Verify JK-B2A8S20P support #105
Labels
Comments
|
I've extracted the frames from the Cell info ( Settings ( Device info ( |
|
|
@EasilyBoredEngineer FYI. Most values are fine. Some are a bit off. Invalid or strange values:
|
|
Hi @syssi, as requested please find attached app screenshots and updated log file. Note:-
|
|
|
If you increase the log level by: you can see the raw traffic if you like. |
|
Thanks @syssi - (deleted hex log above as it was mangled - wasn't obtained properly so was showing the hex in the ascii section) Very verbose logs attached, captured right now after the 4s change. |
syssi
added a commit
that referenced
this issue
May 14, 2022
|
reupload of hex dump response, fyi. This is how the BMS responds to the following command:-
|
|
I've fixed the decoding of the "under temperature protection" and "under temperature protection recovery". It's a signed integer now. |
Thankyou! Now I just have to remember how to synchronise git..... :) |
|
Let's remove this parts from your config yaml: and |
|
@EasilyBoredEngineer If you use this project as clones the main branch of this repository every time ( |
|
great! Thankyou. Yaml changes made. Attached is new log and a screenshot. Note the circled items do nothing when clicked. |
|
additional screenshots attached - looking good! |
|
Are you sure the buttons doesn't work? Because it's not documented it try to explain what's happing: If you press the "retrieve settings button" the If you pree the "retrieve device info" button the BMS should beep and the component should write this to the log: Could you test again? Does it work or doesn't beep the BMS? |
|
@syssi - question for you - if I wished to adapt your code to send the bms data out as a struct via serial IO, as well as sending to home assistant, which part of the code would be best to work on / fork? |
No such section noted in log at very verbose level when / after pressing button. Accessing remotely at the moment so cannot confirm whether the BMS beeps or not, apologies, will revisit when next to the ESP. |
|
This is the most important file about the JK02 (and JK04 protocol): https://github.com/syssi/esphome-jk-bms/blob/main/components/jk_bms_ble/jk_bms_ble.cpp Take a look at the If the Android app connects to the BMS the device info frame is requested first: https://github.com/syssi/esphome-jk-bms/blob/main/components/jk_bms_ble/jk_bms_ble.cpp#L159 The second step is to request the settings + enable the periodic sendout of the cell info frame: https://github.com/syssi/esphome-jk-bms/blob/main/components/jk_bms_ble/jk_bms_ble.cpp#L275 Please make sure between the first request (0x97) and the second (0x96) must be a gap of ~0.5sec. |
|
Thankyou very much @syssi - really appreciated :) FYI - just changed log level to INFO and CONFIRM - both buttons work. attached log shows the relevant responses. |
|
Lucky guy! I was already afraid that you might have damaged the RX pin of your BMS or Jikong has introduced some encryption / verification bits at the random part of the request to avoid unsecured/unauthorized traffic. |
Ha! So was I!! Nope, looks like all is working well now with your package.. Now that's not to say that I haven't possibly killed something else on the bms during my 'investigations' (I did make a big spark with my multimeter probe down towards the mosfet end at one stage - lol) - but I'll know that soon enough :) Thanks again for all your help - close as successful if you wish? |
|
One piece of the puzzle is missing unfortunately: The meaning of each bit of the error bitmask of your device is unknown. Your android app doesn't show a red error message anymore but some bits of the error bitmask are still |
|
Ok - understood. I've just gone down to the office and tried to find a way to factory reset the bms, but alas on my app there is no such option.. another potential cause may be that I have strange private user data (0123456789012), and can't change it using the android app - wondering if that might be bleeding over into the next byte... I'll try to do some investigations over the coming days:- EBE 'to do':-
|
|
I've seen another user/log yesterday with the same strange error bitmask. I assume everything is fine with your BMS and something has changed here. How to identify/reverse engineer the new error bitmask:
|
|
New finding: The error bitmask of the JK02 protocol has always been broken / isn't implemented properly. I can reproduce the issue with my BMS model too. |
|
I've fixed the error bitmask. Just some error bits are well known but may be we can find more in future: |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
jblance/mpp-solar#224
The text was updated successfully, but these errors were encountered: