LSM9DS0 Chip Calibration for Antarctic HiCal Balloon Experiment

Names: Margaret Lockwood, Lawrence High School, Lawrence KS

Research Teacher Mentor: James K. Deane, Ottawa High School, Ottawa KS

Research Mentor: Prof. Dave Besson, University of Kansas, Lawrence KS

Purpose:The LSM9DS0 chip calibration project  is part of the much larger Antarctic HiCal Balloon Experiment. This experiment focuses on the detection of cosmic rays to learn more about neutrinos and protons. These particles can be detected by sending showers of radio frequency radiation. The main balloon, ANITA, detects the reflected radio signals created by an air-proton collision. To detect particles successfully and learn more about wave relationships, the  surface roughness of the ice is studied using HiCal. HiCal is a smaller balloon which trails ANITA as it emits kiloVolt scale signals that are measured by both ANITA directly and in the surface reflection. The LSM9DS0 chip will be used to keep track of the orientation of the HiCal balloon relative to ANITA. The chip needs to be calibrated to give accurate and precise data.  

Methods: The chip has a gyroscope, accelerometer and magnetometer, all which need calibration.  The LSM9DS0 chip is connected and run on an Arduino as it outputs data from the sensors. First the hardware was soldered and set up to communicate with the computer. Since the LSM9DS0 data is a stream of numbers, a visualization of the data would aid in calibration and make the data easier to interpret. This was done by connecting Processing to Arduino. After attempting to use a tutorial to display a graphic on processing, it was apparent that something was wrong with the 3D display in the Java environment.  After a lot of troubleshooting, it was decided to move forward without the 3D graphic. Mimicking the original graphic sketch, a 2D sketch of an ellipse was created to change the height, change the width and rotate the ellipse according with the pitch, yaw and roll values. This graphic was helpful but distracted from the calibration goal.

The minima and maxima of the data were found with an Arduino sketch and were used to calibrate the accelerometer and magnetometer. Some issues surfaced; first regarding moving the chip too fast hence causing the accelerometer min/max to be impacted by forces other than gravity. Secondly, the magnetism from the computer affects the magnetometer data. To deal with these, the accelerometer was moved slowly at the approximant minimums and maximums of all three axises. Then using this data, histograms were made to find the absolute peaks. These values are then scaled in the code using the map function. Similarly to the accelerometer, the magnetometer minimums and maximums of all three axes were found in relation to magnetic north and the magnetometer was moved as far from the computer a possible.

To calibrate the gyroscope a turntable with adjustable speeds was used to compare the actual angular velocity with data from the gyroscope. This produced a factor that could modify the gyroscope data to improve accuracy.  The speed of the turntable was found by using a stopwatch and then converting the value to degrees per second. Dividing the measured angular speed by the gyroscope’s indicated angular speed produced the correction factor.

Results:The calibration was moderately successful. Using the stated methods, the data from gyroscope, accelerometer and magnetometer is considerably more accurate . Out of the three, the magnetometer is the least accurate due to the magnetism in the lab from computers. The chip is not completely ready for launch, but the calibration project made progress. The methods to calibrate the chip are accurate and precise.

Meaning to Larger Project: Proper calibration and testing of the LSM9DS0 chip will permit monitoring of the HiCal balloon package orientation and support the analysis of signals generated by HiCal.

Future Research:A few additional things will be needed to completely finish the calibration. The heading of the chip needs to be tested for accuracy and precision. The already completed calibration should be tested over time to check for drift in the sensors as well as recheck the accuracy of the calibration. For launch a housing for the chip,  a power supply, and way to store data will be needed. A battery, for example, could be used instead of  the  computer for power. Connecting an external drive could be used to store data.  Possible things to consider are ways to make sure the chip is still calibrated after launch and figuring out a reference point (like the sun).

We appreciate the assistance and guidance of the following students during this project.

  • Jessica Stockham

  • Steven Prochyra