LZW31-SN turns off randomly

Hi @Dbolot1 Welcome! I personally would update the FW (here). The changelog reflects that this is one of the issues fixed.

I had the same issue in my 3 ways. Prior to the new FW being available, I opted to replace the dumb switches in my 3 ways with add-ons. The ability to set the level at the add-on as well as the smart dimmer was well worth the extra $20 to me. The added benefit, the add-on switches also fixed the issue you are describing.

Besides fixing this issue, the new FW addresses the persistent notification issue… As well as others.

Thanks for reply, will look into getting smart remote as well as updating firmware, just need to get a z-wave stick. Any suggestions on a stick? I was looking into Z-wave.me.

I’ve seen that one recommended somewhere on this site before. I’ve been using one made by Goodway and have had good luck with it as well. I believe there are several other models that would work.

I am experiencing these exact symptoms with my LZW31-SN on a Hubitat Elevation. It seems that my only way out of this is to set Parameter22 to 0 instead of 1, and keeping Parameter21 at 1. Of course, I lose 3 way dumb switch capability but the light stays on. I’ll have to give the GE Companion (Auxiliary) switch a try when it comes here from Amazon. I suppose Inovelli will have to push a firmware update to fix that bug (will cut off randomly if Parameter21 = 1 and Parameter22 = 1: The settings for 3 way dumb switch).

I am having exact same issue. Any solution so far? I have used GE switches before in my previous home. never had any issue. Innovelli switcher definitely gave me more trouble, including this random turn off issue. I am wondering why it is still not fixed after such a long time.

I had the same issue (Strange behavior Red Series Dimmer - 3 Way w/neutral) and the suggestion by @Eric_Inovelli of setting Parameter 6 to 80-85% fixed it for me. I now have the parameter set to 90% and it’s been working for about 2 months now.

@oldgarfield - I’d recommend flashing to the latest firmware. It should be resolved. Otherwise, as @rakeshg mentioned, you can set maximum output to 80-85% (YMMV) to eliminate the switch from turning off.

Hi all, I’m running a LZW31 with latest FW (1.48) and still having this issue. Setup is the same: 3 way, w/one dumb, and tabs removed on one side. After a few minutes, the switch just turns off.

I posted in the HA forum on how to obtain detailed logs, LMK how I can help debug…

Hey @nic – I know the latest firmware should fix this, but just for giggles, can you set the max to 80-85% and see if that helps?

Other questions if you have a second:

  • What is the load on your switch (approx wattage)?
  • When the light turns off, does the switch go dead or is the LED bar still lit up?
  1. Load: 9.8w x 6 = 58.8W
  2. The LED bar is NOT lit up when switch turns itself off. Cycle is: OFF (bar is dimly lit) -> push ON (briefly lower quarter of bar lights up, then full bar lights up) -> turns OFF (bar is dimly lit)

I’m having the damnedest time getting HA to show the config options, so I can’t get into the 80% game right now.

Logs are below:

  1. Turn off via dumb paddle
  2. Turn on via switch, watch it turn itself off
  3. Turn on via switch, watch it turn itself off
  4. Turn on via dumb paddle, watch circuit remain on.


# Switch Debugging:

== Turned circuit off via dumb paddle switch (switch in down position):
2020-09-30 08:49:08.446 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x00, 0xb5, 0x00, 0x6f
2020-09-30 08:49:08.447 Detail,
2020-09-30 08:49:08.451 Info, Node012, Received SwitchMultiLevel report: level=0
2020-09-30 08:49:08.451 Detail, Node012, Refreshed Value: old value=99, new value=0, type=byte
2020-09-30 08:49:08.451 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:08.451 Detail, Node012, Notification: ValueChanged
2020-09-30 08:49:26.079 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x00, 0xae, 0x00, 0x74
2020-09-30 08:49:26.079 Detail,
2020-09-30 08:49:26.079 Info, Node012, Received SwitchMultiLevel report: level=0
2020-09-30 08:49:26.079 Detail, Node012, Refreshed Value: old value=0, new value=0, type=byte
2020-09-30 08:49:26.080 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:26.080 Detail, Node012, Notification: ValueChanged

