Ok, thanks for the reply – can you confirm I’m understanding this setup correctly?
Fan #1 = Always worked – is power monitoring off on this one?
Fan #2 = Stopped working – but when power monitoring was turned off, it has continued to work?
Fan #3 = Worked, then didn’t work – was power monitoring turned off here?
Sorry, just trying to get all the facts straight. I thought maybe we were onto something here with the power monitoring fix
Yeah, I can promise we won’t be a stickler on the return window with these. There’s clearly something happening that we need to figure out, so until we do, the return window is open indefinitely. Thanks for hanging in there with us!
Fan 1 - always worked. Power monitoring is off now but was on before and no issues.
Fan 2 - was always intermittent, I have the antenna outside the shroud on this one. Power monitoring is off and as of this morning its still connected. Hasnt been 24 hours yet.
Fan 3 - Power monitoring is turned off on this one. Worked after I did a breaker reset. Turned the fan off this morning when I woke up but now I can no longer turn lights or fan on or off.
I have 2 more switches in boxes here. I’m going to swap fan 3 out and see if I have better luck with a completely different switch.
Also to note, Fan 1 and Fan 2 have always maintained a paired connection between module and switch, just no comms. Fan 3 has lost the paired connection twice. Not sure if that info helps at all.
So I have been running an experiment for the last 24 hours unsuccessfully trying to replica the issue. My power monitoring is still obviously on.
Every 5 minutes automation
Verifies Power monitoring indicates power is off
Turn the fan on for 45 seconds
Verify power monitoring indicates fan is on
Turn Fan off.
For a few hours during the time the power was off, I had the four Ilumin bulbs randomly change color every 15 seconds. The goal being to generate a lot of z-wave traffic (the bulbs are updated to 2.30 beta firmware, so they aren’t as noisy as they could be.) Still I was unable to force a disconnect.
One thing that it did get me thinking, could having the switch being a repeater contribute to the likelihood of this happening? Do those having a problem know if their fan/light switch is likely to be a repeater? An interesting test setup might be having the switch setup between the hub and some noisy bulbs and see if that helps trigger?
For reference, here is my hubitat RM doing the testing:
Just an update … the switch stayed connected this time for between 19 and 23 hours with the power monitoring still enabled. Considering most connects last for under two, I’m still trying to come up with some change that may have played into it besides the weather.
@Eric_Inovelli I haven’t shut of the power monitoring yet, there’s plenty of people trying that approach with success, so I’m still tinkering with other things to see what else can be learned that might be helpful. I’m curious though, what is the logic on the power monitoring for these? Does the switch poll the canopy module for updates?, or does the canopy module push updates to the switch on a schedule?, or on a trigger (like a calculated consumption change). Or some other process?
@kitt001, @Eric_Inovelli It is my understanding that the switch polls for power reports the canopy only sends on requests from the switch. I tried to find info out on Dr. Google but was not able to locate any. This is just based on looking at the product and playing with the power monitoring from within my firmware settings!
When Fan 3 disconnects, will a 30-second air gap at the switch restore the connection…or will only a breaker reset restore the connection?
I have personally not ruled out that there are two different problems happening, one being in the switch related to data traffic that an air gap will reset, and the other one being in the canopy that requires a breaker flip to reset. Hence, why am I am asking if a 30-second air gap restored the connection…
I’ve had power monitoring off for more than 72 hours. I just bombarded the switch with zwave commands from the hub as well as button presses on the switch by the dozen. While I did see a slow down in the response of the switch in executing on the commands, I could not replicate a failure that would cause the switch to be non-responsive.
I’ll turn the power monitoring back on here shortly to see if I can make it fail that way and report back.
UPDATE: As of 10:43a eastern time, I’ve removed the zero values from the 3 power monitoring options to force the default. I’ve turned informational logging on as well. More to come.
Still running tests on my install from yesterday in ST. It has been about 20hrs connected successfully without any hiccups besides my own, lol, with power monitoring off. I will post the set up after I pull the canopy to retrieve make and model, forgot to take a pic as the young lad was waiting to get the network back up(darn electrician tying my office and dining room on the same circuit).
Another thing that I was thinking about with some disconnects maybe related to humidity as well as heat. I worked for an electronics manufacturing company that built cap lamp assemblies for miners and while testing some units, communications were affected by the humidity under some circumstances causing delays or disconnects from service altogether. A few well placed toroids and changes in conformal coatings helped. Other instances of Industrial HVAC Controllers resulted in the same thing added toroids to reduce the interference by the stray RF signals bouncing around in the boiler room.
Something else that I noticed is that with protection on and the light relay disabled I can still adjust the LZW36 Light Switch physical level via the app, but have associated LZW42s which result in the actual bulbs dimming physically (crazy buzzing) and not via Z Wave which maybe the cause of the buzzing bulbs in other posts… Is there a setting I missed to change this, like association behavior?
