Figure 3-6, Figure 3-7, and Figure 3-8 provide insights into how the dual sensing element devices deviate in ideal placement for x or y offset. Figure 3-6 shows the stacked die placed for these tests. However, as both die are centered on the magnet's center point, the device package can be rotated 90° and the impact from x or y offset shall be the same. Figure 3-7 and Figure 3-8 show two different scenarios that can be observed for side-by-side die depending on board layout.

As in the prior analysis involving the diametric magnet, absolute angle error, standard deviation in angle difference, and max deviation in angle difference are all important parameters in determining accuracy and redundancy of the different dual die configurations. Figure 3-9 through Figure 3-11, show the deviation in calculated angle from the ideal angle different x-offsets. Figure 3-12 through Figure 3-15 show the standard deviation in differences between angles calculated from the different die. Figure 3-16 through Figure 3-18 show the max difference in angle calculation between die that is observed for a die configuration. Each set of figures precedes with the stacked die figure and proceed with the side-by-side die figures.

Here, we see that depending on the orientation of the side-by-side die it is worse or equivalent to the stacked die design. When the side-by-side die is aligned along the x-axis, the error observed is only identical at a 0-mm offset, otherwise as one sensing element error decreases, the other sensing element increases as the sensing element moves even further from the ideal sensing location leading to max error that is a full degree higher than the stacked die.

In this case, the variation in error for side-by-side die depends on the orientation with the sensor elements aligned along the x-axis exhibiting the highest variation all greater than half a degree. When the sensing elements for the side-by-side are aligned along the y-axis, then it has slightly better variation than the stacked die.

Not only does the side-by-side die have greater variation when the sensing elements are aligned along the x-axis, it has significantly higher difference between angle calculations, with the max difference being above 1.9°. As for the stacked die, it does have a slightly higher max difference than the other side-by-side configuration; however, the max difference is relatively the same over the swept range.

If the device is offset along the y-axis, the stacked has the lowest worst case error at around 0.6° when the device is furthest from the fulcrum of the rotating magnet.

In this instance the stacked die difference in calculation has a little more variation than the side-by-side. However, the expected variation is relatively the same across sensor y-offset.

Here, we can see that side-by-side for one orientation only has a low max angle difference for a small offset range and outside of that it quickly exceeds the max observed for the stacked die configuration. Additionally, if the horizontal orientation is used, the max difference can be 10x larger than the stacked die.