ZW121 Aeotec LED Strips - No command converter

Hello All,

Running Openhabian on latest 2.5.0.M4 on a Raspberry PI.

I have a lot of devices, about 80, and will have about 100 when done.

I am using 4 ZW121 Aeotec LED Strips. I installed the first three months ago. They were installed using the prior version of the Milestone release. I can’t remember which one, but it was a recent one.

I finally got a forth Aeotec LED strip discovered and have items for it, etc.

When I change the color temperature on ALL of the LED strips, I get the following error:

NODE 85: No command converter set for command zwave:device:d0c1b56b:node85:color_temperature type PercentType.

I’ll get this error changing the color temperature from HABPanel OR through a rule. The funny thing is that all of the other ZW121 Aeotec LED Strips behave just fine. Only this latest one - node85 - gives this error.

Changing the colors or the Brightness Dimmer works just fine on all of the LED Strips.

I get the command converter error only on this latest installed one. Is this somehow related to 2.5.0.M4

Thanks.

All at once or individually? Are the all the same firmware version?

Hi Bruce,

All of the commands are issued by a rule. All of the color commands work. The dimmer (brightness) works fine. Any change to the color temperature works on the 3 existing strips I had. The most recently added item one gives me the command converter error. It seems like an OpenHab error, doesn’t seem like a ZWAVE error.

Manufacturer firmware changes sometimes need database updates. That is why I asked about firmware versions. It is easily visible in HABmin

1 Like

Hello Bruce,

Just checked. All LED Strips have the same firmware version. All are 1.2 I have seen other thread citing command converter errors. But none appear helpful to me. Could the database update from the latest version be a factor?

Thanks.
Wes

I have the same problem with the ZW121 Stripe and OH2.5M4. This is my only Stripe of this type so I have nothing to compare.
Mine is firmware version 1.3

I just has updated to 2.5M5, I get the same error.
Is there any I can looking for?
I has reinitialized the device and captured the log and the resulting XML file.
Is there any useful in it?

openhab.log (335.8 KB)
network_da332033__node_43.xml (11.8 KB)

Many many thanks for all your help :slight_smile:

Daniel

Hi all again,

sorry for my monologue, but I’m helpless. I set up an openHAB 2.4 Instance to test the Stripe there. And… it is possible to adjust white.
Ok, after adjusting white, the dimmer jumps back to 0, but the stripe keeps the color temperature.

I hope, somebody have an idea, what can I do to get the stripe working in openHAB 2.5.

Here is the XML File from 2.4 Binding. I can’t see any (relevant) difference between this one and from previus post except the order of command classes and colors.

network_c4b56210__node_4.xml (12.7 KB)

Thanks all,
Daniel

This log doesn’t seem to contain the error that is reported, so it’s hard to know what is the issue. Please provide a log that shows the problem you’re having.

Hi Chris,
thanks for your help.

There is not a lot in the log, only:

2019-11-17 20:31:53.154 [DEBUG] [ding.zwave.handler.ZWaveThingHandler] - NODE 43: Command received zwave:device:32303136-3131-3033-3030-303032383733:node43:color_temperature --> 46 [PercentType]
2019-11-17 20:31:53.154 [WARN ] [ding.zwave.handler.ZWaveThingHandler] - NODE 43: No command converter set for command zwave:device:32303136-3131-3033-3030-303032383733:node43:color_temperature type PercentType

Thats what I have. But… Only for the recently added LED Strip. The ones I added whan the last Milestone Release was current (About a month ago) work fine.

I say fine reluctantly. I am now in that club of Zwave installs that is have all of the problems ever posted. Devices going off line. Devices that respond after minutes of delays, etc.

Hi,
I have break it down to the Thing Definition in the JsonDb.

