I designed a sensor target combination similar to the TIDA-00508 proposal.
The accuracy at room temperature at a fixed z-axis distance of the target (10% of sensor diameter) is better that 1°.
But with temperature variation over the range 0 to 80°C I observe a
linear offset drift of the differential signals A and B with a rotating target.
The effect on the amplitude is neglible if the offset does not change.
But the measured offset could rise above 6% of the amplitude of A and B over temperature. This increases the position error.
I did some static measurements (target not rotating) and fixed z-axis distance of 10% of coil diameter:
Is the target shape aligned that way, that a sensor pair, for example A+ and A- is covered with the same width of the conducting material from the target shape,
I measure no drift.
Is the target shape aligned that way, that sensor A+ is covered with the maximum width of the conducting material from the target shape,
and sensor A- with the minimum width, then the drift is maximum.
What could cause this effect, do I need an additional temperature compensation?