MD FDCAN communication

MD FDCAN communication#

The easiest way to communicate with MD controllers is to use a CANdle device connected to a PC. Even though we are aware some customers want to integrate the MD controllers in their product with minimal setup to reduce the costs and the system’s complexity. This manual will guide you through the process of communicating with MD actuators from your custom CAN master controller.

Hardware requirements#

From firmware version 2.5.0 upwards, MDs are capable of either FDCAN mode communication or CAN2.0 compatibile mode. FDCAN is a preffered protocol to use, as it allows for far greater flexibility and bandwidth. CAN2.0 is supported but not recommended, as the protocol was not optimized for it, and basic operations and control may require multiple CAN frames to be exchanged, significantly reducing the bandwidth, especially with multiple actuators on a single bus.

FDCAN#

The main requirement for the host system is to be equipped with an FDCAN peripheral (either a built-in one or an external one) and an FDCAN transceiver capable of speeds up to 8Mbps. Lower maximum speed transceivers can be used as well, however for the cost of limited update rates. Depending on your custom setup you should be able to integrate a 120 ohm terminating resistor on both ends of your CAN bus.

CAN2.0#

While MD was designed with FDCAN protocol in mind, CAN2.0 compability was introduced. In CAN2.0 mode, the driver can only operate in 1M baudrate, and some registers (that are more than 4 bytes in size) are not available for modification, i.e motorName register. Additionally only some of the Frame Types are supported in this mode - READ_REGISTER_CAN2.0 and WRITE_REGISTER_CAN2.0. In most cases, access in 2.0 mode has to happen 1 register at a time (one register per can frame), with an exception of accessing two U8 type registers. Apart from maximum frame length, the contents of the frames will be the same as in FDCAN version of protocol.

Note

MD controllers can be upgraded to software controlled termination on demand. Please contact us for more information before placing your order.

Communication Structure#

Comminication with MD, happen is a strict Master-Slave structure. The MD itself will never produce a CAN frame by itself, it will only respond to direct commands from a host. Each MD device has configurable CAN ID, that serves as its unique identifier on the CAN bus. There may never be more than one MD with the same CAN ID, as this will lead to conflicts and errors.

All communication with the particular drive will happen only via messeges with particular ID. For example:

Note

Host sends a request (command) to drive with Id 100, the drive will execute the command and respond with a CAN frame that also has the ID of 100

The communication stack is based on a register access using two frames - register read and register write. The list of available registers can be found at the end of this chapter. All fields are little-endian - least significant byte first, and all float fields are 4 bytes long (32 bit) encoded in IEEE-754 standard.

Warning

Wrong access, incorrect data values or other communication errors are not reported explicitly. Command that failed (regardless of the reason), will result in the drive not producing any response. Generally if the drive does not start producing a CAN frame with 100us of the the last bit of the command, the host may consider the command has failed.

Frame Structure#

All frames (FDCAN and CAN2.0) are composed in the same fasion. The first byte is a Frame Type, then a padding byte, followed by the contents of a message.

Message contents are similar in both read and write operations. They follow a pattern of repeating sequence: register id (2 bytes) and register value (1-24 bytes depending on a register). The number of registers to be accessed is only limited by a size of message - 64 bytes for FDCAN, and 8 bytes for CAN2.0.

BYTE 0	BYTE 1	BYTE 2-3	BYTE 4-X	BYTE X+1-X+2	BYTE X+4-X+Y	...
FRAME ID	PADDING (0x00)	REG ID 1	VALUE 1	REG ID 2	VALUE 2	...

For example, a frame that would read a value of the current position, velocity and torque from a drive, would, have a length of 20 bytes, and look like the following:

BYTE 0	BYTE 1	BYTE 2-3	BYTE 4-7	BYTE 8-9	BYTE 10-13	BYTE 14-15	BYTE 16-19
FRAME ID	PADDING (0x00)	regId - mainEncoderPosition	PADDING	regId - mainEncoderVelocity	PADDING	regId - motorTorque	PADDING
0x41	0x00	0x0062	0x00 00 00 00	0x0063	0x00 00 00 00	0x0064	0x00 00 00 00

Frame Types#

Although there are basically only two possible variants of frames - read and write register - the frames can trigger different behaviours, apart of reading and writing. Here is a brief description of the frame behaviours by their id:

Frame ID	Name	Description
0x40	WRITE_REGISTER_LEGACY	performs write operation, and makes the drive respond with legacy response. This is sometimes usefull in high frequency control loops, to minimize number of frames exchanged. For frame contents, refer to Write Register Frame .
0x41	READ_REGISTER	performs read operation, and responds with state of the registers. Refer to Read Register Frame
0x42	WRITE_REGISTER	performs write operation, and responds with state of the registers AFTER operation. Usefull for verification of write operation. For more info, refer to Write Register Frame
0x43	READ_REGISTER_CAN2.0	performs read operation, and responds with state of the registers. Works same as Read Register Frame , but is limited to 8 bytes and produces CAN2.0 compatibile response.
0x44	WRITE_REGISTER_CAN2.0	performs write operation, and responds with state of the registers AFTER operation. Usefull for verification of write operation. Works same as Write Register Frame , but is limited to 8 bytes and produces CAN2.0 compatibile response.
0x0A	LEGACY_RESPONE	RESPONSE ONLY. A response produced as a result of some frames. Contains data, most commonly required in fast control loops - more info here.

Write register frame#

Write register frame is used to modify values of the user-modifiable registers. Only registers with write access can be modified.

FRAME NAME	DRIVE ID	LENGTH	BYTE 0 [ID]	BYTE 1	BYTE 2-3	BYTE 4-X	BYTE X+1-X+2	BYTE X+4-X+Y
WRITE_REGISTER	10-2000	X (64 max)	0x42	0x00	reg ID	value	reg ID	value

Params:

regID (uint16_t) - first register ID (please see the end of this section)
value (uint8_t/uint16_t/uint32_t/float/char[]) - first register value to be written
regID (uint16_t) - second register ID (please see the end of this section)
value (uint8_t/uint16_t/uint32_t/float/char[]) - second register value to be written
… (up to 64 bytes total)

Read Register Frame#

Read register command is used to retrieve certain register values. The actuator will respond with a frame consisting of the addresses and values of the registers issued in the master request. The master request should have the following form:

FRAME NAME	DRIVE ID	LENGTH	BYTE 0 [ID]	BYTE 1	BYTE 2-3	BYTE 4-X	BYTE X+1-X+2	BYTE X+4-X+Y
READ_REGISTER	10-2000	X (64 max)	0x41	0x00	reg ID	0x00	reg ID	0x00

When all read operations succeed the 0x00 fields will be filled with appropriate register data when transmitted back to master by the MDxx controller.

FRAME NAME	DRIVE ID	LENGTH	BYTE 0 [ID]	BYTE 1	BYTE 2-3	BYTE 4-X	BYTE X+1-X+2	BYTE X+4-X+Y
Response to register read	10-2000	X (64 max)	0x41	0x00	reg ID	reg value	reg ID	reg value

Legacy response#

	BYTE 0	BYTE 1-2	BYTE 3	BYTE 4-7	BYTE 8-11	BYTE 12-15	BYTE 16-19	BYTE 20-23
NAME	FRAME ID	QUICK STATUS	MOTOR TEMPERATURE	MAIN ENCODER POSITION	MAIN ENCODER VELOCITY	MOTOR TORQUE	OUTPUT ENCODER POSITION	OUTPUT ENCODER VELOCITY
TYPE	uint8_t	uint16_t	uint8_t [*C]	float [rad]	float [rad/s]	float [Nm]	float [rad]	float [rad/s]
VALUE	0x0A	0x0000 - 0xFFFF	0 - 255	-	-	-	-	-

Available registers#

reg name	address	read/write	size	limits	description
canId	0x001	RW	uint32_t	[10-2000]	FDCAN bus id number
canBaudrate	0x002	RW	uint32_t	[1e6;2e6;5e6;8e6]	FDCAN bus baudrate
canWatchdog	0x003	RW	uint16_t	[0-2500]	FDCAN bus watchdog period in ms
canTermination	0x004	RW	uint8_t	[0-1]	CAN termination (available upon request)

motorName	0x010	RW	char[24]	-	motor name
motorPolePairs	0x011	RW	uint32_t	[2;225]	motor pole pair count
motorKt	0x012	RW	float	>0	motor torque constant
motorKt_a	0x013	RW	float	>0	optional parameter for phase specific torque constant
motorKt_b	0x014	RW	float	>0	optional parameter for phase specific torque constant
motorKt_c	0x015	RW	float	>0	optional parameter for phase specific torque constant
motorIMax	0x016	RW	float	[1 - peak controller current]	maximum phase current
motorGearRatio	0x017	RW	float	-	actuator gear ratio (ex 2:1 should be 0.5) <1 - reducer >1 - multiplier
motorTorqueBandwidth	0x018	RW	uint16_t	[50-2500]	torque bandwidth in Hz
motorFriction	0x019	RO	float32	-	UNUSED
motorStiction	0x01A	RO	float32	-	UNUSED
motorResistance	0x01B	RO	float	[5mOhm-20Ohm]	motor resistance in d axis
motorInductance	0x01C	RO	float	[5nH-100mH]	motor inductance in d axis
motorKV	0x01D	RW	uint16_t	-	motor KV rating [RPM/V]
motorCalibrationMode	0x01E	RW	uint8_t	[0;1]	FULL = 0, NOPPDET = 1
motorThermistorType	0x01F	RW	uint8_t	-	-

outputEncoder	0x020	RW	uint8_t	[0;1;2;3]	NONE = 0, ME_AS_CENTER = 1, ME_AS_OFFAXIS = 2, MB053SFA17BENT00 = 3, CM_OFFAXIS = 4, M24B_CENTER = 5, M24B_OFFAXIS = 6
outputEncoderDir	0x021	RW	uint8_t	0	RESERVED
outputEncoderDefaultBaud	0x022	RW	uint32_t	115200	optional parameter for setting default output encoder baudrate
outputEncoderVelocity	0x023	RO	float	-	output encoder velocity in rad/s (calculated in a 5kHz loop)
outputEncoderPosition	0x024	RO	float	-	output encoder position in rad (read in 5kHz loop)
outputEncoderMode	0x025	RW	uint8_t	[0;1;2;3;4]	NONE = 0, STARTUP = 1, MOTION = 2, REPORT = 3, MAIN = 4
outputEncoderCalibrationMode	0x026	RW	uint8_t	[0;1]	FULL = 0, DIRONLY = 1

motorPosPidKp	0x030	RW	float	-	position PID proportional gain
motorPosPidKi	0x031	RW	float	-	position PID integral gain
motorPosPidKd	0x032	RW	float	-	position PID derivative gain
motorPosPidWindup	0x034	RW	float	-	position PID integral windup limit

motorVelPidKp	0x040	RW	float	-	velocity PID proportional gain
motorVelPidKi	0x041	RW	float	-	velocity PID integral gain
motorVelPidKd	0x042	RW	float	-	velocity PID derivative gain
motorVelPidWindup	0x044	RW	float	-	velocity PID integral windup limit

motorImpPidKp	0x050	RW	float	-	impedance PD proportional gain
motorImpPidKd	0x051	RW	float	-	impedance PD derivative gain

mainEncoderVelocity	0x062	RO	float	-	main encoder velocity in rad/s (calculated in a 40kHz loop)
mainEncoderPosition	0x063	RO	float	-	main encoder position in rad (read in 40kHz loop)
motorTorque	0x064	RO	float	-	motor output shaft torque in Nm (read in 40kHz loop)


runSaveCmd	0x080	WO	uint8_t	other than 0 to run	save non-volatile memory
runTestMainEncoderCmd	0x081	WO	uint8_t	other than 0 to run	runs main encoder test routine
runTestOutputEncoderCmd	0x082	WO	uint8_t	other than 0 to run	runs output encoder test routine
runCalibrateCmd	0x083	WO	uint8_t	other than 0 to run	runs main calibration routine
runCalibrateOutputEncoderCmd	0x084	WO	uint8_t	other than 0 to run	runs output encoder calibration routine
runCalibratePiGains	0x085	WO	uint8_t	other than 0 to run	runs current PI loop calibration routine
runRestoreFactoryConfig	0x087	WO	uint8_t	other than 0 to run	reverts config to factory state
runReset	0x088	WO	uint8_t	other than 0 to run	resets the controller
runClearWarnings	0x089	WO	uint8_t	other than 0 to run	clears all warnings
runClearErrors	0x08A	WO	uint8_t	other than 0 to run	clears non-critical errors
runBlink	0x08B	WO	uint8_t	other than 0 to run	blinks onboard LEDs
runZero	0x08C	WO	uint8_t	other than 0 to run	sets new zero position
runCanReinit	0x08D	WO	uint8_t	other than 0 to run	reinitializes can peripheral

