Geomagnetic Motion Tracking Products

June 1, 2013

Figure 1. Calibration Coverage at Equator with Standard Calibration Program.

Any compass incorporated into a host system requires calibration so that the compass properly compensates for ferrous and magnetic components. If this is not done correctly, heading readings can be very inaccurate.

Figure 2. Calibration Coverage at High Latitude with Standard Calibration Pattern.

Figure 3. Calibration Coverage at High Latitude with Modified Calibration Pattern.

PNI Sensor Corp. recently assisted a customer operating in northern Scandinavia, where the high latitude corresponds to the magnetic field having a large dip angle, or inclination. Magnetic calibration entails taking calibration sample points at various relative headings and attitudes. A total of twelve sample points were taken at various relative headings and attitudes. The 12 points were well distributed, ensuring that the calibration sphere can be modeled accurately. Conceptually, the x, y, and z magnetic field readings for the sample points were plotted, and by extrapolation a “calibration sphere” could be constructed. Figure 1 shows such a sphere when working at the magnetic equator, with a dip angle of 0°. As can be seen, the 12 points are well distributed, helping to ensure the calibration sphere can be modeled accurately. Using PNI’s standard calibration routine and working at typical latitudes, 0.3° rms heading accuracy can be achieved with PCI’s TCM compasses.

Figure 2 shows the resulting sample points for PNI’s standard calibration pattern when working in northern Scandinavia, where the dip angle was 76°. Due to the high dip angle, the sample points are concentrated in one area and it is difficult to accurately extrapolate the calibration sphere. The result is a less accurate heading reading.

The essential problem is to improve sample point coverage for generating the calibration sphere. PNI subsequently performed MatLab modeling and devised a number of improved calibration patterns. After evaluating the options and the customer’s constraints, PNI recommended increasing the roll during calibration, and following a different sequence. As can be seen in Figure 3, the effect is much better coverage.

The ultimate result of using the highroll calibration pattern at high latitude was the ability to retain a heading accuracy of ~0.3° rms, thus meeting the customer’s expectations and requirements.