Aqara T1M (CL-L02D): no effect on endpoints except on/off #2321
Comments
|
Created a log when trying to change RGB color (using color_rgb, which is type level.color.rgb -> writing #ff00ff into the datapoint). This is not creating an error code, however the lamp still does not change its color: when writing e.g. 255,0,255 instead of #ff00ff hex-code style: |
|
When using brightness_white datapoint (0...100), this is the log I get (incl. the error code): |
|
Datapoint colortemp_move (entered as -50...50) is effective for the primary light, creating this log: |
|
Important: all logs have been created while device was excluded. Let me know if I should re-do the test without the exclusion as well, and if other datapoints are of interest. Thanks for your support! |
|
Please set the adapter back to info instead of debug, and instead run the same test with the IEEE of the device(s) entered into the As far as I see it currently, the issue is with the device, as the command is sent to the device but it does not respond. A. |
|
@asgothian: I changed the logs as requested in the comments above, let me know if I can support debugging further. |
|
Thanks for the change. Please make the following by entering the given strings into the "send_payload" state for the device and check for the lamps reaction (and the logs) Please also post the logs for each of the tests separately.
This will indicate if it is an issue with the converters or the zigbee Adapter. A. |
|
Just copying screenshots from the device in the adapter view to have all information on one place: |
|
Logs... Device consists of a inner lamp that should have dimmable brightness and changeable white-tone (CCT) and an RGB outer ring {"state_white":"ON"}
{"brightness_white":128}
{"state":"ON"}
Decided to also try payload {"state_rgb":"ON"}, nothing happens: {"brightness":128}
{"brightness_move":25}
{"brightness_move":-25}
Maybe there is a mapping issue between datapoints shown in iobroker and the 2 sections of the lamp? I couldn't find any action that works for its RGB part, while the white section works on some datapoints. |
|
Possibly. I’ll need to take a close look at the implementation in the code. Are you comfortable using the developer tab in the adapter to see if the lamp reacts properly ? |
|
Hi @asgothian!
Since I'm handing over all data to a Loxone Miniserver, I only need datapoints to set acutal values, means only "moveTo..." type of messages. My results in detail: Switching on/off: Works for both endpoints where 1…CCT white part and 2…RGB part of the lamp Logs for on command (independent from endpoint): Dimming: Works for both endpoints where 1…CCT white part and 2…RGB part of the lamp Log (independent from endpoint): CCT White Tuning: Expects values between 0…65535, where 65535 is the warmest tone Logs (independent from endpoint): Color Tuning on RGB part Works with endpoint 2 (RGB part of the lamp) Log: Final remark: |
|
what happens if you do not use the expert mode to include the trans time, just add the level / color / colortemp ? Does it refuse or will it at least set the correct value ? I will need to build additional logging to see if the iobroker states send the data to the correct EP. That could explain the issue. I can't promise if I manage that before Christmas, though. A. |
|
I didn't add the reason why I used expert mode in my previous posting: Maybe I am not using the dev-mode in the right way, please correct me if I'm wrong:
The problem I'm facing is in step 6, since I can't find the right named attributes from the dropdown. Even if I use the closest possible thing I find in the dropdown of attribute ID, it doesn't work. Example: 
There is no "level" attribute available in the dropdown, choosing "onLevel" instead. Log: Please let me know which attribute IDs I should use / where I could find the right ones. Thanks! fyi: I am allowed to send the lamp back to the vendor until Jan 31st, 2025. I cannot and do not want to put any pressure on you. If you either think that it's not even feasible to make the lamp compatible or that you don't have the time to support it, I'm 100% fine if you tell me any time before end of January. Then I'd return the lamp and find something compatible (found a nice looking Philips Hue Lamp that is only slightly more expensive and seems supported an all websites from blackadder to zigbee2mqtt). Thank you! |
|
I do hope that I will be able to get to the bottom of this before Jan 31st. I already have a suspicion on what is going wrong. As far as the expert mode with the moveToLevel is concerned - this can be a remnant of the old system. There was a refactor in the underlying library at some point where some keywords were changed, which explains why the attribute-ID is not mapped out correctly. So you are using the expert mode correctly. I will try to fix a log output first to confirm this, and then dig through where the endpoint might have gotten lost. Unfortunately, I cannot run the tests myself as I do not have any light with 2 endpoints. A. |
