Our top priority

We are as proud of our service as we are of our robots.

Forgot your password?
Register

Splash Forums Development SDK Some questions about force control

This topic contains 29 replies, has 6 voices, and was last updated by  cortesao 1 year, 4 months ago.

Viewing 15 posts - 16 through 30 (of 30 total)
  • Author
    Posts
  • #2147

    Anonymous

    Hi jiangxihj

    My guess is that your robot does not switch to torque mode. Usually, it does so when the gravity torques are not accurate. Could you break in your code after ‘MySetTorqueControlType(DIRECTTORQUE);’ and tell me if you have switched in torque control (the robot should be very easy to move by hand, and should not react to joystick commands)?

    Thanks

    Martine

    #2157

    Anonymous

    Hi, Martine

    Thank you for your reply. I add break in my code after ‘MySetTorqueControlType(DIRECTTORQUE);’ The arm did not switch into torque control mode(the arm can not be moved by hand and it still react to joystick command). I also test the GUI interface program “TorqueConsole.exe” in ..\JACO2SDK\GUI. The arm also can not be switched into torque control mode.

    Jin

    #2252

    gurusub90
    Participant

    Hi Martine,

    Response to this:
    Hi Guru
    IN the code you just sent, do you get the joint impedance problem?
    Do you get the same problem if you comment the line MySetCartesianControl?
    Thanks
    Martine

    Sorry for not getting back to you earlier. I missed your earlier message. I’m still getting this issue after commenting out
    result = (*MySetCartesianControl)();
    The only work around is pressing the home button on the joystick.
    Please let me know if you’ve found a work around. I can send you the entire file so you can reproduce the issue on your end.

    Thanks for your help!
    Guru

    #2276

    martine
    Participant

    Dear Guru

    I am really sorry it took me so long to answer you. We got some problem with our login to this forum. I wanted to tell you that I succeeded to reproduce the issue you are experiencing on my end. I also found the source of this error in our control code. This is a bug in our current robots. We will integrate my fix for this problem in the next code release.

    Sorry about that

    Martine

    #2286

    gurusub90
    Participant

    Hi Martine,
    Thanks for looking into it. If you would like I can beta test your fix. I need it for my experiments anyways. Do you have a workaround before the next release?
    Thanks
    Guru

    #2293

    gurusub90
    Participant

    Hi Martine,
    Do you have a fix for the bug that I can use now? This is important since our experiments depend on this.
    Thanks a lot!
    Guru

    #2306

    martine
    Participant

    Dear Guru

    Thank you for your e-mail. Right now, the only way to get around the bug is to give a small joystick input. I could provide you with a beta-version with my fix as I do not think it will be officially released before several months. Could you remind me if you have a Jaco 2 or a Mico arm (you can send me your robot’s serial number). Also, I will need a e-mail address to send you the new firmware.

    Thanks and sorry for the inconvenient

    Martine

    #2307

    gurusub90
    Participant

    Hi Martine,
    Please send it to gsubramani@wisc.edu. Thank you for helping me out with the new firmware.
    Guru

    #2308

    gurusub90
    Participant

    Hi Martine,
    Is the URDF for the Mico and Mico2 the same? Are they the same robot? I would like to know since I am using the URDF files provided by kinova’s ros library and my robot seems slightly off when I use ROS instead of the KinovaUSB api.
    Thank you
    Guru

    #2309

    gurusub90
    Participant

    Hi Martine,
    Sorry to bother you again and again. I’m having difficulty with changing the Inertia and Damping parameters while the robot is in reactive force control mode. When I initiate a change using “SetCartesianInertiaDamping” the robot simply stops moving. Is this normal behavior?
    Thanks again Martine for your continued support!
    Guru

    #2310

    martine
    Participant

    Yes Mico and Mico 2 are the same physical robot. Mico 2 only has some extra functionalities

    As for the SetCartesianInertiaDamping, it works fine on my robot. What are the inertia and damping parameters you tried to put?