Here is the (relevant part) Thing from OH 2.4:

        {
          "acceptedItemType": "Dimmer",
          "kind": "STATE",
          "uid": {
            "segments": [
              "zwave",
              "device",
              "a6ba1f28",
              "node4",
              "color_temperature"
            ]
          },
          "channelTypeUID": {
            "segments": [
              "zwave",
              "color_temperature"
            ]
          },
          "label": "Color Temperature",
          "description": "The color temperature channel allows to set the color temperature of a light from 0 (cold) to 100 (warm).",
          "configuration": {
            "properties": {}
          },
          "properties": {
            "binding:Command:OnOffType": "COMMAND_CLASS_SWITCH_MULTILEVEL;",
            "binding:*:PercentType": "COMMAND_CLASS_SWITCH_COLOR;colorMode\u003dDIFF_WHITE"
          },
          "defaultTags": []
        }

Here the Version vom OH 2.5

{
          "acceptedItemType": "Dimmer",
          "kind": "STATE",
          "uid": {
            "segments": [
              "zwave",
              "device",
              "32303136-3131-3033-3030-303032383733",
              "node43",
              "color_temperature"
            ]
          },
          "channelTypeUID": {
            "segments": [
              "zwave",
              "color_temperature"
            ]
          },
          "label": "Color Temperature",
          "description": "The color temperature channel allows to set the color temperature of a light from 0 (cold) to 100 (warm).",
          "configuration": {
            "properties": {}
          },
          "properties": {
            "binding:Command:OnOffType": "COMMAND_CLASS_SWITCH_MULTILEVEL",
            "binding:*:PercentType": "COMMAND_CLASS_SWITCH_COLORcolorMode\u003dDIFF_WHITE"
          },
          "defaultTags": []
        }

I think the Problem is the missing ; between COMMAND_CLASS_SWITCH_COLOR and colorMode.

Edit: the missing semicolon is already seen in the Z-Wave Device Database:

https://www.cd-jackson.com/index.php/zwave/zwave-device-database/zwave-device-list/devicesummary/639?layout=openhab2new

But I don’t know to do it right…

Is this possible an general problem with the Device Database?
I’ve checked several tunable white bulbs and there is at all the semicolon missing.
One example: https://www.cd-jackson.com/index.php/zwave/zwave-device-database/zwave-device-list/devicesummary/619?layout=openhab2new

Edit:
I also browsed the history of the Zwave Github Repository of all the devices, in the older ones is the semicolon existent.

Well. The XML is valid. I suppose it would need to be to work at all.

I have not reached the level of understanding in HAB to know what the application expects these element text items to be. It may be a syntax issue.

Why would it have changed though. Did somebody “fix” something?

There does appear to be a missing semicolon. I am Openhabian. Where does this file live in my application. Can I put the semicolon in locally and test it?

Hi @wesav, yes I can confirm that the semicolon is the breaking part.
As a workaround you can manually edit the jsondb thing file /var/lib/openhab2/jsondb/org.eclipse.smarthome.core.thing.Thing.json. After that you need to restart openHab I guess (I don’t use openhabian). Do it on your own risk!

@chris can you confirm my assumption is right, that the openhab2 XML export from your Database is broken?

Hi,
So… Why do three of them work and one (the latest install not work)?

Does Openhab use and retain device data that exists at the time the device is discovered. So even though I have 4 of the exact same device, the data that Openhab is using could be different versions?
Wes

Yes, take a Look at the file I mentioned.

@chris, @Bruce_Osborne, @sihui, @5iver you are all firm in the z-wave device database.
I searched the Database for tunable white devices (colorMode=DIFF_WHITE).
I found 20 devices, 19 of them are broken.
I think the sunricher 9003T3 is “hot-fixed”

https://www.cd-jackson.com/index.php/zwave/zwave-device-database/zwave-device-list/devicesummary/1024

Is the semicolon right on this place?
If so, then I will fix the 19 other devices in the same way. But I don’t think so. The devices was well before the complete database update. So I think the wrong part is the openHAB2 Export.
Can anyone confirm that? So that can be fixed befor 2.5 release :slight_smile:

Thanks, Daniel

Yes, looks odd. Chris should take a look at this, thx.
Could you open a github issue, please?