I used the pull chain (High->Medium->Low->OFF) for fan speed and this triggered ZERO power reporting from the switch as seen in the screen shot.
Looking at the event History the switch must have locked up some time after we put my daughter to bed, as the last activity was at 9:15pm and then you can see my interations at 10:30 when I detected the switch is not working
Hi @Eric_Inovelli,
Any thoughts on how the antenna wire should be oriented?
Pointed at the switch, perpendicular to the line to the switch, vertical (parallel to the switch)…?
For a basic antenna, the RF signal comes out the side of the antenna in a doughnut shape (as if you stuck the antenna through the hole of a doughnut). For best results, the tip of the antenna should be pointing perpendicular to the switch.
FYI, mine finally locked up again, after trying to make it fail since Sunday. I’m soooo excited…in a geeky way!
Power usage does NOT appear to be updating in Hubitat, even if I poll the switch.
Changes from the switch and Hubitat are showing up on both.
More importantly, part of my test was leaving breeze mode on…and it is still working! My rationale is that I suspectbreeze mode executes in the canopy module, and this shows that he canopy module’s processor is still running even though there is a communications issue somewhere between the switch and canopy.
Any other tests that I should run before I air-gap the switch to see if it restores the connection???
Replying to myself, let me also add that I noticed that the light/fan were not responding after stopping my Hubitat light/dimmer exercise loop and trying to turn the brightness up to 100% from the Hubitat dashboard. The Hubitat dashboard button icon turned into an hourglass when I changed the brightness to 100% from it, and didn’t change the light’s displayed state or toggle the state on/off for at least 10-15 seconds worth of clicking before the switch and Hubitat were back in sync. Just throwing it out there because I am still personally suspecting a problem on the switch side of the link where some Z-Wave event is causing the RF radio to the canopy to drop connection. Air-gapping the switch to restore the canopy connection makes me suspect the problem is there and not in the canopy module, and seeing breeze mode still operational appears to be a sign that the canopy module is still running (sans link to the switch)…
Jumping on this thread with a “me too” … I’m having the same issue. Bought three, installed one … and now I’m trying to solve this also.
Like a lot of other people on here, I also have a Unifi access point with quite a heavy 2.4 ghz load (probably around 30 devices) … but I don’t feel like this is the most likely suspect. My thinking here is that once connected, these things work for a random period, meaning that if it is an outside force that’s breaking the communication; obviously, we all need to have it in common, and it needs to be transient for all of us.
If it’s the 2.4ghz environment signal it would either need to be flirting with barely connected from the get go, or disrupted by the normal process of devices coming and going. My 2.4Ghz environment doesn’t change much, and it’s unlikely that we’re all sitting at the same congestion level that causes a “barely there” connected state between the switch and the canopy where one more device confuzzles the communication … though anything is possible
I’m wondering if the thing that we have in common is related to the actual power. Perhaps the switch, or the canopy module is unbuffered to minor power fluctuations (most likely the switch, based on the thread), and the 2.4ghz hardware in one or the other hits a bad state and gets stuck when the power dips out of range. In my house, when ever the central AC kicks on, due to the load change, there is a visible diming of lights, etc … so I know that the switch feels that too. Power fluctuations are pretty common and it would explain the randomness that’s being described. I’m still trying to work out a test, but even if power is the root cause, proving it could be hard without the right equipment since it could be related to severity and/or duration. So … I’ll ask first - is anyone else consciously aware that they have power fluctuations?
I just air-gapped the switch for 60 full seconds, and it restored the connection with the canopy module.
When connectivity was lost, the light dimmer was at 15-20% and the fan on breeze mode. Before I pulled the air gap, I set the light to 100% and fan on high using Hubitat…and it was shown on the switch’s LED status bars. About 2 seconds after I pushed the air gap back in, the light went up to 100% and the fan went into high mode.
One extra variable that I am going to throw into the equation for @kitt001 is that we had a slow-moving blob of lazy thunderstorms hanging overhead for about three hours during the time that the switch and canopy module stopped communicating. There were no near strikes, and I did not hear the relays click on the UPS that I was sitting next to or hear the beeper on the bigger UPS in the West Wing Server Room.
