Robot Arm, Beckman Coulter Orca

ORCA: Optimized Robot for Chemical Analysis;

5 (6) Axes; [1] Longditudonal motion along rail, [2] 1st lat. joint, [3] 2. lat. joint, [4] 3. lat. joint, [5] 1. rotational joint, [6] 1. Finger/Gripper.
Almost certainly 24V power supply
<5MBit RS485 multidrop bus to all three control boards.
Axes [1,2], [3,4] and [5,6] have their own control boards with differing layouts but nearly i
dentical components.
There is a separate 2-board (probably) inclinometer board attached to the 2. lateral joint. Every joint consists of a brushed DC motor with axles protruding from both ends, with significant gear reduction on one side and a Honeywell glass disk encoder on the other.

Motor Control Board Guts:

MC68HC711E9CFNE2: MCU, 8BIT 12KB OTP ROM 2MHz 52PLCC
L4960: Switchmode controller, 2.5A 9-46Vin 5V Out
SN75176A: Differential Bus Transceiver
DA05P: ESD Protection Diode Bridge
74HC4538: Dual retriggerable precision monostable multivibrator, probably for RS485/something timing alignment
LMD182003A: 55V H-Bridge
- Inputs: Direction, Brake, PWM
- Outputs: Thermal flag, Analog current sense
82C54: CMOS Programmable Interval Timer
TL7705ACP: Voltage Supervisor
HCTL2000: 12/16-bit Quadrature Decoder/Counter

125000, 62500,

Hints to comms protocol:

Supposed to have Force and Torque feedback for at least fingers, maybe all joints. Video online of the robot painting and moving batteries corroborate this, Each 2-joint motor control board probably runs a PID-ish control loop, possibly with full motion control loops but perhaps just point-to-point with motion parameters.
4 near identical modules communicating via a <2MBit RS485 multidrop bus
MCU has one UART (SCI):
- Datasheet §7.3; NRZ signaling, simple data format: Idle high, start bit 0, 8/9 bit packet LSB first, stop bit 1, break period of 0-bit strings for at least 1 frame.
- Datasheet §7.1 Table 1; Baud rate table: 8MHz Crystal → 2MHz Bus. There are 32 possible rates, but something really standard,
- Probably means multi-packet instructions, 8 bits for both operation bits and data bits seems insufficient.

Inspiration: