Encoder Use (EMOD) in MBS file

[Myzhar]

Hello,

I'm not sure about the correctness of the following line:
https://github.com/g/roboteq/blob/indigo-devel/roboteq_driver/mbs/script.mbs#L30

I don't understand why you set EMOD to 50, while according to manual (p210) it should be
aa + mot1 * 16 + mot2 * 32

so

18 for motor 1 with feedback
34 for motor 2 with feedback

Walter

[Myzhar]

Hi,

I have found a post on the RoboteQ's forum that confirms my hypothesis:
http://www.roboteq.com/index.php/forum/9-controller-configuration-operation/29529115-encoder-setup-after-controller-reset

[romainreignier]

I would be also interested in the answer.

[rbonghi]

I @romainreignier I've fixed the @Myzhar with a new node in ROS https://github.com/rbonghi/roboteq_control This node works with ROS control and with the dynamic reconfigurator you can dinamically update the roboteq configuration. It has look like RoboRun+

[romainreignier]

@rbonghi yes, I have seen your node. The concept to use only the serial comm and no script and the dynamic reconfigure for everything is very interesting.
But I do not like in your code that you call the nh.getParam() every time in each callback because it takes some time so it is not very efficient. And I do not have the time right now to modify it.

[rbonghi]

yes, I know only this way :-( to dynamically update the the parameters when the motors run. If you know other solution I am interested to change this approach.

At this time I've tested with a Jetson TX2 and works better without delay.

[romainreignier]

This issue is not the right place for this discussion. But usually we use class members updated in the dynamic_reconfigure callback for this.