Support for newer type #11
Comments
|
thanks, i try to develop integration for your strip, i need time, and i sent you some changes to try it , because i can't test it |
|
Please, try to test the code in main repository in: I think you can integrate in HA your strip and do operations "turn on" and "turn off" If you have any problem, please, report here thanks |
|
I create new release 1.0.0 , you can try is and test your led strip. please report results here Thanks |
|
It's detecting them, but when adding: They show up as |
|
Update version 1.0.1 and try in HomeAssistant use debug log to see more information of posible errors. i need device discover information that show in debug logs In configuration.yaml: |
|
I need know too if you use "manual mac" or HA discover the strip Mac and show its name. Thanks |
|
I used discovery. There's not much more logged with that enabled: Then when trying to add, same backtrace as before. |
|
You have old version of elkbledom integration, please update it to last version 1.0.1 and remember restart Home Assistant first to test it again. Thanks |
|
So I updated and the error goes away, but it doesn't turn on and off. I'll try to figure out the correct startup sequence once i have some more time, I must've missed something. |
|
Please, can you send me in separate , init command sequence, turn on command and turn off command, it is posible i confuce it. |
Closed
|
Hi I have the almost same model and behave same as it show up as MELK-OA10 F0 and successfully add in HA. But cannot turn on and off. |
|
I cannot try or discover strip information i have not a strip to try. I only need you send me in separate, init command (if is needed) , turn on command and turn off command. You can google search this information and try it with your led strip, if works for you post it here and i change integration to work |
dave-code-ruiz
added a commit
that referenced
this issue
Apr 26, 2023
Please, test with new turn on and turn off commands in version 1.0.3 and post results, thnks |
dave-code-ruiz
added a commit
that referenced
this issue
Apr 27, 2023
|
Please, test too with new init command in version 1.0.5 and post results, thnks |
|
After update to 1.0.5, It still can auto discover the light, but unable to add it with unknown error. |
Issue solved in version 1.0.6, Sorry, my mistake |
|
No issue to add now. But still unable to control. |
|
I have the same lights as @sasukebinbin, it runs with the same app and control. |
|
You can use gatttool to try discover your init, turn on/off command with: @infernix detect you need use init commands -n 7e0783 and then -n 7e0404 and before you can turn on off or |
|
I tested all the commands using gatttools and it works locally, but when I tried to add it to HA I got the following error: |
|
Ok, first, if you tested all the commands using gatttools and it works locally, please, send me how to with all commands you set in gatttool, it is very helpful for me. It is very difficult to develop code without having the led strip to test. The error you post is in init command, it is the first led strip that needs a init command. I created an error handler for the init command and change turn on command in elkbledom.py file, you need to do this to try it: When homeassistant restart you can test it, send me DEBUG log |
|
Running this sequence, I was able to turn off and turn on my light strip: Now, I'll update the file that you sent me and check the logs in HA to see what is failing. Thanks for your help and support! |
|
I have two devices and I tried it on both, this is the log: |
|
Ok, I see you have a problem with start_notify i now not call it, not is needed, you need to do this to try it: When homeassistant restart you can test it, send me DEBUG log |
|
This is the new log: |
|
Ok, i have new change, you need to do this to try it again:
When homeassistant restart you can test it, send me DEBUG log |
|
Still having errors 😕 |
dave-code-ruiz
added a commit
that referenced
this issue
May 5, 2023
|
we are in the right track, i publish new change, please, to try it again:
When homeassistant restart you can test it, send me DEBUG log |
dave-code-ruiz
added a commit
that referenced
this issue
May 5, 2023
dave-code-ruiz
added a commit
that referenced
this issue
May 5, 2023
|
just an observation as i havent had the time to debug further myself. i bought several of these, each came with a remote. I can turn them on with any remote, but they only respond to color commands from a single remote each. So it looks to me as if there is some init and/or login/binding set of commands which then limits from which controller it accepts commands. @dave-code-ruiz if you drop me a mail with your home address (can find it on my github profile page) i'll buy you one and have it shipped to you, if you're interested. |
That's weird. I have two and I can change the color of both with the same control. But I was also considering the option of sending one of the lights 😅 |
|
Yes you can control multiple from one controller at the same time, but not one strip from multiple controllers. Or that's what it looks like to me |
This is the new log information: |
dave-code-ruiz
added a commit
that referenced
this issue
May 8, 2023
|
ok, try again, i think i have problems with login/init commands, when code wants to read information to led strip show an error, i think it is for login/init commands don't work. Solution: Do not read state information, my integration already had a bug about reading the state of the led strip When i receive the led strip courtesy of @infernix i will continue with this |
Thank you! It's working now. If you could improve that in the future that would be awesome, but now it's really useful to be able to use the lights from home assistant. |
|
Great!!, can you tell me in addition to turn on and turn off commands, the color and brightness also works? |
|
Thnks for your coffee, I have refunded it by mistake, hehehe |
Yes, everything is working. It could be more precise when changing color, but I think it is also a problem with the device because with the native app it does not achieve the desired color either |
dave-code-ruiz
added a commit
that referenced
this issue
May 10, 2023
- Finish changes for #11 - readme description - Ajust config params
|
I create new version 1.0.7 with changes creates in this issue. |
|
It works great now. All function can work inside the compontent. But whatever it's already awsome. |
Thanks @infernix for buy me your led strip, please post amazon link here, thnks Initialy i have same issue. in logs show cannot connect to Led strip. Before send inits commands i connect it and works without problem. Where discover you init commands ?? |
|
Install the DuoCo StripX android app and then turn on bluetooth packet logging. it generates a logfile which you can then feed into wireshark. You can find the exact strip if you search amazon for |
|
Did some digging with packet logging, i think this integration can handle brightness better: #41 |
|
Did some changes to better support
Not sure how to combine it with the |
Closed
|
@ynsgnr since the protocol is backwards compatible, wouldn't it be possible to create an inheritance-based approach where there's:
|
Probably, but I do not own any ELK device so I can not implement or test it. Could also done with interfaces or abstract classes but all of those implementations and testing takes time which I do not have |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi,
I've got some strips that report as MELK (
found device: BE:16:71:00:18:91 0 MELK-OA10) instead of ELK and they use the DuoCo StripX app from the play store. By default they don't seem to work. For what I can tell, it looks as if these are looking for some initial (auth?) packet before they accept commands.After I send
char-write-cmd 0x0009 7e0783and then7e0404I got them to work. From then on they seem happy to accept all existing commands.gattool reports:
Here's the base64-encoded btsnoop log:
(save as
melk.txt, thencat melk.txt | base64 -d > melk.logand open that in wireshark)Any ideas what more you would need to support these?
The text was updated successfully, but these errors were encountered: