Loading ...
Sorry, an error occurred while loading the content.

Re: [SeattleRobotics] Re: Gyro ADXRS300 - does it need to be in thecenter?

Expand Messages
  • Peter Harrison
    ... All parts of a rigid body rotate at the same rate - wheels or no wheels. Gyros can be mounted anywhere. And thank goodness they do. How in the world are
    Message 1 of 11 , Apr 5, 2008
    • 0 Attachment
      dunkerjeff wrote:
      > The answer is yes and no. Ideally you should be at the center of mass
      > or the rotation axis not always the same as is the case in a wheeled
      > vehicle but is for a rocket. If you are not at this point then
      > translational/rotational compensations is needed to make data appear
      > as it the unit was. You have one of two choices, live with the error
      > caused by being off axis or compensate it the processor. The
      > compensation algorithm can be a resource hog depending on how many
      > error sources you are trying to compensate for. This is less on a
      > concern when speeds are low and with fewer number of degrees of
      > freedom (DOF). The 6 DOF inertial measurement units (IMU) used in
      > aerospace are typically not at the center of mass of the rocket and
      > require the translational/rotational compensation to make it look like
      > it was once this done the the compensation for the various error
      > sources are address. The gyro's and accelerometer's require different
      > compensations. A gyro that has it axis in line or parallel to the
      > axis of rotation will only require translation compensation. If it is
      > not then it also would need a rotational compensation.


      All parts of a rigid body rotate at the same rate - wheels or no wheels.
      Gyros can be mounted anywhere. And thank goodness they do. How in the
      world are you going to predict where the centre of rotation will be
      except for the simplest cases?

      Accelerometers are a whole different game.

      Pete Harrison
    • fred miller
      havent tested the gyro ADXRS300 ,but have tested others on my robots i have made ,and found out that if not in center ,will give a small error or more
      Message 2 of 11 , Apr 5, 2008
      • 0 Attachment
        havent tested the gyro ADXRS300 ,but have tested others on my robots i have made ,and found out that if not in center ,will give a small error or more depending how far off from the center,but can be corrected in software

        ---------------------------------
        You rock. That's why Blockbuster's offering you one month of Blockbuster Total Access, No Cost.

        [Non-text portions of this message have been removed]
      • Robert F. Scheer
        ... Could you post how you do the compensation and maybe what the physics behind needing to do the compensation would be? - Robert
        Message 3 of 11 , Apr 5, 2008
        • 0 Attachment
          On Sat, 2008-04-05 at 15:38 -0700, fred miller wrote:
          > havent tested the gyro ADXRS300 ,but have tested others on my robots i have made ,and found out that if not in center ,will give a small error or more depending how far off from the center,but can be corrected in software

          Could you post how you do the compensation and maybe what the physics
          behind needing to do the compensation would be?

          - Robert
        • dunkerjeff
          ... wheels. ... You are correct for a rigid body and a perfect gyro. But unfortunately gyro s are also sensitive to acceleration. When off axis they will
          Message 4 of 11 , Apr 7, 2008
          • 0 Attachment
            --- In SeattleRobotics@yahoogroups.com, Peter Harrison
            <peter.harrison@...> wrote:
            >
            > dunkerjeff wrote:
            > > The answer is yes and no. Ideally you should be at the center of mass
            > > or the rotation axis not always the same as is the case in a wheeled
            > > vehicle but is for a rocket. If you are not at this point then
            > > translational/rotational compensations is needed to make data appear
            > > as it the unit was. You have one of two choices, live with the error
            > > caused by being off axis or compensate it the processor. The
            > > compensation algorithm can be a resource hog depending on how many
            > > error sources you are trying to compensate for. This is less on a
            > > concern when speeds are low and with fewer number of degrees of
            > > freedom (DOF). The 6 DOF inertial measurement units (IMU) used in
            > > aerospace are typically not at the center of mass of the rocket and
            > > require the translational/rotational compensation to make it look like
            > > it was once this done the the compensation for the various error
            > > sources are address. The gyro's and accelerometer's require different
            > > compensations. A gyro that has it axis in line or parallel to the
            > > axis of rotation will only require translation compensation. If it is
            > > not then it also would need a rotational compensation.
            >
            >
            > All parts of a rigid body rotate at the same rate - wheels or no
            wheels.
            > Gyros can be mounted anywhere. And thank goodness they do. How in the
            > world are you going to predict where the centre of rotation will be
            > except for the simplest cases?
            >
            > Accelerometers are a whole different game.
            >
            > Pete Harrison
            >
            You are correct for a rigid body and a perfect gyro. But unfortunately
            gyro's are also sensitive to acceleration. When off axis they will
            experience a centrifugal force. Depending on the sensor design method
            of the gryo some will be more sensitive than others. The rotation
            axis can be calculated but is probably easier to measure, but getting
            it right on may not even be possible due to other structures. The best
            thing is that the sensitivity to this error source should be small and
            the turn rate low so the error can probably be ignored. But it is
            there and is still an accuracy determination for the robot designer.

            Jeff
          Your message has been successfully submitted and would be delivered to recipients shortly.