Installation Tips and Tricks


Transient accelerations

The 3D linear accelerometers in the industrial motion tracker are primarily used to estimate the direction of gravity to obtain a reference for attitude (pitch/roll). During long periods (more than tens of seconds) of transient “free” accelerations (i.e. 2nd derivative of position), the observation of gravity cannot be made. The sensor fusion algorithms can mitigate these effects to a certain extent, but nonetheless, it is impossible to estimate the true vertical without additional information.

 

The impact of transient accelerations can be minimized when you take into account a few things when positioning the device when installing it in the object you want to track/navigate/stabilize or control.

 

If you want to use the industrial motion tracker to measure the dynamics of a moving vehicle, it is best to position the measurement device at a position close to the centre of rotation (CR) of the vehicle/craft. Any rotations around the centre of rotation translate into centripetal accelerations at any point outside the centre of rotation. For a GNSS/INS device with a valid GNSS-fix, the detrimental effect of transient accelerations on orientation estimates is overcome by integrating GNSS measurements in the sensor fusion engine. 

 

Vibrations

The industrial motion tracker samples IMU signals at high frequency per channel, processing them using a strapdown integration algorithm with coning/sculling compensation. Proper coning/sculling compensation already mitigates errors that poorly designed signal processing pipelines introduce when the device is under vibration. For best results, however, it is recommended that the industrial motion tracker be mechanically isolated from vibrations as much as possible: since vibrations are measured directly by the accelerometers, the following two conditions can make the readings from the accelerometers invalid;

  1. The magnitude of the vibration is larger than the measurement range of the accelerometer. This will cause the accelerometer to saturate, which may be observed as a “drift” in the zero-level of the accelerometer. This will show up as an erroneous roll/pitch.
  2. The frequency of the vibration is higher than the bandwidth of the accelerometer. In theory, such vibrations are rejected, but in practice they can still give rise to aliasing, especially if close to the bandwidth limit. This can be observed as a low frequency oscillation. Further, high frequency vibrations often tend to have large acceleration amplitudes (see item 1).

 

There is an effect on the gyroscopes as well and, especially when the vibrations include high-frequent coning motion, the gyroscope readings may become invalid.

 

Xsens has tested a set of vibration dampeners on the industrial motion tracker. Vibration dampeners are low-profile rubber cylinders that allow the industrial motion tracker to be mounted on an object without a direct metal to metal connection that transduces vibrations from the object to the device. The vibration dampeners have been tested with frequencies up to 1200 Hz that caused aliasing when the industrial motion tracker was mounted directly on the vibration table had no effect with the vibration dampeners fitted. The dampeners tested are manufactured by Norelem and have part number 26102-00800855, www.norelem.com

 

Magnetic materials and magnets

When an industrial motion tracker is placed close to or on an object that is either magnetic or contains ferromagnetic materials, the measured magnetic field is distorted (warped) and causes an error in the computed heading. The earth's magnetic field is altered by the presence of ferromagnetic materials, permanent magnets or power lines with strong currents (several amperes) in the vicinity of the device. The distance to the object and the amount of ferromagnetic material determines the magnitude of disturbance introduced. Errors in estimated yaw due to such distortions can be quite large, since the earth's magnetic field is very weak in comparison to the magnitude of the sources of distortion.

 

By default, the AHRS and the GNSS/INS versions (when using the GeneralMag filter profile) stabilize heading using the local Earth's magnetic field. In other words, the measured magnetic field is used as a compass. In addition, the gyroscope biases are continuously estimated by the industrial motion tracker's on-board filter. For the rate of turn around the x-axis and the y-axis (roll and pitch axes), the gyroscope bias is estimated using gravity (i.e. by using the accelerometers). In a homogeneous magnetic field, the gyroscope bias around the z-axis can be successfully estimated as well by monitoring the direction of the magnetic field.

 

The magnetic field can be distorted by the presence of ferromagnetic materials, permanent magnets or power lines with strong currents (several amperes) in the vicinity of the device. The distance to the object and the amount of ferromagnetic material determines the magnitude of disturbance introduced. If the local Earth magnetic field is temporarily disturbed, the on-board filters will initially track this disturbance instead of incorrectly assuming that the device has rotated. However, in case of continuous magnetic disturbances (>10 to 30 s, depending on the filter settings), the computed heading will slowly converge to a new solution using the 'new' local magnetic north. Note that the magnetic field has no direct effect on the inclination estimate.

 

In the special case that the industrial motion tracker is rigidly strapped to an object containing ferromagnetic materials, constant magnetic disturbances will be present. Using a so-called 'magnetic field mapping' (MFM, i.e. a 3D calibration for soft and hard iron effects), these magnetic disturbances can be completely calibrated for, allowing the industrial motion tracker to be used as if it would not be secured to the object containing ferromagnetic materials.

 

For more information please review the Magnetic Calibration Manual.

 

GNSS antenna

Xsens GNSS/INS Development/Starter Kits include a GNSS patch antenna. In contrast to other antenna types such as helix antennas, patch antennas require a ground plane underneath them to operate properly. A ground plane will reduce errors due to multipathing effects, by blocking signals that can normally reach the GNSS antenna from low or sub-horizon elevations.

Adding a ground plane is not necessary when mounting the antenna directly onto a flat metal surface, such as the roof of a car. Otherwise, we recommend mounting the antenna on top of a metal plate (thickness irrelevant) with a minimum diameter of 10 cm.

For best practices or tailor-made ground planes, we recommend contacting the original manufacturer of the used GNSS antenna. Antennas that are sold by Xsens, as well as their product code and original manufacturer, are listed in this BASE article. For Tallysman patch antennas, best practices were discussed here.

In addition to creating a proper ground plane, avoid mounting the antenna underneath or in close proximity of other metal structures and electronics in order to ensure the best GNSS signal reception.

 

For the devices with RTK enabled GNSS receivers, the antenna placement becomes more critical. Because RTK enabled GNSS receivers can measure the position down to centimetre level, the used antenna must be properly fixed with respect to the industrial motion tracker. An accurate measurement of the lever-arms from the sensor's origin of measurements to the antenna’s phase center should be made. Note that an antenna’s phase center is not always the physical center of the antenna. Also, an antenna with low phase center variation should be used. Further details on the GNSS lever arm (i.e. the relative distance of the antenna with respect to the industrial motion tracker) can be found on BASE: The GNSS lever arm (antenna offset) and its role in the GNSS/INS sensor fusion algorithm.

 

A normal (code phase) GNSS receiver needs at least 4 satellites to get a three-dimensional position fix. However, RTK (carrier phase) initialization demands that at least 5 common satellites must be tracked at base and rover sites. Furthermore, carrier phase data must be tracked on the 5 common satellites for successful RTK initialization. Once initialization has been gained, a minimum of 4 continuously tracked satellites must be maintained to produce an RTK solution. Therefore, it is even more critical for an RTK GNSS antenna to have a clear view of the sky to receive data from as many satellites as possible.

 

 

Create your own Knowledge Base