[ZeroErr eRob] Joint not moving using EcCia402Drive + Joint trajectory position

[alesof]

Hi there!

I am trying to move an actuator using the generic plugin EcCiA402Drive. The device cannot switch to OP state due to a configuration error, can someone provide any help?

The error i get is: AL status message 0x001E: "Invalid input configuration". The actuator name is: erob ZeroErr

sudo dmseg:

[ 7812.078596] EtherCAT DEBUG 0: Adding datagram pair with expected WC 3.
[ 7812.078597] EtherCAT 0: Domain0: Logical address 0x00000000, 20 byte, expected working counter 3.
[ 7812.078598] EtherCAT 0:   Datagram domain0-0-main: Logical offset 0x00000000, 20 byte, type LRW.
[ 7812.078599] EtherCAT DEBUG 0: Stopping master thread.
[ 7812.078632] EtherCAT DEBUG 0: Master IDLE thread exiting...
[ 7812.078651] EtherCAT 0: Master thread exited.
[ 7812.078652] EtherCAT DEBUG 0: FSM datagram is 0000000088c1f073.
[ 7812.078653] EtherCAT 0: Starting EtherCAT-OP thread.
[ 7812.078815] EtherCAT DEBUG 0: Operation thread running with fsm interval = 4000 us, max data size=45000
[ 7812.078817] EtherCAT DEBUG 0: mmap()
[ 7812.078818] EtherCAT WARNING 0: 1 datagram UNMATCHED!
[ 7812.078820] EtherCAT DEBUG 0: Vma fault, offset = 0, page = 00000000dfd09629
[ 7813.086330] EtherCAT DEBUG 0: Configuration changed.
[ 7813.086333] EtherCAT DEBUG 0-0: Checking system time offset.
[ 7813.106353] EtherCAT DEBUG 0-0: DC 64 bit system time offset calculation: system_time=751649702317787824 (corrected with 16000000), app_time=751649702316035076, diff=-1752748
[ 7813.106356] EtherCAT DEBUG 0-0: Setting time offset to 751641523853848985 (was 751641523855601733)
[ 7813.122353] EtherCAT DEBUG 0: Requesting OP...
[ 7813.162369] EtherCAT DEBUG 0-0: Changing state from PREOP to OP.
[ 7813.162374] EtherCAT DEBUG 0-0: Configuring...
[ 7813.190944] EtherCAT DEBUG 0-0: Now in INIT.
[ 7813.190945] EtherCAT DEBUG 0-0: Clearing FMMU configurations...
[ 7813.201000] EtherCAT DEBUG 0-0: Clearing sync manager configurations...
[ 7813.211005] EtherCAT DEBUG 0-0: Clearing DC assignment...
[ 7813.221006] EtherCAT DEBUG 0-0: Configuring mailbox sync managers...
[ 7813.221007] EtherCAT DEBUG 0-0: SM0: Addr 0x1000, Size 128, Ctrl 0x26, En 1
[ 7813.221008] EtherCAT DEBUG 0-0: SM1: Addr 0x1080, Size 128, Ctrl 0x22, En 1
[ 7813.231004] EtherCAT DEBUG 0-0: Assigning SII access to PDI.
[ 7813.260944] EtherCAT DEBUG 0-0: Now in PREOP.
[ 7813.260945] EtherCAT DEBUG 0-0: Assigning SII access back to EtherCAT.
[ 7813.270968] EtherCAT DEBUG 0-0: SM2: Addr 0x1100, Size   0, Ctrl 0x64, En 0
[ 7813.270969] EtherCAT DEBUG 0-0: SM3: Addr 0x1400, Size   0, Ctrl 0x20, En 0
[ 7813.300998] EtherCAT ERROR 0-0: Failed to set SAFEOP state, slave refused state change (PREOP + ERROR).
[ 7813.310940] EtherCAT ERROR 0-0: AL status message 0x001E: "Invalid input configuration".
[ 7813.330965] EtherCAT 0-0: Acknowledged state PREOP.

