Checking the data plots in Figure 3-12 or Figure 3-14 do show some scattering and uncertainty of absolute position. This variation occurs due to looseness of the mechanical tolerances of the assembly. The sensor is sensitive enough to capture the minor variations in magnetic position, and the control was intentionally wiggled through mechanical tolerances of each position to clearly define the limits for operating states.

Aside from this, input referred noise and sensitivity error are the key factors to consider when determining measurement error. Input referred noise produces random variation on the output signal which influences the final angle calculations. Averaging consecutive measurements can reduce this error source. In effect the RMS noise value is reduced by a ratio of the square root of the number of samples.

Sensitivity error is the result of gain variation in the device signal chain and can result in amplitude variations between axes and between devices. The significance of this error on angle measurements tends to vary based on the mismatch between individual axes.

Additionally, one must consider temperature and lifetime drift of the sensor. This can cause minor changes to offset and sensitivity. The significance of any drift can impact the Alpha and Beta angle calculations and result in some small variance to the defined positions of the lever. Applying temperature compensation internal to the sensor greatly reduces any impact due to temperature variations.

For reference, please consider the following 3D Hall-effect sensors and their data sheets electrical table for a complete summary of performance.