openlumi / JN-ZigbeeNodeControlBridge-firmware

15 stars 4 forks source link

Is that a firmware backward compatible with the original fork ? #7

Open pipiche38 opened 1 year ago

pipiche38 commented 1 year ago

Looks like you have done some recent update and it might be interesting to have more fixes than on the orginal https://github.com/fairecasoimeme/ZiGate which seems to be dead.

(1) to run as a coordinator which firmware needs to be used :

(2) I found some strange behaviour when using a Xiaomi Vibration

MsgType: 8102, Data: 1453ef01010100552100020001, LQI: 180

G1K commented 1 year ago

This firmware has the same roots with zigate this is an example of JN-AN-1216 These firmwares have the same message format on the serial port. (Difference from original NXP in LQI at the end) we adapted and fixed the assembly of the nxp firmware from under Linux and made the assembly in github actions, because we needed to build for the JN5169 chip ( Used in xiaomi gateway )

History: This version appeared at the time when I was writing an nxp / zigate adapter for zigbee2mqtt, and I constantly encountered chip overflow and hang errors. Therefore, I tried to transfer the minimum part of the patches from zigate to the NXP JN-AN-1216 example. This brought results, and I stopped encountering these errors (0x8000 with the status A6 are received) that were fixed a little later in zigate https://github.com/fairecasoimeme/ZiGate/issues/348 .

Since I use raw zcl frames E_SL_MSG_DATA_INDICATION=0x8002 in the adapter for zigbee2mqtt, many fixes aimed at processed messages, for example, 8102, are not transferred to this firmware (in zigate it is difficult to deal with the commitjd history. often all the contents of the files changed there due to for indentation and line breaks and the first commits were already with fixes different from the nxp JN-AN-1216 example)

The latest changes in the firmware are implemented and tested by the guys from the chat https://t.me/xiaomi_gw_hack , aimed at using the chip in zigbee2mqtt: 1) An attempt to implement a raw backup proposal and restore PDM (on DGNWG05LM gateways with a JN5169 chip, sometimes after a power failure, the chip starts with the message E_SL_MSG_NODE_FACTORY_NEW_RESTART = 0x8007 , which seems to indicate the absence of part of the data in the PDM, the same message is sent by the chip after Erase PDM )

2) Fixes in stability and support for devices connected to this gateway.

(1) We use the version without GP_Proxy, I was not able to start and test Green Power in conjunction with zigbee2mqtt.

(2) Unfortunately, I do not understand what the question is about

MsgType: 8102, Data: 1453ef01010100552100020001, LQI: 180