zeroerr_config.yaml

# Configuration file for ZERO ERR drive
vendor_id: 0x5A65726F
product_id: 0x00029252
assign_activate: 0x0300  # DC Synch register - 
auto_fault_reset: false  # true = automatic fault reset, false = fault reset on rising edge command interface "reset_fault"
rpdo:  # RxPDO = receive PDO Mapping
  - index: 0x1600
    channels:
      - {index: 0x607a, sub_index: 0, type: int32, command_interface: position, default: .nan}  # Target position
      - {index: 0x60fe, sub_index: 0, type: int32, default: 0}  # Digital Output Functionalities
      - {index: 0x6040, sub_index: 0, type: uint16, default: 0}  # Control word
tpdo:  # TxPDO = transmit PDO Mapping 
  - index: 0x1A00
    channels:
      - {index: 0x6064, sub_index: 0, type: int32, state_interface: position}  # Position actual value
      - {index: 0x60fd, sub_index: 0, type: int32}  # Digital Input Functionalities
      - {index: 0x6041, sub_index: 0, type: uint16}  # Status word
sm:  # Sync Manager
  - {index: 0, type: output, pdo: ~, watchdog: disable}
  - {index: 1, type: input, pdo: ~, watchdog: disable}
  - {index: 2, type: output, pdo: rpdo, watchdog: enable}
  - {index: 3, type: input, pdo: tpdo, watchdog: disable}

$ ethercat cstruct

/* Master 0, Slave 0, "ZeroErr Driver"
 * Vendor ID:       0x5a65726f
 * Product code:    0x00029252
 * Revision number: 0x00000001
 */

ec_pdo_entry_info_t slave_0_pdo_entries[] = {
    {0x607a, 0x00, 32}, /* Target Position */
    {0x60fe, 0x00, 32}, /* Digital outputs */
    {0x6040, 0x00, 16}, /* Control Word */
    {0x6064, 0x00, 32}, /* Position Actual Value */
    {0x60fd, 0x00, 32}, /* Digital inputs */
    {0x6041, 0x00, 16}, /* Status Word */
};

ec_pdo_info_t slave_0_pdos[] = {
    {0x1600, 3, slave_0_pdo_entries + 0}, /* R0PDO */
    {0x1a00, 3, slave_0_pdo_entries + 3}, /* T0PDO */
};

ec_sync_info_t slave_0_syncs[] = {
    {0, EC_DIR_OUTPUT, 0, NULL, EC_WD_DISABLE},
    {1, EC_DIR_INPUT, 0, NULL, EC_WD_DISABLE},
    {2, EC_DIR_OUTPUT, 1, slave_0_pdos + 0, EC_WD_ENABLE},
    {3, EC_DIR_INPUT, 1, slave_0_pdos + 1, EC_WD_DISABLE},
    {0xff}
};

Thank you!

[alesof]

EDIT: the config file is correct and I'm able to set the OP state. For some reason the config file it was being launched wasn't updated.

Anyway now when I send a new position the controller doesn't work and the motor doesn't move. I am able to move the motor, and the communication is ok as if I set the default position to a value different then zero it moves there. Therefore, the problem is when using the trajectory controller, any help would be appreciated.

[mcbed]

Hi, can you include the log of your terminal running ros2_control ?

[JensVanhooydonck]

Which operation mode is used?

[MEJamnezhad]

Hi, I have the same problem with the erob ZeroErr. I can't reach OP state. Can you describe how you reached the OP state? Thanks

[JensVanhooydonck]

You must include the operation mode in the tpdo, otherwise your position isn't forwarded. I searched long after this..

I will try to open a PR which falls back to the setup operation mode if there is no config in tpdo.

[alesof]

Hi there, the working configuration file for me was:

zeroerror_config.yaml

# Configuration file for ZERO ERR drive
vendor_id: 0x5A65726F
product_id: 0x00029252
assign_activate: 0x003c
auto_fault_reset: true 
rpdo:  
  - index: 0x1600
    channels:
      - {index: 0x607a, sub_index: 0, type: int32, command_interface: position, default: 0}  
      - {index: 0x60fe, sub_index: 0, type: int32, default: 0} 
      - {index: 0x6040, sub_index: 0, type: uint16, default: 0}
tpdo:  
  - index: 0x1A00
    channels:
      - {index: 0x6064, sub_index: 0, type: int32, state_interface: position}
      - {index: 0x60fd, sub_index: 0, type: int32}
      - {index: 0x6041, sub_index: 0, type: uint16} 
sm:  
  - {index: 0, type: output, pdo: ~, watchdog: disable}
  - {index: 1, type: input, pdo: ~, watchdog: disable}
  - {index: 2, type: output, pdo: rpdo, watchdog: enable}
  - {index: 3, type: input, pdo: tpdo, watchdog: disable}

One thing was that the default position couldn't be set to .nan but works with zero. Of course this produces another error, if the motor is powered on when its position is not near zero the OP state won't be set to ready but in fault.

