Lookup-Table PWM Circuit for Low-Distortion Class D Amplifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Class D amplifiers face limitations in achieving high fidelity and efficiency due to distortion issues in PWM modulation, particularly in reaching 0 dBFS with controlled frequency, and require complex pre-processing algorithms, while sigma-delta modulators offer noise shaping but not fixed frequency outputs, and multi-bit signals are unsuitable for class D amplifiers.
Innovation Solution
A class D amplifier incorporating a noise-shaping modulator, pulse width modulator, and pulse amplifier with a lookup table for PCM to PWM conversion, along with a compensation circuit to optimize performance, which reduces total harmonic distortion and power consumption, and includes an optional demodulator filter for analog signal conversion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If direct PCM to PWM conversion is used, then the amplifier achieves high efficiency and simplicity, but distortion terms appear in the baseband signal
Solution Approach 1:
The patent applies preliminary action by implementing pre-processing of the PCM signal before PWM modulation. The noise-shaping modulator and lookup table perform preprocessing operations that shape the quantization noise and prepare the signal in advance, preventing distortion from appearing in the audible baseband while maintaining the efficiency benefits of direct PCM-to-PWM conversion.
2Manufacturing precision
If sigma-delta modulators with noise shaping are used, then signal fidelity improves, but the output frequency is not fixed
Solution Approach 1:
The patent segments the modulation function into two distinct parts: a noise-shaping modulator that improves signal fidelity, and a separate lookup table that converts the modulator output to fixed-frequency PWM signals. This segmentation allows each component to specialize in one function, achieving both high fidelity and frequency stability.
Solution Approach 2:
The lookup table acts as an intermediary between the noise-shaping modulator and the final PWM output. It takes the variable-frequency output of the modulator and transforms it into fixed-frequency PWM signals, mediating between the conflicting requirements of noise shaping and frequency stability.
3Manufacturing precision
If complex pre-processing algorithms are used, then distortion is reduced, but device complexity increases
Solution Approach 1:
The patent uses a lookup table that stores pre-computed PWM patterns corresponding to different input signal levels. Instead of performing complex real-time calculations, the system copies pre-prepared pulse patterns from the lookup table based on the input signal amplitude, significantly reducing computational complexity while maintaining high signal fidelity.
Solution Approach 2:
The complex processing is performed in advance during the creation of the lookup table, not during real-time operation. The table contains pre-calculated optimal PWM sequences that minimize distortion, allowing the runtime system to simply retrieve and apply these pre-optimized patterns without performing complex calculations.
4Measurement precision
If multi-bit PWM signals are used, then signal resolution improves, but compatibility with class D amplifiers is lost
Solution Approach 1:
The patent segments the high-resolution information across multiple sequential PWM cycles rather than attempting to encode it in a single multi-bit signal. The lookup table stores multi-cycle pulse patterns that convey fine resolution information through temporal sequencing, maintaining compatibility with standard class D amplifiers that expect conventional PWM inputs.
Solution Approach 2:
The patent transitions from encoding resolution in the amplitude dimension (multi-bit signals) to encoding it in the temporal dimension (multi-cycle pulse patterns). By distributing resolution information across time rather than using vertical multi-bit encoding, the system maintains compatibility with standard class D amplifiers while achieving high signal resolution.
Data Source
AI summary
A class D amplifier includes a noise-shaping modulator, a pulse width modulator, and a pulse amplifier. The noise-shaping modulator receive a pulse code modulated (PCM) signal and produces an oversampled PCM signal. The pulse width modulator produce a pulse width modulated (PWM) signal from the oversampled PCM signal. The pulse amplifier amplifies the PWM signal to produce an amplified PWM signal. The PWM uses a lookup table to convert from PCM to PWM. A compensation circuit optimizes amplifier performance. An optional demodulator filter converts the amplified PWM signal to an analog signal. The amplifier is ideal for integrated audio applications.