PDM Digital Filter Using Quantizer Feedback Without Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital filters are inefficient when processing Pulse-Density Modulated (PDM) signals, requiring sample-rate or data format conversions, which incur delays and necessitate high precision coefficients, leading to hardware complexity and potential artefacts like limit cycles.
Innovation Solution
A digital filter design that directly processes PDM signals without conversion, using a filter circuit with quantizer and sigma-delta modulator to attenuate frequency components, where filter variables are updated based on output signals, allowing for efficient hardware implementation as a high pass, low pass, band reject, or band pass filter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If PDM signal is converted to PCM for filtering, then filtering can be performed using conventional digital filters, but delays are incurred and hardware complexity increases
Solution Approach 1:
The patent extracts the quantizer from the conventional digital filter structure and places it at the input stage to directly process PDM signals. This allows the filter to operate on PDM data without requiring conversion to PCM, eliminating conversion delays while maintaining filtering functionality.
Solution Approach 2:
The patent introduces a specialized PDM-compatible filter structure that acts as an intermediary between PDM signal sources and subsequent processing stages. This intermediate filter processes PDM signals directly, avoiding the need for PCM conversion while preparing the signal for downstream applications.
2Adaptability or versatility
If sample-rate conversion is performed on PDM signals, then filtering at different rates is enabled, but hardware complexity and power consumption increase
Solution Approach 1:
The patent implements dynamic filter coefficients that can be adjusted based on the desired frequency response without requiring sample-rate conversion. The filter adapts its characteristics through programmable coefficients, providing versatility while maintaining a fixed hardware structure that avoids complex conversion circuits.
3Measurement precision
If high precision filtering coefficients are used, then filtering accuracy is improved, but hardware area and power consumption increase
Solution Approach 1:
The patent employs fixed-point arithmetic with carefully selected coefficient precision that balances filtering accuracy with hardware efficiency. Rather than using high-precision floating-point coefficients, the design uses optimized fixed-point representations that achieve sufficient accuracy with reduced hardware area and power consumption.
4Ease of manufacture
If PDM to PCM conversion is performed, then conventional filtering can be applied, but additional conversion stages are required
Solution Approach 1:
The patent designs a universal filter structure that can process both PDM and PCM signals, with the quantizer positioned to handle PDM input directly. This multi-functional design eliminates the need for separate conversion stages while maintaining compatibility with conventional filtering approaches when needed.
Data Source
AI summary
A digital filter and a method for filtering a pulse density modulation (PDM) signal are presented. The digital filter has a first filter circuit to receive an input signal with input values at successive time steps to provide a filtered input signal with filtered values at successive time steps. The digital filter does not require sample-rate or data format conversions. Also, the digital filter is area and power efficient when implemented in hardware. Optionally, the digital filter has a sigma-delta modulator including the quantiser, the sigma-delta modulator being used to receive the filtered input signal and to process the filtered input signal before and/or after being quantised by the quantiser. This digital filter does not require sample-rate or data format conversions. This digital filter is area and power efficient when implemented in hardware.