This is the ros2_control config file:

<?xml version="1.0"?>
<robot xmlns:xacro="http://www.ros.org/wiki/xacro">

  <xacro:macro name="motor_drive">

    <ros2_control name="motor_drive" type="system">
      <hardware>
        <plugin>ethercat_driver/EthercatDriver</plugin>
        <param name="master_id">0</param>
        <param name="control_frequency">1000</param>
      </hardware>

      <joint name="joint_1">
        <state_interface name="position"/>
        <command_interface name="position"/>
        <ec_module name="ZeroErr">
          <plugin>ethercat_generic_plugins/EcCiA402Drive</plugin>
          <param name="alias">0</param>
          <param name="position">0</param>
          <param name="mode_of_operation">8</param>
          <param name="slave_config">$(find ethercat_motor_drive)/config/zeroerr_config.yaml</param>
        </ec_module>
      </joint>
    </ros2_control>

  </xacro:macro>

</robot>

Sorry if I didn't answer earlier but I didn't have the motor to make any test for a while. I'll try in the next days to make it work with ros2 control, if anyone else has attempted and succeeded let us know!

[brennand]

I would also be interested in the solution for these motors. I can talk to them and also set and read the values for example the position and write the control words eta. But I still can't move the motors and have them all hooked up to a ros2 controller, as I can print the desired position. Any input on how you managed to get them to work would be good.

[JensVanhooydonck]

The main problem is that the mode of operation display is set to 0 -> ModeOfOperation::MODE_NO_MODE https://github.com/ICube-Robotics/ethercat_driver_ros2/blob/9657cf3c53d21337c6aedacea25e483b5afb4ab7/ethercat_generic_plugins/ethercat_generic_cia402_drive/include/ethercat_generic_plugins/generic_ec_cia402_drive.hpp#L50

This should initially be set to the operation mode given in de config mapping. Otherwise no commands will be send to the position adres.

Temporary work around is to add - {index: 0x6061, sub_index: 0, type: uint8 } # Mode of operation to tpdo:

example from above:

# Configuration file for ZERO ERR drive
vendor_id: 0x5A65726F
product_id: 0x00029252
assign_activate: 0x003c
auto_fault_reset: true 
rpdo:  
  - index: 0x1600
    channels:
      - {index: 0x607a, sub_index: 0, type: int32, command_interface: position, default: .nan}  
      - {index: 0x60fe, sub_index: 0, type: int32, default: 0} 
      - {index: 0x6040, sub_index: 0, type: uint16, default: 0}
tpdo:  
  - index: 0x1A00
    channels:
      - {index: 0x6064, sub_index: 0, type: int32, state_interface: position}
      - {index: 0x6061, sub_index: 0, type: uint8 }  # Mode of operation
      - {index: 0x60fd, sub_index: 0, type: int32}
      - {index: 0x6041, sub_index: 0, type: uint16} 
sm:  
  - {index: 0, type: output, pdo: ~, watchdog: disable}
  - {index: 1, type: input, pdo: ~, watchdog: disable}
  - {index: 2, type: output, pdo: rpdo, watchdog: enable}
  - {index: 3, type: input, pdo: tpdo, watchdog: disable}

This wil set the vallue correctly here: https://github.com/ICube-Robotics/ethercat_driver_ros2/blob/9657cf3c53d21337c6aedacea25e483b5afb4ab7/ethercat_generic_plugins/ethercat_generic_cia402_drive/src/generic_ec_cia402_drive.cpp#L78

[alesof]

Hi @JensVanhooydonck! I tried adding the Mode of operation using - {index: 0x6061, sub_index: 0, type: uint8 } as you suggested, but this makes the motor go into preop error state. Did you manage to use it on the Zero err?

