Class D Amplifier Feedback Switching for Low-Level Noise Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Class D amplifier circuits face limitations in portable devices due to size and power consumption constraints, primarily because of the noise and linearity requirements of analogue-to-digital converters (ADCs) used in feedback paths, which are not optimized for low signal amplitudes.
Innovation Solution
A Class D amplifier circuit that selectively varies its operation between closed-loop and open-loop modes based on the amplitude of the input signal, using a signal selector block and controller to determine when to include an error signal in the modulator input, thereby reducing the need for high-performance ADCs at low signal levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If a continuous-time ADC is used in the feedback path to suppress noise, then noise performance is improved, but the device size and power consumption increase
Solution Approach 1:
The patent applies dynamics by making the feedback loop operational status dynamic rather than static. The feedback path is selectively enabled or disabled based on the instantaneous amplitude of the input signal. When the signal amplitude exceeds a threshold, the feedback loop is activated to suppress noise; when the signal amplitude is below the threshold, the feedback loop is deactivated to reduce power consumption and device complexity. This dynamic adaptation resolves the contradiction between maintaining good noise performance and reducing device size.
2Reliability
If a continuous-time ADC is used in the feedback path to maintain linearity, then linearity is improved, but power consumption increases
Solution Approach 1:
The patent implements periodic action by intermittently activating the feedback loop based on signal amplitude conditions. Rather than continuously operating the ADC and feedback path, the system periodically engages the feedback mechanism only when the input signal amplitude justifies its use. This periodic activation maintains linearity when needed while significantly reducing average power consumption, resolving the contradiction between linearity and power efficiency.
3Reliability
If feedback is always applied, then signal quality is improved, but noise floor increases at low signal levels
Solution Approach 1:
The patent applies local quality by making the feedback application selective rather than uniform across all signal conditions. The feedback loop is locally activated only in the high signal amplitude region where it provides benefit, while being deactivated in the low signal amplitude region where it would degrade performance by raising the noise floor. This localized application of feedback resolves the contradiction between maintaining signal quality and minimizing noise floor.
Data Source
AI summary
This application relates to Class D amplifier circuits. A modulator controls a Class D output stage based on a modulator input signal (Dm) to generate an output signal (Vout) which is representative of an input signal (Din). An error block, which may comprise an ADC, generates an error signal (ε) from the output signal and the input signal. In various embodiments the extent to which the error signal (ε) contributes to the modulator input signal (Dm) is variable based on an indication of the amplitude of the input signal (Din). The error signal may be received at a first input of a signal selector block. The input signal may be received at a second input of the signal selector block. The signal selector block may be operable in first and second modes of operation, wherein in the first mode the modulator input signal is based at least in part on the error signal; and in the second mode the modulator input signal is based on the digital input signal and is independent of the error signal. The error signal can be used to reduce distortion at high signal levels but is not used at low signal levels and so the noise floor at low signal levels does not depend on the component of the error block.