Evaluating Thermal Hysteresis Limits in MEMS Accelerometer Static Leveling
MEMS thermal hysteresis creates non-repeatable zero-g offset shifts up to 3 mg, driving uncompensable arcsecond tilt errors in static leveling systems.
Bias Repeatability
Baseline Stability
Metrological stability is the baseline metric against which successive zero-point measurements are evaluated under identical operating conditions. For many inertial and pressure sensors, bias repeatability determines the long-term reliability of the device by defining how consistently the output returns to its null state after cycling. This characteristic is distinct from bias stability, which tracks drift over a continuous duration rather than across discrete power cycles or environmental transitions.
The parameter is typically expressed in units of the measured physical quantity, such as micro-g for accelerometers.
Measurement Protocol
Verification of this metric requires a structured testing sequence where the sensor is powered down, subjected to a thermal or mechanical transition, and then powered up again to capture the zero-offset value. Technicians repeat this cycle multiple times to calculate the standard deviation of the resulting bias values. A low variance indicates high stability, allowing the system computer to apply a constant correction factor with confidence.
When the repeatability is poor, the control system must perform frequent auto-zero routines to prevent error accumulation.
Environmental Drift
Temperature fluctuations represent the primary source of drift, as thermal gradients within the sensor package induce physical stresses that alter the electrical null point. Protective shielding and thermal insulation help minimize these gradients, although high-performance applications often require active temperature control. Manufacturers specify the valid thermal range for the declared repeatability value on the calibration sheet.
Calibration Routine
Routine calibration procedures must distinguish between a constant bias offset and the random variations characterized by repeatability. The constant offset is easily subtracted in software, whereas the random repeatability limits the ultimate resolution of the instrument. High-precision systems use high-stability reference standards during production to characterize these limits before the sensor is deployed in the field.