Adaptive Capacitive Load Driving With Dynamic Filter Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driving circuits for capacitive loads, particularly in battery-operated devices, inefficiently manage voltage and current, leading to unnecessary filtering and potential artefacts due to worst-case scenario assumptions, which can limit driving capability and introduce signal distortions.
Innovation Solution
An adaptive filter dynamically adjusts its characteristics based on the load capacitance and power source voltage, using a dynamic model to track actual conditions and compute expected currents, allowing efficient driving without introducing artefacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a low pass filter is designed for worst-case scenario (maximum load capacitance, maximum input signal amplitude, maximum frequency, lowest battery level), then protection against artefacts and overcurrent is improved, but excessive filtering occurs during normal operation reducing driving capability
Solution Approach 1:
The filter characteristics are made dynamically adjustable based on real-time monitoring of load capacitance and battery voltage. The system transitions from a static worst-case filter design to a dynamic filter that adapts its cutoff frequency and attenuation characteristics according to actual operating conditions, thereby preventing excessive filtering during normal operation while maintaining protection when needed.
Solution Approach 2:
The invention changes the filter parameters (cutoff frequency, attenuation) based on measured load capacitance and battery voltage values. By continuously monitoring these parameters and adjusting the filter characteristics accordingly, the system avoids the fixed worst-case design that causes excessive filtering, while still providing adequate protection against artefacts and overcurrent conditions.
2Device complexity
If a fixed load model is used based on worst-case assumptions, then design simplicity is improved, but the model does not track load variations leading to excessive filtering
Solution Approach 1:
The system implements feedback by continuously measuring the actual load capacitance and battery voltage, then using these measurements to adjust the filter characteristics. This closed-loop approach replaces the open-loop fixed worst-case model with an adaptive system that responds to actual conditions, improving filtering efficiency without significantly increasing design complexity.
Solution Approach 2:
The system performs self-characterization by automatically measuring its own load capacitance and power supply voltage, then using this self-acquired information to optimize filter settings. This eliminates the need for external characterization equipment and complex pre-programmed models, achieving adaptive filtering with minimal additional complexity.
3Power
If maximum voltage and current are output by the amplifier, then driving power is improved, but artefacts are introduced and protection circuits are triggered
Solution Approach 1:
The system takes preliminary anti-action by pre-calculating the expected current based on measured load capacitance and input signal characteristics, then using this prediction to proactively adjust the filter settings before artefacts or overcurrent conditions occur. This preventive approach allows the amplifier to operate at maximum power while avoiding the triggering of protection circuits and introduction of artefacts.
Data Source
Figure 1A~1C
Figure 2~3A
Figure 4A~5B
AI summary
The present invention relates to a circuit (2000) for driving a capacitive load (L) comprising: an amplifier (1100) for driving the load (L) based on an input signal (VIN), the amplifier comprising at least a boost converter, a dynamic model (2400) configured to track a capacitance of the load (CL) and a voltage of the source (VBAT) for powering at least parts of the circuit (2000), an adaptive filter (2300), configured to filter the input signal (VIN) based on an output of the dynamic model (2400).