== Turned circuit ON via LZW31 switch
2020-09-30 08:49:27.122 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x63, 0xb1, 0x00, 0x08
2020-09-30 08:49:27.122 Detail,
2020-09-30 08:49:27.122 Info, Node012, Received SwitchMultiLevel report: level=99
2020-09-30 08:49:27.122 Detail, Node012, Refreshed Value: old value=0, new value=99, type=byte
2020-09-30 08:49:27.122 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:27.122 Detail, Node012, Notification: ValueChanged

== Switch turned OFF by itself
2020-09-30 08:49:27.958 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x00, 0xb1, 0x00, 0x6b
2020-09-30 08:49:27.959 Detail,
2020-09-30 08:49:27.959 Info, Node012, Received SwitchMultiLevel report: level=0
2020-09-30 08:49:27.959 Detail, Node012, Refreshed Value: old value=99, new value=0, type=byte
2020-09-30 08:49:27.959 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:27.959 Detail, Node012, Notification: ValueChanged
2020-09-30 08:49:39.752 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x00, 0xb0, 0x00, 0x6a
2020-09-30 08:49:39.752 Detail,
2020-09-30 08:49:39.752 Info, Node012, Received SwitchMultiLevel report: level=0
2020-09-30 08:49:39.752 Detail, Node012, Refreshed Value: old value=0, new value=0, type=byte
2020-09-30 08:49:39.752 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:39.752 Detail, Node012, Notification: ValueChanged

== Turned circuit ON via LZW31 switch
2020-09-30 08:49:41.421 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x63, 0xae, 0x00, 0x17
2020-09-30 08:49:41.421 Detail,
2020-09-30 08:49:41.421 Info, Node012, Received SwitchMultiLevel report: level=99
2020-09-30 08:49:41.421 Detail, Node012, Refreshed Value: old value=0, new value=99, type=byte
2020-09-30 08:49:41.421 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:41.421 Detail, Node012, Notification: ValueChanged

== Switch turned OFF by itself
2020-09-30 08:49:43.479 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x00, 0xb3, 0x00, 0x69
2020-09-30 08:49:43.479 Detail,
2020-09-30 08:49:43.479 Info, Node012, Received SwitchMultiLevel report: level=0
2020-09-30 08:49:43.480 Detail, Node012, Refreshed Value: old value=99, new value=0, type=byte
2020-09-30 08:49:43.480 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:43.480 Detail, Node012, Notification: ValueChanged

== Turned circuit ON via dumb paddle (stays on)
2020-09-30 08:49:50.398 Detail, Node012,   Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x63, 0xb5, 0x00, 0x0c
2020-09-30 08:49:50.398 Detail,
2020-09-30 08:49:50.398 Info, Node012, Received SwitchMultiLevel report: level=99
2020-09-30 08:49:50.398 Detail, Node012, Refreshed Value: old value=0, new value=99, type=byte
2020-09-30 08:49:50.398 Detail, Node012, Changes to this value are not verified
2020-09-30 08:49:50.398 Detail, Node012, Notification: ValueChanged

@Eric_Inovelli, I’m still seeing random light switch off issues, here is where I am

  1. Tried a brand new LZW31. Flashed to 1.48 (same as the other)
  2. Dimming does NOT work 1.48. Cannot do a MAX level (param sets but lights always come on at 100)
  3. Confirmed wiring tightness on both switches.
  4. Confirmed switch is set to 3 Way Toggle.

See the logs, wattage and answers to your questions in my comment above.

There are lots of oddities here, meaning I’ll watch the light status in HA toggle back and forth, even though the light itself isn’t changing. Dimming in HA or on switch has no impact. I’ll turn the switch off via HA, watch the LED bar dim down, and then go back to ON brightness right away (and no difference to lights)

I’m at my end here, with 15+ hours of troubleshooting. I could review the wiring, but it doesn’t seem like a cable wiring issue, but more the switch itself. The LZW31 Line is wired to the circuit breaker Line. The light Load is in the box w/the dumb switch, and it’s wired to the black nut opposite the brass nut (where the Traveller is connected).

