MASS post-eject attitude

This is chapter 2 on the MASS post-eject attitude:

Above plot shows the magnetometer data, so the cosines of the angle with respect to the local magnetic field line of the Earth, for 3 orthogonal axes, in still arbitrary scale.

It can be seen that the data quality is good and the noise level small and identical for each of the channels (except some spurious data on Channel 2). At first glance the data seems to confirm the following things:

• The derivative of the data correlates nearly perfectly with the gyro data (see below image)
• Therefore, the MASS/Fotino was in an oscillation, not a rotation (which would have caused severe nonlinearities on the relationship), confirming proper ejection
• The amplitude of oscillation must also be quite small (under 30 degrees) and initially quire perpendicular to the field lines (such that the data resembles a sine wave of small angles and can be interpreted as those angles directly such that the derivate matches angular rate)
• The magnetometer and gyro both show significant linearity within the region of interest
• Maximum amplitude was achieved at 17, 36 and around 50 seconds
• These times correspond perfectly with the suspected zero level (first data point) of the gyroscope data, making the gyro data better understood (apparently the zero level of the gyroscope moved by about 15 points out of 1023 during the 11 days in space)
• Therefore, the first data point represents pre-ejection conditions, such that also: 1. the post-ejection attitude can be understood relative to initial conditions, and 2. the time of ejection [t=0] can be placed accurately within 0.125 s (between first and second data point).
• The initial datapoint of magnetometer 1 is not yet understood, as it seems there is a discontinuity in attitude. Perhaps it is related to the difference in magnetic surroundings after ejection (no more Foton). 
• Indeed during deployment of ripstitching, the oscillation was nearly stopped near maximum angle, at this point there was (coincidentally) little tension, then started again to swing back towards Foton when tension was rising due to turning of the barberpole.
• The second time that the maximum angle in one direction is achieved, the signal is corrupted (data gap), possible this can be correlated to the antenna pattern and position. The data gets lost again a last time once the MASS/Fotino presumably is oriented along the tether.  This has to be better understood
• Apparently there is no spin induced by the ejection (magnetometer 2, TBC), so pure pitch-off
• The pitch-off may be introduced under an angle of about 10 degrees from the plane of magnetometer 1 (since magnetometer 3 amplitude is about 1/6th of that of magnetometer 1) from the magnetometer axes (to be correlated w.r.t. ejection system). This could match nicely with the angle between MBOX and ejection spring. N.B. amplitude of magnetometer 2 is again 4 times smaller. However, all 3 magnetometers seem to follow the same trend, rather than a sine vs. cosine behavior as may have been expected for at least one of the axes. This behavior is not yet understood at this first glance.