ethercat slaves returns: 0 0:0 PREOP + ZeroErr Driver

@mcbed when i launch it with the configuration that i published in previous comments the ros2_control returns:

[ros2_control_node-1] [INFO] [1706526480.322534728] [EthercatDriver]: Starting ...please wait...
[ros2_control_node-1] {0, 0, 0x5a65726f, 0x29252, 0x607a, 0x0}
[ros2_control_node-1] {0, 0, 0x5a65726f, 0x29252, 0x60fe, 0x0}
[ros2_control_node-1] {0, 0, 0x5a65726f, 0x29252, 0x6040, 0x0}
[ros2_control_node-1] {0, 0, 0x5a65726f, 0x29252, 0x6064, 0x0}
[ros2_control_node-1] {0, 0, 0x5a65726f, 0x29252, 0x60fd, 0x0}
[ros2_control_node-1] {0, 0, 0x5a65726f, 0x29252, 0x6041, 0x0}
[ros2_control_node-1] [INFO] [1706526480.322851982] [EthercatDriver]: Activated EcMaster!
[ros2_control_node-1] 1 slave(s).
[ros2_control_node-1] Master AL states: 0x02.
[ros2_control_node-1] Link is up.
[ros2_control_node-1] Slave: State 0x02.
[ros2_control_node-1] Slave: online.
[ros2_control_node-1] STATE: Not Ready to Switch On with status word :0
[ros2_control_node-1] [INFO] [1706526481.323316459] [EthercatDriver]: updated!
[ros2_control_node-1] [INFO] [1706526481.323329363] [EthercatDriver]: System Successfully started!
[ros2_control_node-1] [INFO] [1706526481.323336627] [resource_manager]: Successful 'activate' of hardware 'motor_drive'
[ros2_control_node-1] [INFO] [1706526481.325584271] [controller_manager]: update rate is 1000 Hz
[ros2_control_node-1] [WARN] [1706526481.325740030] [controller_manager]: Could not enable FIFO RT scheduling policy
[ros2_control_node-1] Slave: State 0x01.
[ros2_control_node-1] Slave: State 0x02.
[ros2_control_node-1] Domain: WC 3.
[ros2_control_node-1] Domain: State 2.
[ros2_control_node-1] STATE: Fault with status word :4616
[ros2_control_node-1] Slave: State 0x08.
[ros2_control_node-1] Slave: operational.
[ros2_control_node-1] STATE: Switch on Disabled with status word :4688
[ros2_control_node-1] STATE: Ready to Switch On with status word :4657
[ros2_control_node-1] STATE: Switch On with status word :4659
[ros2_control_node-1] Master AL states: 0x08.
[ros2_control_node-1] STATE: Operation Enabled with status word :4663
[ros2_control_node-1] [INFO] [1706526481.508627839] [controller_manager]: Loading controller 'joint_state_broadcaster'
[ros2_control_node-1] [INFO] [1706526481.513056653] [controller_manager]: Loading controller 'trajectory_controller'
[spawner-3] [INFO] [1706526481.515047005] [spawner_joint_state_broadcaster]: Loaded joint_state_broadcaster
[ros2_control_node-1] [WARN] [1706526481.516423306] [trajectory_controller]: [Deprecated]: "allow_nonzero_velocity_at_trajectory_end" is set to true. The default behavior will change to false.
[ros2_control_node-1] [INFO] [1706526481.517045064] [controller_manager]: Configuring controller 'joint_state_broadcaster'
[ros2_control_node-1] [INFO] [1706526481.517089710] [joint_state_broadcaster]: 'joints' or 'interfaces' parameter is empty. All available state interfaces will be published
[spawner-4] [INFO] [1706526481.517324299] [spawner_trajectory_controller]: Loaded trajectory_controller
[ros2_control_node-1] [INFO] [1706526481.519045399] [controller_manager]: Configuring controller 'trajectory_controller'
[ros2_control_node-1] [INFO] [1706526481.519093188] [trajectory_controller]: No specific joint names are used for command interfaces. Using 'joints' parameter.
[ros2_control_node-1] [INFO] [1706526481.519102575] [trajectory_controller]: Command interfaces are [position] and state interfaces are [position].
[ros2_control_node-1] [INFO] [1706526481.519108770] [trajectory_controller]: Using 'splines' interpolation method.
[ros2_control_node-1] [INFO] [1706526481.519489758] [trajectory_controller]: Controller state will be published at 50.00 Hz.
[ros2_control_node-1] [INFO] [1706526481.520063999] [trajectory_controller]: Action status changes will be monitored at 20.00 Hz.
[spawner-3] [INFO] [1706526481.523317968] [spawner_joint_state_broadcaster]: Configured and activated joint_state_broadcaster
[spawner-4] [INFO] [1706526481.525604155] [spawner_trajectory_controller]: Configured and activated trajectory_controller
[INFO] [spawner-3]: process has finished cleanly [pid 71207]
[INFO] [spawner-4]: process has finished cleanly [pid 71209]