Besides that the Google VD works to send a Z wave command from dimmer to bulbs but the Alexa VD appears to adjust the physical dimmer only resulting in reduced voltage to the bulbs.
Looking forward to getting the updated DTH and running it through its paces.
I have now run the power-monitoring all set for 0 for about 30hrs now, without any problems pairing. Before I never had a single day where it would not have lost pairing overnight (and throughout the day when I checked). Came in this morning and the light button worked right off the bat.
I know it is hardly “definitive” but it is another case for something related to the power monitoring, or at least something to look into in more detail.
The “disable relay” feature actually just disables local control of the output load (prevents physical buttons from controlling the load). You can still control the load via zwave with the relay disabled.
My other devices when controlled via app send the message->switch->lights. (->=Zwave signal lol). The LZW36 is not forwarding the Z wave commands to the bulbs but is dimming the physical switch. This is causing the bulbs to drop voltage and isn’t good for them. They are meant to be dimmed programmatically( z wave commands) but it does not appear that you can disable remote control which means that the app will dim the physical switch not send the z wave commands from what I can tell from the logs. Physical interaction with the switch appears to work as expected. I can’t just dim from the app, or Alexa, Google seems to work.
The way I got around this was to bypass the module for the fan lights. I then used association group 3 to keep the bulbs in sync with the light. This way the LED notification shows the correct dim level, and the association group works as expected.
I agree thought that it would be nice to have a virtual switch of sorts. The actions of the switch, the LED strip dimming level, and the association messages would be based on this. Then there would be a parameter that indicates the state of the physical light dimmer, if it was 100% (on), 0% (off) or kept in sync with the virtual switch.
@zavex thats basically how I had it prior to the LZW36. Fan was on a dummy switch powered on LZW31-SN controlling bulbs in association 3. Well I said I’d get a pic of the make/model lol, now I have an excuse. Yeah all of my Zooz switches show scene activation and levels with the relay disabled on the LEDs, and will stay in sync with commands, child devices can be natively controlled as well.
Just a little disappointed after dropping $100 (CDN) on this switch and not seeing too many success stories, that’s why I waited to install it and now will be waiting for the rest of the product to be available(DTH).
Just venting some frustrations, I love Inovelli Products but unless you are constantly watching the boards or a techie you have no idea what changes have been made. The downloads area is a mess and hard to follow, nothing is deprecated so as to what the most current firmware/DTH are for the products. A sign up for email notifications regarding the products you wish to subscribe to for updates would be great.
I am looking forward to updates to this as I am a techie with 30yrs electronics behind me, I love it but the family not so much when lights don’t work.
Another update on my stock setup that still has power monitoring enabled, and all default driver settings … this time the connection has surpassed 30 hours and counting. The weather outside here is 90F and sunny today. The first couple days I was playing with this and had constant disconnects, it was the tail end of a heat wave and 95-98 out … so the grid was probably strained, ever since the temperature drop on Friday, it’s been increasingly cooperative.
There are a couple things that I’d like to try, but now I’m riding this one out as well. Until I get a concrete-ish baseline; there’s no way to know what changes are beneficial. If my other ceiling fan wasn’t 20 feet in the air, I’d setup the second switch so I could have two to play with.
Apparently the forum software is configured to disallow post edits after a certain period of time, so I cannot add this to my install description above from 8 days ago.
Since I am almost certain that the disconnect problem is in the switch and not the canopy module, I unboxed my second (of four) LZW36’s and dressed the wiring the same as the first one…only this time I took a picture to share:
It will slide into the ball mount with the vents up (as pictured), with the wires facing upward (with Wago connectors on them) passing around the side of the bracket and up into the ceiling box and the wires facing downward (with the female blade connectors on them) hanging down into the canopy where there are shortened fan wires with male blade connectors waiting to connect the two.
The new fan that will be installed with this second switch is a Home Depot special Windward IV 52" light/fan:
This (and the 44" version I am using as fan #1) appear to have been made by Hampton Bay, and surprisingly there are several differences between the 44" and 52" model. The 44" model came with pull chains and the reverse switch on the light module housing. The 52" model has no pull chains, the reverse switch hidden on top of the motor housing, and an RF remote control kit (with the same wiring as the LZW36-C) that will soon be listed on eBay after my diagonal cutters liberated it.
One thing that I noticed while reading the fan’s instructions this time is that they recommend coiling the excess wire around the ball socket instead of shortening the wires. This will bury any canopy module tucked into the ball socket behind a major source of possible electro-magnetic interference! My wires in the canopy are obviously shortened, but this could be an additional variable for anyone who follows those directions.
FWIW, this switch is now tucked back into its Inovelli box for at least another week, as I cannot physically hang a ceiling fan and my installer friend is in self-quarantine…