Documenting the reverse engineering efforts of my Robomow City 120
J105 pinout: RX1: 4 TX1: 5 RX0: 2 TX0: 3 GND: 8
Terminal sends the following packets periodically (with varying -depending on the traffic from the robot- rate between 10-60ms). I never saw traffic on the TX and RX at the same time (the terminal and the robot seems to be handshaking on the bus).
* Idle: 0x23 0x00 0x18 0xC4
* Forward: 0x23 0x40 0x18 0x84
* Backward: 0x23 0x80 0x18 0x44
* Left: 0x23 0x10 0x18 0xB4
* Right: 0x23 0x20 0x18 0xA4
* Rabbit: 0x23 0x00 0x19 0xC3
* C: 0x23 0x08 0x18 0xBC
* Go: 0x23 0x04 0x18 0x44
* Stop: 0x23 0x00 0x1A 0xC2
* Factory: 0x23 0x00 0x1C 0xC0
* Up: 0x23 0x01 0x18 0xC3
* Down: 0x23 0x02 0x18 0xC2
Strapped to Mode 5
SDA - P53 SCL - P52
64KByte
2K on chip RAM present between 0xFFF400 to 0xFFFC00
The contents of the external RAM is initialized from the flash in multiple portions
Both UART0 (connected to the terminal) and UART1 (debug?) seems to be initialized and used.
UART0 (terminal) has 31 byte RX buffer, UART1 (debug?) has 15 byte buffer
UART1 (debug?) initialized to 4800 baud (F_CPU == 16M, BRR1 == 103), the baud rate is adjustable with some commands.
Both UART reception handled with RX ISR.
The UART1 looks to have simpler protocol, and it might be involved in bootloader/flashing/config upload activites (it ends in one case using a "DownLoad..." string).
UART1 (debug?) frames (only 8 byte is handled).