[alesof]

@JensVanhooydonck by hardcoding 8 as mode_of_operationdisplay i am able to move it. What's a clean way to solve this? 😅

[JensVanhooydonck]

@JensVanhooydonck by hardcoding 8 as mode_of_operationdisplay i am able to move it. What's a clean way to solve this? 😅

That was my next possible try..

I wanted to add a PR which fills in the inital value of mode_of_operationdisplay with the value defined here:

<ros2_control name="ec_single_axis" type="system">
    <hardware>
      <plugin>ethercat_driver/EthercatDriver</plugin>
      <param name="master_id">0</param>
      <param name="control_frequency">100</param>
    </hardware>

   <joint name="joint_0">
      <state_interface name="position"/>
      <state_interface name="velocity"/>
      <state_interface name="effort"/>
      <command_interface name="position"/>
      <command_interface name="reset_fault"/>
      <ec_module name="MAXON">
        <plugin>ethercat_generic_plugins/EcCiA402Drive</plugin>
        <param name="alias">0</param>
        <param name="position">0</param>
        <param name="mode_of_operation">8</param>
        <param name="slave_config">/path/to/maxon.yaml</param>
      </ec_module>
    </joint>
  </ros2_control>

Currently not had the time for this. I guess this can easily be done in: https://github.com/ICube-Robotics/ethercat_driver_ros2/blob/9657cf3c53d21337c6aedacea25e483b5afb4ab7/ethercat_generic_plugins/ethercat_generic_cia402_drive/src/generic_ec_cia402_drive.cpp#L131 So we also set display instead of only mode_ofoperation

Haven't tested this yet..

Could you let me know if this works?

[alesof]

Ok never mind, I wrote too soon ahah. I am able to set it from the tpdo config but CiA402D_TPDO_MODE_OF_OPERATION_DISPLAY is a uint16!

For whoever is watching, the correct config file, in order to move the motor using ros2_control is the following:

# Configuration file for ZERO ERR drive
vendor_id: 0x5A65726F
product_id: 0x00029252
assign_activate: 0x003c
auto_fault_reset: true 

rpdo: 
  - index: 0x1600
    channels:
      - {index: 0x607a, sub_index: 0, type: int32, command_interface: position, default: .nan} 
      - {index: 0x60fe, sub_index: 0, type: int32, default: 0}
      - {index: 0x6040, sub_index: 0, type: uint16, default: 0}
tpdo:  
  - index: 0x1A00
    channels:
      - {index: 0x6064, sub_index: 0, type: int32, state_interface: position} 
      - {index: 0x60fd, sub_index: 0, type: int32} 
      - {index: 0x6041, sub_index: 0, type: uint16} 
      - {index: 0x6061, sub_index: 0, type: uint16, default: 8}

sm: 
  - {index: 0, type: output, pdo: ~, watchdog: disable}
  - {index: 1, type: input, pdo: ~, watchdog: disable}
  - {index: 2, type: output, pdo: rpdo, watchdog: enable}
  - {index: 3, type: input, pdo: tpdo, watchdog: disable}

[JensVanhooydonck]

Ok never mind, I wrote too soon ahah. I am able to set it from the tpdo config but CiA402D_TPDO_MODE_OF_OPERATION_DISPLAY is a uint16!