EDIT: I luckily had an SN lying around, which fully works on v1.48! While I’m happy to have a normal house again, it makes me sad about my base LZW31s, especially as one has lost its dimming capabilities after the v1.48 capabilities.

@EricM_Inovelli any thoughts here?

I don’t see this on other hubs, so I’m not sure if this is a HA issue or what?

Happy to ship it in, I just can’t twiddle anymore in the house. This behavior occurred on both ST and HA.

Sorry for the dumb question, but did you flash Target 1 with the bin file? If yes, I would try to flash both Targets again and perform a factory reset afterwards.

@EricM_Inovelli if I get the opportunity to put this back into the wall, I’ll try that. I remember asking checking myself, but actually don’t think I flashed Target 1, because I didn’t see it called out in the instructions but I may have confused those instructions from a different FW version.

1 Like

@EricM_Inovelli and @nic, I just updated my firmware to 1.48 in an effort to get rid of the random shut off issue and have run into the exact issues described by @nic now that it’s running 1.48.
I’m going to re flash it sometime today and will let you know how it goes.

…and it appears I have a new brick in the wall.
I first attempted to downgrade to 1.47 with the hubitat firmware updater. This caused the switch to intermittently turn off with the LED bar turning to red intermittently.
I then excluded the device in Hubitat and used my SiLabs Z wave controller with Z-wave PC Controller 5 to try to update to 1.48. Target 0 seemed to work successfully but at the completion of Target 1, Add Node Security Failed showed up at the bottom of Z wave PC Controller 5. I’ve been unable to control the lights and the LED is black. No noise from the relay inside with button clicks.
When I try to exclude and then include through Hubitat it gets stuck “Initializing” the device. And in PC Controller 5, it includes successfully but when I ‘Get Version’, all it reports is Command Class Version where it reports “4”, but nothing else.
Anyone with any thoughts or assistance on this Thanksgiving Day/Weekend?

Ooof, thanks for the validation, but sorry to hear you’re in the same boat as myself. My switch ended up in a similar zombie state. I don’t remember the exact symptoms, but it wasn’t completely black like you described. However, I had written DEAD across it with sharpie!

If you’re getting near the end of your rope, and in case you didn’t see above, my eventual was to give up and use the Red series I had on hand. Living room lights were a little too publicly visible for another round of testing, and (offtopic) I’m battling another front where some of my switches you really have to press to make them respond.

I am having the random turn off issue with my red series dimmer. I too am using a 3way setup with a dumb switch and neutral. I am on Smartthings. I am driving three 9 watt LED bulbs in a ceiling fan. The issue only happens when the dumb switch is in the down position. I haven’t looked to see if lights are being powered from the Load or Travler wire when in that position. Reducing the max dim level helps with the issue, as did replacing the the LED bulbs with a different brand. When it happens, the LED bar on the inovelli switch turns to it’s normal off state, as if I had flipped the dumb switch off. My theory is that the inovelli switch is monitoring current to see when the dumb switch if flipped off. With the LED bulbs having basically a power supply in them, they don’t need to continuously draw power from every pulse of AC (like incandescent does). If the bulb skips a few pulses before drawing power again, the inovelli switch would see that as the dumb switch being flipped off and shut itself off. Reducing the max dim level likely helps, because the power supply in the bulbs now needs current more often, this reducing the no current period. I would think this could be fixed in firmware via a longer no current period before triggering the shut-off or via adding a bypass. Adding more LED bulbs would likely help too. I’m not really interested in the bypass, as it somewhat defeats the purpose of LED bulbs. The problem exhibits itself in two different light sets I have. I have tried 4 different inovelli switches. I also replaced the dumb switch twice. Has this problem been fixed?

@daveslifeontheedge what version of firmware are you running? v1.48 addresses just the symptom/fix you are discussing about the issue going away when the power is reduced (below 80%).

If you are at 1.48 or above, please let us know so we can keep tracking.

1 Like