Example of ethercat_generic_plugins/EcCiA402Drive

[jaronski]

Hello,
would it be possible to show an example of how to integrate the EcCiA402Drive into a xacro and the corresponding config.yaml?
i have a CiA402 Drive that i would like to use with the ethercat_driver_ros2 but i am having trouble with the configuration. I adjusted the Configuration File as shown in the MaxonEpos3 example to the appropriate values to match my drive (rpdo, tpdo). I created a urdf.xacro with the ros2_control tag where the ethercat_driver is started and configured 1 joint with command interface position and velocity as ec_module i choose the plugin ethercat_generic_plugins/EcCiA402Drive put the alias and position on the ethercat bus mode_of_operation to 8 or 9 depending on what i want to test out and the param slave_config i put the config.yaml i adjusted.

how can i claim the <command_interface name="reset_fault"/> and to what controller? Can you show this for the EPOS3?
i am able to start the drive and move it to a position sometimes and then i restart and when i launch everything again sometimes the drive doesn't get into Operation Enabled and i don't understand why.

Any help or pointers in the right direction would be much appreciated
best regards
jaronski

[mcbed]

Hi @jaronski,
If your drive is in fault it will not be able to switch to OE mode and the fault needs to be reset. This can be done in 2 ways:

1. use the control word to do fault reset using SDO by running ethercat -m 0 -p 0 download -t uint16 0x6040 0x00 128 to set the fault reset bit to 1.
2. use the command interface reset_fault and interface it with a controller. It does fault reset on rising edge signal on the interface (0 -> 1). There is a controller that can do this on this branch. You configure this controller like this:

 controller_manager:
   ros__parameters:
     update_rate: 100  # Hz

     gpio_command_controller:
       type: gpio_controllers/GpioCommandController

 gpio_command_controller:
   ros__parameters:
     gpios:
       - joint_1
     command_interfaces:
        joint_1:
         - reset_fault

Note: you need to add <state_interface name="reset_fault"/> to your urdf.

[jaronski]

Hi @mcbed
thanks for your response. I will check that out first thing on Monday 😉

best regards and a great weekend

[iamdrfly]

Hi @mcbed,
may i ask why he needs to add <state_interface name="reset_fault"/> instead of <command_interface name="reset_fault"/>
?

Best Regards
Alessio

[mcbed]

Hi @AlessioMosca,
You need both for the controller to configure. It is a incomplete feature but I did not have the time to fix it yet.

[mcbed]

Hi @jaronski,
There is now the possibility to reset fault using a ros service, check out PR #61 if interested

[qleonardolp]

Dear, @mcbed
About the fault reset feature, would it be a problem to use the control_word command interface directly to perform the fault reset, writting 0x80, or 128 on it?

I'm writting a custom controller, based on the joint trajectory controller, and I'm interested on have a controller able to do fault handling.