Response Recovery
Algorithm implementation to correct for the inherent delays and finite bandwidth of physical transducers improves the speed and accuracy of real-time measurements. Applying sensor dynamic compensation involves executing an inverse filter on the digitised sensor output to reconstruct the rapid changes in the physical variable. This digital processing extends the effective frequency response of the sensor beyond its physical limits, which is vital in fast-moving control loops.
The process relies on an accurate dynamic model of the transducer to avoid amplifying high-frequency sensor noise.
Filter Design
Inverse transfer functions are discretized and implemented on the microcontroller using digital signal processing structures. The designer must balance the speed of the transient recovery against the noise amplification that occurs when boosting high-frequency signals. Implementing a low-pass filter in series with the compensating filter limits the out-of-band noise while retaining the improved response time.
Model Identification
Estimating the thermal or mechanical time constants of the transducer is accomplished by subjecting the device to step inputs in a controlled test rig. If the identified time constant deviates from the true physical value, the compensation algorithm will either undercorrect or overcorrect the sensor output. This mismatch is minimized by running automated calibration routines at scheduled intervals.
Processing Budget
Execution of complex filtering arithmetic adds computational load and latency to the embedded processor. Low-power sensors often utilize simplified first-order models to run within the real-time constraints of the hardware.