    Thanks

    Martine

    #2312

    gurusub90
    Participant

    These are the inertia/damping values:
    Set before force control is active:
    Inertia.X = 5;
    Inertia.Y = 5;
    Inertia.Z = 10000;
    Inertia.ThetaX = 1000;
    Inertia.ThetaY = 1000;
    Inertia.ThetaZ = 1000;
    Damping.X = 60;
    Damping.Y = 60;
    Damping.Z = 10000;
    Damping.ThetaX = 10000;
    Damping.ThetaY = 10000;
    Damping.ThetaZ = 10000;

    They were changed to the following after forcecontrol was turned on:

    Inertia.X = 5;
    Inertia.Y = 5;
    Inertia.Z = 10000;
    Inertia.ThetaX = 1000;
    Inertia.ThetaY = 1000;
    Inertia.ThetaZ = 1000;
    Damping.X = 10000;
    Damping.Y = 60;
    Damping.Z = 10000;
    Damping.ThetaX = 10000;
    Damping.ThetaY = 10000;
    Damping.ThetaZ = 10000;

    The robot freezes at this point. I’m not sure why.
    Thanks
    Guru

    #2417

    cortesao
    Participant

    Hi guys,

    Can we send just a torque command to the robot?
    Say, can we just send a gravity compensation torque tau=gravity(q)? (q represents joint angles). If yes, what happens to the robot when we touch it? Does it move smoothly? Do we feel a lot of friction? Is it back-drivable in this situation?

    I am about to buy a JACO arm, and would like to know how the robot behaves.

    Another question: what is the sampling time for control purposes?

    Cheers,
    Rui

    #2418

    martine
    Participant

    Hi Cortesao

    Our control loop runs at 100 Hz (so sampling time is 10 ms)

    In torque mode, you can send a torque command to the robot. When you do not touch the robot in torque mode, it will stay still because it only compensates its own weight. When you touch it without adding any torque command, the robot will move smoothly in the direction you push it (it feels a little bit like the robot would be in space. You should not feel a lot of friction. The robot is not back-driven in this situation (it is back-driven when the motors are off). You will have the impression that you are moving the robot (as you can do when the robot is off), but really the robot is doing the motion, so the motors are not mechanically back-driven. At all time, the controllers inside the motors are compensating for the motion you are doing.

    I hope that makes it clearer and enjoy your new JACO 🙂

    Martine

    #2419

    cortesao
    Participant

    Hi Martine,

    Thanks for your reply. I am trying to figure out if we can use JACO has a haptic device. 10ms is a rather high sampling time, though. Can’t we make it smaller?
    The movies available online only show JACO has a slave arm. Do you know if there is some demo showing JACO as a master arm?

    Does anyone know the maximum force that the JACO can apply at the end-effector?

    Cheers,
    Rui

Viewing 15 posts - 16 through 30 (of 30 total)

You must be logged in to reply to this topic.

Trending now

News & Events

Kinova’s Vice President of Medical Robotics, Appointed to 2 Boards, 2 Committees

2018 is bound to be a busy year for Kinova’s own Stuart Kozlick. In 2017, Stuart has been appointed to sit on four prestigious committees and boards, effectively positioning Kinova among key industry thought leaders. These strategic partnerships will allow Kinova to accelerate the adoption of new technologies within the healthcare system and encourage alignment within the industry, healthcare communities and governments facilitating accessibilities and communities.

Read more
Humans & Robots

THE NEW NORMAL: UPROOTING ASSISTIVE NORMS in 2018

A new year is a time for new beginnings and starting things afresh. In the spirit of making a change, there’s a specific paradigm that needs to be addressed, or, as many in the industry would say, “disrupted.” Consider 2018 the year of “The New Normal” for the Assistive market. The New Normal is a term coined by our colleague, Abe Clark, in the United States.

Read more
Expert Talk

STEERING TOWARDS CO-CREATION, COLLABORATION, AND INNOVATION – #KINOVAONTHEROAD in 2018

My objective at Kinova is to build bridges between people, businesses, educational organizations, and governments to effectively fuel the cycle of innovation. Unsurprisingly, that means hopping on a plane and going to find the best people in the world to exchange with. This is why the concept of Steering Committees as I came to realize is, and will increasingly be, vital.

Read more