For whoever is watching, the correct config file, in order to move the motor using ros2_control is the following:

# Configuration file for ZERO ERR drive
vendor_id: 0x5A65726F
product_id: 0x00029252
assign_activate: 0x003c
auto_fault_reset: true 

rpdo: 
  - index: 0x1600
    channels:
      - {index: 0x607a, sub_index: 0, type: int32, command_interface: position, default: .nan} 
      - {index: 0x60fe, sub_index: 0, type: int32, default: 0}
      - {index: 0x6040, sub_index: 0, type: uint16, default: 0}
tpdo:  
  - index: 0x1A00
    channels:
      - {index: 0x6064, sub_index: 0, type: int32, state_interface: position} 
      - {index: 0x60fd, sub_index: 0, type: int32} 
      - {index: 0x6041, sub_index: 0, type: uint16} 
      - {index: 0x6061, sub_index: 0, type: uint16, default: 8}

sm: 
  - {index: 0, type: output, pdo: ~, watchdog: disable}
  - {index: 1, type: input, pdo: ~, watchdog: disable}
  - {index: 2, type: output, pdo: rpdo, watchdog: enable}
  - {index: 3, type: input, pdo: tpdo, watchdog: disable}

I think this is dependend on the mapping of the drive. On my drive this was a uint8 (: I still think the proper solution would be to alter https://github.com/ICube-Robotics/ethercat_driver_ros2/blob/9657cf3c53d21337c6aedacea25e483b5afb4ab7/ethercat_generic_plugins/ethercat_generic_cia402_drive/src/generic_ec_cia402_drive.cpp#L131 to also set the mode_of_operationdisplay

But good that it works now!

EDIT Also the default: 8 should not be necessary.. This value is now read from the drive it self.

[alesof]

Thanks again for your help 😁 Time permitting I'll also try a more robust approach!

[brennand]

Hi, I've have the drives sort of working with ros2 control at the moment, but there aren't that robust e.g. you have to restart them a couple of times before they are working all together properly. I made some minor changes to the code that might not be needed but made my life easier:

   last_value = factor * last_value + offset;
    return last_value;

I removed the factor / offset so we can get the raw values from the motors

   {
      ec_read(domain_address);
      if (interface_index >= 0) {
        state_interface_ptr_->at(interface_index) = factor * last_value + offset;
      }

And then set the factor and offset here. This way I can get the raw motor position to set as the initial position for the motors.

Then in the drive file I just made this change:

if (mode_of_operation_display_ != ModeOfOperation::MODE_NO_MODE)

I change it to this:

if (int(last_position_) != 0 && !std::isnan(last_position_))

The thinking behind it. We initial the last_position as a nan in the header so check to make sure we have a value from the motor and then to make sure we have a none zero, as the motor position flickers +- 2 raw position values. I also had to add this as the first compel of readings I get from the motor are zero.

This is my working yaml:

# Configuration file for ZERO ERR drive
vendor_id: 0x5A65726F
product_id: 0x00029252
#assign_activate: 0x003c
auto_fault_reset: true
rpdo:
  - index: 0x1600
    channels:
      - {index: 0x607a, sub_index: 0, type: int32, command_interface: position, default: 0.0, factor: 83443.4044753}
      - {index: 0x60fe, sub_index: 0, type: int32, default: 0.0} # Digital Inputs
      - {index: 0x6040, sub_index: 0, type: uint16, default: 0.0} # Control word
tpdo:
  - index: 0x1A00
    channels:
      - {index: 0x6064, sub_index: 0, type: int32, state_interface: position, factor: 0.00001198417}
      - {index: 0x60fd, sub_index: 0, type: int32} # Digital Inputs
      - {index: 0x6041, sub_index: 0, type: uint16, state_interface: motor_status_word}
sm:
  - {index: 0, type: output, pdo: ~, watchdog: disable}
  - {index: 1, type: input, pdo: ~, watchdog: disable}
  - {index: 2, type: output, pdo: rpdo, watchdog: enable}
  - {index: 3, type: input, pdo: tpdo, watchdog: disable}

[brennand]

@alesof Where did you find this value:

assign_activate: 0x003c

I've looked in the data sheet and can't find it any where?

[alesof]

@brennand, I initially started with a configuration file I had written for another Chinese actuator (Maxon and others share the same address). I subsequently modified it to suit the new actuator and left it there even if I didn't actually find a reference in the datasheet. I'm just experimenting with them during my free time and not actively working on them. Nonetheless, I've created a public repository for the automatic setup and usage, I'll share any update there. I'll take a look at your solution. Thanks!

