Drotek RM3100 sensor performance test

August 6, 2019 5:23 pm

RM3100 Geomagnetic sensor is used to measure the Earth’s magnetic field to provide absolute reference and heading. But a large challenge is the changing magnetic fields that temporarily distort the heading information. Compensating for these and other transient magnetic anomalies requires the geomagnetic sensor to be able to precisely distinguish between sensor noise and real changes in magnetic field.

Generally, drones avoid using internal magnetometers because they are easily disturbed, their calibration yields high offsets that tend to drift over time.  A well-positioned external magnetometer as RM3100 can help you to get rid of magnetic disturbances due to motor coils and metal parts.

 

What is RM3100 Geomagnetic Sensor?

When it comes to heading measurements and accurate motion tracking, we base our calculations on Earth’s magnetic field. In real conditions, numerous surrounding fields can distort it such as system components in furniture (batteries, shielding materials, motors), passing vehicles, or nearby cell phones and computers. That is why the use of a geomagnetic sensor become essential.

RM3100

Drotek Professional Grade Magnetometer RM3100 eliminates any noises in your magnetic field measurements, making distortion error correction a breeze, allowing you to easily and accurately calculate absolute direction and heading. This new generation of geomagnetic sensor provides high resolution, low power consumption, large signal noise immunity, a large dynamic range, and high sampling rates. Measurements are stable over temperature and inherently free from offset drift. Other technologies require expensive and complex signal processing to obtain similar resolution and noise.

RM3100 VS LIS3MDL

The goal of this test is to compare the performances of two magnetometers used in the drone industry: ST LIS3MDL and PNI RM3100. On an autonomous vehicle, the magnetometer is used to measure an heading angle relative to the magnetic North. This absolute angle is used to improve the estimation of the navigation filter.

 

Test setup

We fixed the LIS3MDL and the RM3100 sensor, on a wooden plate, close one to another and facing in the same direction. The two magnetometers are connected to a Drotek Pixhawk 3 Pro via I2C. The firmware used in the Pixhawk 3 Pro is PX4 v1.9.0.

This wooden plate is fixed to a four-wheeled rover (the Dagu Wild Thumper from Pololu). The two sensors were calibrated on the rover, with QGroundControl. The rover and the sensors are placed outside, in the center of a field to avoid external magnetic perturbations. The rover is elevated (wheels off the ground). Using the compass from an iPhone, the rover is placed facing the magnetic North.

rm3100 sensor magnetic field performance test

 

During this test, we measured the heading angle from the two magnetometers with the motors turned off and with the motors at full throttle.

 

The following figure represents the measurements of the two sensors.

RM3100 sensor performance test log59 alcal

On the first part of this graph the rover motors are turned off, then on the second part of this graph the motors are at full throttle.

The following figure represents a zoom on the measurements with the motors turned off.

RM3100 sensor performance test log59 alcal without motors

For these measurements, we computed the mean values and the standard deviation for the two magnetometers.
RM3100 tabs sensor

We can see that our alignment to the North with a smartphone is not perfect, because the magnetometers mean values are around -5°. The absolute heading error of the two sensors is not computed in this test, because a more precise magnetic compass is required to compare its measurement with the measurements of the LIS3MDL and the RM3100. The RM3100 heading angle is more stable and less noisy than LIS3MDL angle. The RM3100 is less noisy than the LIS3MDL by a factor of 7.5 .

The following figure represents a zoom on the measurements with motors at full throttle.

RM3100 performance test sensor log59 alcal with motors

For these measurements, we computed the mean values and the standard deviation for the two magnetometers.

RM3100 sensor perfo test tabs 2
The RM3100 heading angle is still more stable and less noisy than LIS3MDL angle. The RM3100 is less noisy than the LIS3MDL by a factor of 4.8. With the motors at full throttle, the heading angle from the two sensors are noisier. The brushed motors create magnetic perturbations on the magnetometers measurements.

Get all documentation for RM3100 professional grade magnetometer here.