One of the northeast power grid’s major interconnect points is a couple of miles away as the crow flies on the same road that I live on, and our power is usually rock-solid having the big grid, local substation, and distribution lines all within a few miles of the main sources for NYC and the entire northeast. Not going to say that it’s the cleanest power in the world, but we seldom see spikes or sags.
My test included 4 days of running the fan in breeze mode, which I’m sure blipped the canopy module’s power whenever high speed kicked in. It was also running this Hubitat RM task repeatedly for the past three days, which I’m sure also probably threw some spikes at the canopy module flipping on/of and changing dimmer settings:
Summary: Air gap for 60 seconds restored connectivity, and switch immediately sent settings changed during the communications failure to the canopy module as soon as they were able to communicate again…
The last power report indicated 7 watts - this corresponds to the lights being on, but not the fan. I run with local protection enabled for the Light only.
It was a cooler day here, so room temperature was probably around 74 - might have been warmer at the ceiling.
After taking the wiring pic, I moved the antenna outside the cover, so it was sticking out.
I don’t use Ubiquiti APs, but i do have 2 2.4 GHz wifi networks, one for upstairs, and another for downstairs.
Pulling the airgap fixed the issue.
I was still seeing Z-wave commands being processed.
I have 3 of these installed and of those 3 only 1 is still responding. Breaker reset worked for the first day or two but now they wont come back at all.
Crazy request for someone who is having frequent/repeatable disconnects. Can you disable power reporting and see if it still loses connection?
I have no idea how the LZW36 was engineered under the hood beyond the long project thread, but have a hunch from prior experience that the switch’s 2.4GHz receiver might be overflowing its buffer under certain conditions when the canopy module sends status updates quicker than Z-Wave can repeat them out. I have no idea if this setting stops the canopy module from sending them or just tells the switch to toss them as received, but figured that it is worth a shot in the dark for someone with frequent failures while we all wait for the factory’s responses in the 24-hour timezone turnaround cycle…
@vreihen - I can try this out as soon as my switch loses connection again. I am trapping some logging information now. I’ll reset it and then disable the power reporting and continue to monitor logs
@Eric_Inovelli - Random thought, is it possible/worthwhile to create a debug version of the firmware that could output some form of diagnostics either on demand or under certain conditions? Maybe with a corresponding driver update for compatible hubs to update log the information?
After what amount of time does the power update for you? I had the light and fan on and it was showing 90 watts. Made the changes. Saved. And I turned off the light and fan. Still showing 90 watts (which is what I’d expect) and it has been 8 minutes.
these might have to cycle all the way though to update. when I initially changed the reporting frequencies on the 4 in 1 sensor. it didn’t setup until those values did the full cycle.
Does your fan have pull-chains or is it remote operated? (If pull-chain, make sure to set it to high)
Pull chains, set to high
What wiring schematic are you following? Schematics can be found here: LZW36 Wiring Schematics – NOTE: Please let us know if you have any alternate wiring
I’m a little confused by those - unless I missed it looking through a few times, none of the options were pull strings for both lights and fan My fan already had a receiver wired up which I replaced with the Inovelli one (they were pre-labeled, which made it easy to swap out)
How far from the canopy module is your switch?
10~ feet
What is the estimated temperature in your house?
73F usually, or lower
Can you give pics of the canopy installation (show how it’s sitting in the canopy and how it’s wired?
Sorry, can’t get to it (had to rent a ladder and van to get the ladder home to reach it)
Do you have any access points near the fan?
I have the TP-Link Deco mesh network (5 Access points, mix of M9 and M5, with a M9 Plus is closest to the switch, about 10ft away)
Would you consider your 2.4 GHz network large or overworked?
The majority of my devices are on 5ghz, maybe only 4-5 devices remaining on 2.4ghz that I can think of
I installed the switch a week ago, and it was fine until today suddenly (turning on/off lights/fan usually a few times a day). I tried turning the fan and lights on/off multiple times from the physical switch, but no response. The switch bars did show like it was turning on/off. After pulling the air gap for 30s the fan began responding again