LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Discussions about Z-Way software and Z-Wave technology in general
400HPMustang
Posts: 14
Joined: 09 Jun 2022 22:54

LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by 400HPMustang »

I'm running the Windows version of Z-Wave.Me, version 3.2.1 which according to the download page is the latest. If not, I'd like a link to the latest version.

Anway, I have my Inovelli LZW36 installed/configured and everything is working. I have noticed that if I turn on the light or fan via Zwave that after 2 hours and 10 minutes the light or the fan will turn itself off.

After doing some google searching it appears that this was happening for other Zwave platforms as well.

https://github.com/zwave-js/node-zwave-js/issues/3594
https://github.com/zwave-js/node-zwave-js/issues/2159

While I realize that this is said to be a firmware bug with the device, it does not appear the manufacturer has fixed the issue as I'm running the latest firmware on the device and currently experiencing the problem.

With that said, is it possible to fix or mitigate the issue on the Zwave side in Zwave.Me?

Thank you,
lanbrown
Posts: 279
Joined: 01 Jun 2021 08:06

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by lanbrown »

Have you tried the beta firmware for it?
https://support.inovelli.com/portal/en/ ... red-series

Just my 2 cents, I cannot see the Z-Way developers developing a "fix" for one specific Z-Wave device. A change like that would create complexity and technological baggage that they would need to deal with for years to come. From a technical standpoint I could not see how they could "fix" it if the either the switch or the RF component is the one turning the fan off after 2 hours and 10 minutes. Why would they want to add to the code that if that device is used and the fan turns off at that timeframe to then turn it back on. There could be undesired consequences like not being able to turn the fan off when you actually want it off. There is a reason why KISS (Keep It Simple Stupid) is usually the best approach. The more complex something is, the harder it is to get it working right and keep it working.

You could use a virtual device in Z-Way and that is trigger to be on when the fan is on; if the fan goes off and that virtual is still on you can create a rule to turn it back on. So from Z-Way you would actually be controlling the virtual device to control the fan for both ON and OFF. Then you just have rules in Z-Way for the LZW36 to match the virtual.

You might want to check that parameter 11 is set to 0
https://support.inovelli.com/portal/api ... nline=true

If you want this fixed, then you should probably contact Inovelli. They are quite responsive and see what they have to say about the bug and when they expect to fix it. With this bug impacting multiple platforms, it is not something that they can ignore.
400HPMustang
Posts: 14
Joined: 09 Jun 2022 22:54

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by 400HPMustang »

Hey thanks for your incredibly detailed response.

First, yes I am running that beta firmware and it has not helped.

Second I did reach out to Inovelli support, unfortunately they are out of office or something until the 27th per the auto reply email I got so I'll be waiting a while for them to help.

Third, yes I did check parameter 11 and it is set to of which is how I ended up finding that this is some kind of bug with Zwave. What's happening, according to the github posts I've linked is that when something called "supervision" is run the switch reboots itself which results in turning the accessories off. It was explained better in that second github post I linked. Since the Zwave js folks seemed to be able to correct for it fairly quickly once it was identified I figured it was worth a shot to ask the Z-Way folks too. This does however beg the question of why the Inovelli devs didn't just fix the firmware.
lanbrown
Posts: 279
Joined: 01 Jun 2021 08:06

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by lanbrown »

This might not do anything, but worth a shot. You can always put it back to true just as easily as false.

You will need the IP address of the your z-way system and know what the Z-Wave node ID is for the LZW36.

Try inputing this command into a browser:
http://ZWAY-IP-ADDRESS:8083/JS/Run/zway.devices[LZW36 Node ID].Supervision.data.interviewDone=false

After inputting the command, it will show that the interview is not complete for the LZW36 and that Supervision has not been completed. The hope is that Supervision won't be used since it is viewed as incomplete. If it makes no difference, then just change the =false to =true

So if you Z-Way IP is 192.168.1.100 and the Z-Wave Node ID is 200 the command would be:
http://192.168.1.100:8083/JS/Run/zway.d ... Done=false

It also appears that a number of people have disconnect issues between the switch and the canopy modules; but not everyone. They haven't been able to pinpoint it. Before the product was available and they just announced it, the canopy module put me off. It just seemed to add complexity. They might have had space constraints in the switch itself and had to use a module to control things or maybe the firmware would be too big for 500 series.

I have quite a few of the red switches. Even if I could buy more, I wouldn't and wait for the 700 series. Their supplier for Zigbee is/will be a supplier for the 700 series chipsets as well. So I expect that they will be refreshing their product line after the Zigbee product rollout. I know they were running into issues with the firmware size on the 500 series. The memory available is fixed on the modules. I would expect the 700 series to have a larger memory since more and more features are being added to devices.

You could always just do the virtual device method as well. The rule would compare the actual state (status of LZW36) to what the expected state is (state of the virtual device) and if they are different to act upon it. So if the actual state goes to off and the virtual is on, the rule would turn the LZW36 back on. A second rule would to turn the LZW36 off if the virtual is off. This way you would control the virtual. You might need a third if you manually turn it on.
400HPMustang
Posts: 14
Joined: 09 Jun 2022 22:54

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by 400HPMustang »

I ran the command you gave. I'll see what happens with it tomorrow. I don't want to start running what amounts to a 2+ hour "test" this late in the evening since my house is settling down for the evening but I'll see what the result is tomorrow.

