Isolating Micro-Machined Accelerometer Zero-G Bias Shifts under Static Thermal Cycles
Static thermal zero-g bias shifts in MEMS accelerometers are isolated by enforcing prolonged thermal dwell periods to decouple stress creep from thermal gradients.
B-Spline Thermal Interpolation
Numerical Estimation
Multi-point temperature compensation in precision sensors relies on mathematical curves to estimate correction factors across a wide operating range. This approach, known as B-spline thermal interpolation, uses piecewise polynomial functions to construct a smooth curve that passes through or near calibration points. The method ensures that the first and second derivatives of the curve are continuous, preventing abrupt changes in the correction value.
This continuity is essential for preventing artificial noise in the compensated output of the sensor.
Algorithm Structure
The calculation requires a set of control points and a knot vector that defines the intervals over which each polynomial segment is valid. Unlike global polynomial fitting, which can exhibit large oscillations between data points, this local method restricts the influence of each calibration point to its immediate neighborhood. This localized behavior means that adjusting a single calibration point only recalculates the adjacent segments, leaving the rest of the correction curve unchanged.
This modular structure reduces the computational load on the embedded microprocessor during the calibration process.
Sensor Calibration
Factory calibration procedures utilize this interpolation technique to program the non-volatile memory of digital pressure sensors and accelerometers. The sensor is exposed to a series of known temperatures, and the output error is recorded at each step. These measured errors define the control points for the spline algorithm, which are then stored in the sensor memory.
This storage strategy allows the microprocessor to compute precise correction factors in real time for any measured temperature.
Precision Boundary
Interpolation accuracy depends on the spacing of the calibration points and the stability of the reference sensor. If the calibration points are too far apart, the spline may fail to capture highly non-linear temperature effects, resulting in interpolation errors. This limitation is addressed by placing more calibration points in regions where the sensor output exhibits the highest thermal sensitivity.