4 Summary

There are at least two approaches to sensing lever position such as found in an e-shifter. Tools like TI's TIMSS can streamline the process evaluating field behavior for a given magnet moved in an arc around a fulcrum point. The key metrics for comparing the die configurations for the two approaches demonstrated in this report are deviation from the ideal angle and difference between angle calculations, which affects redundancy. For the diametric magnet approach, the stacked die consistently has less angle error, smaller differences in angle calculation, and less variation in the observed difference. The starkest contrast in performance was observed for smaller magnet diameters, tilted magnet, and magnets with eccentricity. For the alternative lever approach using an axial magnet, the stacked die is roughly as good and, in some cases, better than the side-by-side die.

The main benefit with the stacked die for an axial magnet swinging past the magnet is flexibility in layout, which might be beneficial for boards that are space constrained with odd geometry. Side-by-side die can see degrees of difference between sensing elements if offset for a particular orientation, while stacked die is consistently well under a degree of difference regardless of orientation.