minimal number of devices possible.

Does Z-wave polling help me in any way in this case? What exactly does it do, and does polling somehow improve network stability? Sorry for the ton of questions, but I'm learning a lot and hopefully can pass on the favor once I finish this migration.

It literally polls the device for its status. Older z-wave devices did not report their status back to the controller. So if a switch was controlled physically/manually, the Hub was completely unaware of the change in status. Polling addresses that issue.

It does not improve network stability. If done too frequently it can reduce network stability by bogging down a slow network.