calOutputEncoderStdDev	0x100	RO	float	-	output encoder test result (standard deviation)
calOutputEncoderMinE	0x101	RO	float	-	output encoder test result (min error)
calOutputEncoderMaxE	0x102	RO	float	-	output encoder test result (max error)
calMainEncoderStdDev	0x103	RO	float	-	main encoder test result (standard deviation)
calMainEncoderMinE	0x104	RO	float	-	main encoder test result (min error)
calMainEncoderMaxE	0x105	RO	float	-	main encoder test result (max error)

positionLimitMax	0x110	RW	float	-	maximum valid position
positionLimitMin	0x111	RW	float	-	minimum valid position
maxTorque	0x112	RW	float	-	maximum torque
maxVelocity	0x113	RW	float	-	maximum velocity
maxAcceleration	0x114	RW	float	-	maximum acceleration
maxDeceleration	0x115	RW	float	-	maximum deceleration

profileVelocity	0x120	RW	float	-	profile velocity
profileAcceleration	0x121	RW	float	-	profile acceleration
profileDeceleration	0x122	RW	float	-	profile deceleration
quickStopDeceleration	0x123	RW	float	-	quick stop deceleration in case of a non-critical error
positionWindow	0x124	RW	float	-	position window within position is considered to be reached
velocityWindow	0x125	RW	float	-	velocity window within velocity is considered to be reached

motionModeCommand	0x140	WO	uint8_t	-	commands a motion mode change: IDLE = 0x00, POSITION_PID = 0x01, VELOCITY_PID = 0x02, RAW_TORQUE = 0x03, IMPEDANCE = 0x04, POSITION_PROFILE = 0x07, VELOCITY_PROFILE = 0x08
motionModeStatus	0x141	RO	uint8_t	-	shows the currently set motion mode
state	0x142	RW	uint16_t	-	returns the internal state machine state of the controller

targetPosition	0x150	RW	float	-	sets target position in rad
targetVelocity	0x151	RW	float	-	sets target velocity in rad/s
targetTorque	0x152	RW	float	-	sets target torque in Nm


userGpioConfiguration	0x160	RW	uint8_t	-	0 - OFF, 1 - AUTO-BRAKE, 2 - GPIO INPUT
userGpioState	0x161	RO	uint16_t	-	GPIO input state

reverseDirection	0x600	RW	uint8_t	-	used to change the direction of the main encoder when using other encoders than the onboard one. Always recalibrate after changing this setting

shuntResistance	0x700	RO	float	[0.001 - 0.01]	Current sense resistor value. Setting this register to a value that is not coherent with the hardware may damage the controller. In this cases warranty is not respected.

buildDate	0x800	RO	uint32_t	-	software build date
commitHash	0x801	RO	char[8]	-	commit hash
firmwareVersion	0x802	RO	uint32_t	-	-
hardwareVersion	0x803	RO	uint8_t	-	-
bridgeType	0x804	RO	uint8_t	-	type of the mosfet driver
quickStatus	0x805	RO	uint16_t	-	quick status vector
mosfetTemperature	0x806	RO	float	-	power stage temperature
motorTemperature	0x807	RO	float	-	motor temperature (if thermistor is mounted)
motorShutdownTemp	0x808	RW	uint8_t	-	temperature at which the MD series motor controller will enter IDLE mode
mainEncoderErrors	0x809	RO	uint32_t	-	main encoder errors
outputEncoderErrors	0x80A	RO	uint32_t	-	output encoder errors
calibrationErrors	0x80B	RO	uint32_t	-	calibration errors
bridgeErrors	0x80C	RO	uint32_t	-	bridge errors
hardwareErrors	0x80D	RO	uint32_t	-	hardware errors
communicationErrors	0x80E	RO	uint32_t	-	communication errors
motionErrors	0x810	RO	uint32_t	-	motion errors