I did also make a comment in a post on the Inovelli forum too, maybe there's some hidden gem of information someone will drop on me or maybe I'll just be another datapoint in this ongoing saga.

Either way, thank you for your help.
seattleneil
Posts: 172
Joined: 02 Mar 2020 22:41

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by seattleneil »

My understanding is that disabling Supervision is the correct method (instead of marking the interview for Supervision as not done). Here's an example (for Z-Wave device ID 7, fill in the correct password and IP address):

curl -u admin:mysecretpasswd --globoff http: //myip:8083/JS/Run/zway.devices[7].instances[0].Supervision.data.disabled=true
400HPMustang
Posts: 14
Joined: 09 Jun 2022 22:54

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by 400HPMustang »

seattleneil wrote:
20 Jun 2022 07:19
My understanding is that disabling Supervision is the correct method (instead of marking the interview for Supervision as not done). Here's an example (for Z-Wave device ID 7, fill in the correct password and IP address):

curl -u admin:mysecretpasswd --globoff http: //myip:8083/JS/Run/zway.devices[7].instances[0].Supervision.data.disabled=true
Ok. I made that change. I'll report back in about 2 hours and 10 minutes. Besides running that curl command, was there anything else I was supposed to be doing before testing this out again?
400HPMustang
Posts: 14
Joined: 09 Jun 2022 22:54

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by 400HPMustang »

Well either it didn't work or I didn't run the curl correctly the first time. Interview results now say it's disabled so I'm running my fan again to see how it behaves.

Code: Select all

/: None (12:49)
    0: None (15:11)
        status: 255 (15:11)
        duration: 0 (15:11)
        moreStatusUpdates: false (15:11)
        lastEvent: 530310 (15:11)
    1: None (12:55)
        status: 255 (12:55)
        duration: 0 (12:55)
        moreStatusUpdates: false (12:55)
        lastEvent: 9089 (12:55)
    2: None (12:55)
        status: 255 (12:55)
        duration: 0 (12:55)
        moreStatusUpdates: false (12:55)
        lastEvent: 9089 (12:55)
    3: None (12:55)
        status: 255 (12:55)
        duration: 0 (12:55)
        moreStatusUpdates: false (12:55)
        lastEvent: 9207 (12:55)
    4: None (12:55)
        status: 255 (12:55)
        duration: 0 (12:55)
        moreStatusUpdates: false (12:55)
        lastEvent: 9565 (12:55)
    5: None (12:55)
        status: 255 (12:55)
        duration: 0 (12:55)
        moreStatusUpdates: false (12:55)
        lastEvent: 9565 (12:55)
    6: None (15:11)
        status: 255 (15:11)
        duration: 0 (15:11)
        moreStatusUpdates: false (15:11)
        lastEvent: 530304 (15:11)
    7: None (15:11)
        status: 255 (15:11)
        duration: 0 (15:11)
        moreStatusUpdates: false (15:11)
        lastEvent: 530304 (15:11)
    supported: true (12:49)
    version: 1 (12:49)
    security: false (12:49)
    interviewDone: true (12:49)
    interviewCounter: 9 (12:49)
    lastSession: 7 (15:11)
    disabled: true (15:08)
400HPMustang
Posts: 14
Joined: 09 Jun 2022 22:54

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by 400HPMustang »

seattleneil wrote:
20 Jun 2022 07:19
My understanding is that disabling Supervision is the correct method (instead of marking the interview for Supervision as not done). Here's an example (for Z-Wave device ID 7, fill in the correct password and IP address):

curl -u admin:mysecretpasswd --globoff http: //myip:8083/JS/Run/zway.devices[7].instances[0].Supervision.data.disabled=true
ok so I've run another test and like clockwork the fan shut off after 2 hours and 10 minutes. Is there any logging or diagnostic info I can pull from the Z-Way software to give them??
seattleneil
Posts: 172
Joined: 02 Mar 2020 22:41

Re: LZW36 Fan/Light turn themselves off after 2 hours and 10 minutes

Post by seattleneil »

On the linux version of Z-Way (which I use), the default log file is /var/log/z-way-server.log. Unfortunately, I don't know where the Windows version of the Z-Way log file is located. I highly recommend figuring out where the log file is located since it can be very useful for troubleshooting issues.

I clicked on the github links that described the problem you're experiencing and have 2 suggestions. (1) Avoid buying more Innovelli products - IMHO it's ridiculous to take more than 8 months for them to fix a serious bug. (2) On a more constructive note, the work-around described on github involved disabling Supervision (i.e., the curl command which you know how to do), using an insecure inclusion (this is not the same as S2-Unauthenticated) and removing encapsulation. For lack of a better idea, try excluding your nasty LZW36, and then use the Expert UI (Network->Control) to force an unsecure inclusion. Once the interview process completes, repeat the curl command to disable Supervision (be sure to use the new device ID), and then repeat the auto-off test. Although it doesn't solve the problem, it's interesting that the switch is not automatically turning off, rather, the switch is restarting itself which has the side effect of turning off the switch. By disabling Supervision and Security, hopefully the switch will not restart itself.
Post Reply