LoRaWAN Node for Growatt Photovoltaic Inverter Modbus Data Interface

Use this project to monitor your PV Inverter in case it is out of reach of your WiFi access point (and a WiFi repeater would not help/is not wanted) and you are not able to cover the distance with an RS485 cable. Otherwise, the original Growatt ShineWiFi-S (serial) / ShineWiFi-X (USB), the customized otti/Growatt_ShineWiFi-S or nygma2004/growatt2mqtt can be used.

This is a reimplementation of growatt2lorawan using RadioLib.

• Scheduled uplink (2 message types)
• Network time sync
• Downlink decoding
• Commanded uplink (configuration)
• Battery voltage reading

• Hardware Requirements
• Inverter Modbus Interface Options
  • Modbus Interface Select Input
• Power Supply
• Pinning Configuration
  • Modbus Interface Selection
  • Modbus via RS485 Interface
  • Debug Interface in case of using Modbus via USB Interface (optional)
• LoRaWAN Network Service Configuration
• Library Dependencies
• Software Build Configuration
• LoRaWAN Payload Formatters
  • The Things Network Payload Formatters Setup
• MQTT Integration and IoT MQTT Panel Example
  • Set up IoT MQTT Panel from configuration file
• Remote Configuration Commands / Status Requests via LoRaWAN
  • Parameters
  • Using Raw Data
  • Using the Javascript Uplink/Downlink Formatters
• Loading LoRaWAN Network Service Credentials from File
• Datacake Integration

• ESP32 (optionally with LiPo battery charger and battery)
• SX1276 or SX1262 (or compatible) LoRaWAN Radio Transceiver
• LoRaWAN Antenna
• optional: RS485 Transceiver - 3.3V compatible, half-duplex capable (e.g Waveshare 4777 module)
• optional: USB-to-TTL converter for Debugging (e.g. AZ Delivery HW-598)

1. USB Interface

   The inverter's USB port operates like a USB serial port (UART) interface at 115200 bits/s. If the length of a standard USB cable is sufficient to connect the ESP32 to the inverter (and there are no compatibility issues with the ESP32 board's USB serial interface), this is the easiest variant, because no extra hardware is needed.

   As pointed out in otti/Growatt_ShineWiFi-S, only CH340-based USB-Serial converters are compatible - converters with CP21XX and FTDI chips do not work!

2. COM Interface

   The inverter's COM port provides an RS485 interface at 9600 bits/s. An RS485 tranceiver is required to connect it to the ESP32.

The desired interface is selected by pulling the GPIO pin INTERFACE_SEL (defined in settings.h) to 3.3V or GND, respectively:

The ESP32 development board can be powered from the inverter's USB port which only provides power if the inverter is active.

No sun - no power - no transmission! 😎

But: Some ESP32 boards have an integrated LiPo battery charger. You could power the board from a battery while there is no PV power (at least for a few hours).

See src/growatt_cfg.h

USB-to-TTL converter, e.g. AZ Delivery HW-598

Create an account and set up a device configuration in your LoRaWAN network provider's web console, e.g. The Things Network.

• LoRaWAN v1.1
• Regional Parameters 1.1 Revision A
• Device class A
• Over the air activation (OTAA)

• RadioLib by Jan Gromeš
• Lora-Serialization by Joscha Feth
• ESP32Time by Felix Biego
• ModbusMaster by Doc Walker

• Install the Arduino ESP32 board package in the Arduino IDE
• Select your ESP32 board
• Install all libraries as listed in package.json — section 'dependencies' — via the Arduino IDE Library Manager
• Clone (or download and unpack) the latest (growatt2lorawan-v2 release)
• Set your LoRaWAN Network Service credentials — RADIOLIB_LORAWAN_DEV_EUI, RADIOLIB_LORAWAN_NWK_KEY and RADIOLIB_LORAWAN_APP_KEY — in secrets.h:

// The Device EUI & two keys can be generated on the TTN console

// Replace with your Device EUI
#define RADIOLIB_LORAWAN_DEV_EUI   0x---------------

// Replace with your App Key
#define RADIOLIB_LORAWAN_APP_KEY   0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--

// Put your Nwk Key here
#define RADIOLIB_LORAWAN_NWK_KEY   0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--, 0x--

• Load the sketch growatt2lorawan-v2.ino
• Compile and Upload

Upload Uplink Formatter and Downlink Formatter scripts in your LoRaWAN network service provider's web console to allow decoding / encoding of raw data to / from JSON format.

See The Things Network MQTT Integration and Payload Formatters and TS013-1.0.0 Payload Codec API for more details.

Decode uplink payload (a sequence of bytes) into JSON format, i.e. data structures which are readable/suitable for further processing.

In The Things Network Console:

1. Go to "Payload formatters" -> "Uplink"
2. Select "Formatter type": "Custom Javascript formatter"
3. "Formatter code": Paste scripts/uplink_formatter.js
4. Apply "Save changes"

Encode downlink payload from JSON to a sequence of bytes.

In The Things Network Console:

1. Go to "Payload formatters" -> "Downlink"
2. Select "Formatter type": "Custom Javascript formatter"
3. "Formatter code": Paste scripts/downlink_formatter.js
4. Apply "Save changes"

Arduino App: IoT MQTT Panel

You can either edit the provided JSON configuration file before importing it or import it as-is and make the required changes in IoT MQTT Panel. Don't forget to add the broker's certificate if using Secure MQTT! (in the App: Connections -> Edit Connections: Certificate path.)

Editing IoT_MQTT_Panel_Growatt2LoRaWAN.json

Change USERNAME and PASSWORD as needed:

[...]
"username":"USERNAME","password":"PASSWORD"
[...]

Many software parameters can be defined at compile time, i.e. in BresserWeatherSensorLWCfg.h. A few parameters can also be changed and queried at run time via LoRaWAN, either using raw data or using Javascript Uplink/Downlink Formatters.

• Sleep interval (long): SLEEP_INTERVAL, SLEEP_INTERVAL_LONG; see growatt2lorawan_cfg.h
• LW_STATUS_INTERVAL: see growatt2lorawan_cfg.h

If a valid file secrets.json exists on LittleFS, the settings defined at compile time (in secrets.h) are overridden.

Modify the example data/secrets.json as required and install it to the board's Flash memory using earlephilhower/arduino-littlefs-upload.

For integration with Datacake, there is the script datacake_decoder.js. With Datacake, you can get data reports as CSV files at regular intervals. The Python script datacake_report_pv.py allows to concatenate, sort and filter those files and to create a report with data plots as PDF file (example).