[alejomancinelli]

@brennand @alesof Hello. I'm also trying to make a ZeroErr eRob motor work with ethercat in ROS2. I try with the configuration file discussed early and it worked, at least once. It generally loops through the following states:

[ros2_control_node-1] STATE: Switch On with status word :4787
[ros2_control_node-1] STATE: Fault with status word :4744
[ros2_control_node-1] STATE: Switch on Disabled with status word :4816
[ros2_control_node-1] STATE: Ready to Switch On with status word :4785
[ros2_control_node-1] STATE: Switch On with status word :4787

Sometimes it says it has reached the OP state, but it doesn't work. Some other times it reaches the SAFEOP+ERROR, or just the PREOP+ERROR state.

I also notice, checking the eTunner software, that the operation mode never changes to 8, it's always at 3. The yaml file is missing the corresponding rpdo (i guess, I'm new to Ethercat), but adding it didn't change anything. Apart from that, in the monitor it shows the 33792 error: The velocity error exceeds the limit value, although I've never modify any velocity value. Have you made any progress to a more robust solution? Maybe it would be better to open a separate issue for this particular motor.

[alesof]

Hi there, I'll be updating the public repo on my profile as soon I'll be able to work again on it. In the meantime I can tell you that I experienced the same error (velocity ecxeeds...) and it should be related to the current position beeing too far away from the commanded one. Using .nan (instead of zero) as default value removed the issue for me. But again, i didn't make many test! Probably @brennand solution using last_value is solving the same issue.

As the issue is related to this specific motor I edited the title.

[alejomancinelli]

A little update: I did the changes you suggested and I managed to get it into de OP state every time by launching the "joint trajectory controller" a little after everything else. It seems like, as you said, the current position was too far away from the target position. I let the motor get to the OP state and then I launch the "joint trajectory controller" (so far manually, I've been only doing tests).

It still doesn't work perfectly with MoveIt2, but I didn't modify many configurations parameters yet. It starts moving but reaches the target before the final position, and then brakes, or it diverges too much, and brakes.

Some other notes:

• I changed the scaling factor in the yaml file to work in radians instead of degrees, since MoveIt2 uses radians.
• In the OP state without the "joint trajectory controller", the position slowly decreases. I still don't get why this happens.
• The Operation Mode in the oscilloscope shows 3, Profile Velocity mode, when it should be 8, Cyclic Synchronous Position mode. I did many tests trying to change it but so far I couldn't make it. Nevertheless, it seems like it works anyways.
• Sometimes, just having everything turned on but without movement, the current position gets a little off and the motor brakes. Then it gets back to the OP state. Maybe there is a tolerance factor that can be modify.
• When trying to move, if the current positions diverges too much from the target position, I have to re-launch everything. This happens every time the MoveIt2 fails, as I explained before. The target position should match again the current position to get again into the OP state. Maybe a reset button could be implemented.

[JensVanhooydonck]

If you do not map the tpdo:

• {index: 0x6061, sub_index: 0, type: uint8 } # Mode of operation

The operation mode will not be altered. You must add this to the tpdo. That should change the operating mode to 8 instead of 3..

[brennand]

This is what I learned about the velocity limit error. When we first connect to the motors, we do not get a valid position from the motor. So ROS2 control thinks the initial position is Zero and sets that the the desired position, that will course the motor to try and turn at great speed. I hack my version of this code to add:

        if (static_cast<int>(last_value) == 0 && interface_index == 0) {
          return;
        } 

after the ec_read(domain_address); With my hacky assumption that the motor should never read zero..... but it works. You can see some other small changes I made to the code base on my page.

I still find it very hard to startup 7 motors and they all start with no errors. normally takes a couple of attempts. I need to really write some code to clear the errors and try booting again.

[ZeroErrControl]

Streamline your eRob integration with our latest open-source software, drivers, and examples now available on GitHub: github.com/ZeroErrControl Find resources for: • ROS2 & MoveIt2 • TwinCAT3 & Python • CAN, CANopen, EtherCAT, SOEM • Linux & Windows • USD Files & LLM Control (Isaac Sim) Explore the repositories and get started today!

[alesof]

I am closing this, as there is an official repo now from @ZeroErrControl. I have opensourced the code I've been using too.