|
Thanks a lot, all clear. If I can assist the debugging, please let me know, we could sync over discord or whatever is convenient for you! Merry Christmas! |
|
Please install the 1.10.14 from my GitHub repository (https://github.com/asgothian/ioBroker.zigbee). It includes 2 important changes:
If my suspicion is right, it should either work because there was a bug in the converters or it will not address the correct endpoint - the log should show it. A. p.s. Please note - due to massive changes in the zigbee-herdsman-converters, there will be certain devices which will stop working with the adapter. I am checking if there is a simple way to fix this, else this will be a side effect of implementing the new code. For this test, I would ask that you ignore the issues and check the lamp only, if needed installing the current main GitHub version again after the test. A. |
|
Hi @asgothian, you're mentioning 1.10.14, if I use your link / the fork above, it still shows me 1.10.13 in my adapter overview. Is there a way to be certain that I didn't mess up the change to your adapter version? Let's clarify this before I redo any tests. Thanks! |
|
If you check the log from the adapter startup, you should find this line:
give me 10 minutes to remove some code again, else it will flood your log. A |
|
Looks good... |
|
Any improvement with controlling the lamp / log lines from the lamp ? Any other odd things occurring with this version ? A. |
|
Hi, To summarize: only state_rgb and state_white true/false works to switch the two endpoints on and off, everything else remains ineffective. Details with the logs below: State_white true (effective): State_white false (effective): State_rgb true (effective): State_rgb false (effective): Brightness_white -> (ineffective) Brightness_rgb (ineffective): Colortemp_white (ineffective): Colortemp_rgb (ineffective): Color_rgb (ineffective): using command 255,0,0: Color_rgb (ineffective): using command #ff0000: |
|
Thank you for the additional data. The cause for the entire problem is clear in a way, but lies outside of the zigbee Adapter. I have opened an Issue on the zigbee-herdsman-converters project for this problem. There are a number of items to test, so if you are up for that, I would love to arrange for a debug session using discord / anydesk so we can get a faster test cycle for various code changes to make more headway. I can be available pretty much at any time. Just let me know if you are up for it, either here or on the ioBroker Discord (nick: Asgothian) A. |
|
@grandslam18 Can you test again with the current 1.10.13 adapter from the official GitHub ? My test were successful, now we need to verify that the lamp still works, as I don't have that. A. |
|
@asgothian: verified, all ok! |
|
Perfect.. closing this. Again thanks for the help. |
|
Thank you so much @asgothian! Great support! |
describe the problem in Detail. What do you expect to happen ? What does happen ?
Installed an aqara T1M ceiling light today and added it to my zigbee adapter instance. Lamp marked as supported in zigbee2mqtt: https://www.zigbee2mqtt.io/devices/CL-L02D.html
No issues during interview:
Interview successful: {"friendly_name":"0x54ef441000b107d2","model":"CL-L02D","vendor":"Aqara","description":"Ceiling light T1M","supported":true} New device joined '0x54ef441000b107d2' model CL-L02D Device '0x54ef441000b107d2' announced itself Interview started: 0x54ef441000b107d2link_quality, power_outage_count and device_temperature look accurate, msg_from_zigbee updates itself in intervals of few seconds:
{"type":"attributeReport","data":{"65522":{"type":"Buffer","data":[170,117,19,132,44,30,7,65,0,15,12,35,1,1,0,8,10,35,8,0,12,100,11,32,2]}},"linkquality":32,"groupID":0,"cluster":"manuSpecificLumi","meta":{"zclTransactionSequenceNumber":72,"manufacturerCode":4447,"frameControl":{"frameType":0,"manufacturerSpecific":true,"direction":1,"disableDefaultResponse":true,"reservedBits":0}},"endpoint_id":1}state_rgb and state_white are effective / let me switch off both lights independently and as expected.
All other datapoints such as brightness_rgb, brightness_white, color_rgb, colortemp_white just produce error codes in the logs like this and show no effect:
zigbee.0 2024-12-15 21:44:39.033 error Send command to 0x54ef441000b107d2 failed with no error code (undefined)Tried to exclude the device in adapter settings, tried multiple pairings and state cleanups without any effect.
Adapter version
1.10.13
Coordinator firmware version
1.0
Coordinator
cod.m ZigBee Coordinator CC2652P7 (CZC-1.0)
Setup
Raspberry Pi 4 (no docker)
Protocol of the error (as text, encapsulated in multiline code tags <>). Separate different occurrances into clearly split log sections
No response
Mod-Edit: Fixed Link
The text was updated successfully, but these errors were encountered: