Parallel Sigma-Delta Channels for Quantization Noise Decorrelation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multichannel transmit and receive systems face challenges in decorrelating quantization noise without impacting signal quality or stability, particularly when high bit depth DACs are used, leading to increased complexity and power consumption.
Innovation Solution
Implement a multichannel system with sigma-delta modulators having distinct transfer functions for each channel, decorrelating quantization noise by modifying the second term of the transfer function without altering the first term, using techniques such as hysteresis quantizers, multibit quantizers, and cascaded sigma-delta modulators with channel-specific adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a high bit depth DAC is used to reduce quantization noise, then the signal-to-noise ratio is improved, but the device complexity and power consumption increase substantially
Solution Approach 1:
The patent divides the single high-resolution DAC into multiple parallel low-resolution DACs (one per channel). Each DAC operates at reduced bit depth, but the collective output of all channels provides the necessary resolution through spatial segmentation. This allows the system to achieve high effective resolution without requiring any single DAC to have high bit depth, thereby reducing individual DAC complexity and power consumption.
Solution Approach 2:
The patent combines the outputs of multiple parallel processing channels, each with its own low-resolution DAC, to achieve the equivalent performance of a single high-resolution DAC. By merging the quantization noise characteristics across channels and applying decorrelation techniques, the system achieves high signal-to-noise ratio while using multiple simple DACs instead of one complex high-bit-depth DAC.
2Device complexity
If the same digital signal is input into all channels to simplify routing, then the device complexity is reduced, but the quantization noise adds coherently and its power increases as 20 log(N)
Solution Approach 1:
The patent applies local quality by making each channel's quantization noise characteristic unique through different noise transfer functions. While all channels receive the same input signal for routing simplicity, each channel processes it differently through distinct sigma-delta modulator configurations, creating locally differentiated noise characteristics that prevent coherent addition and reduce overall noise power.
Solution Approach 2:
The patent introduces asymmetry in the noise transfer functions across parallel channels. By configuring each channel's sigma-delta modulator with different parameters (such as different quantizer thresholds or feedback coefficients), the quantization noise in each channel becomes asymmetric and uncorrelated with others, preventing coherent addition and reducing the noise power increase from N channels.
3Object-generated harmful factors
If different dither is injected on each channel to reduce quantization noise correlation, then the quantization noise correlation is reduced to 10 log(N), but the device complexity increases due to N uncorrelated noise sources and additional processing
Solution Approach 1:
The patent achieves channel-specific noise decorrelation using a universal sigma-delta modulator architecture that can be instantiated N times in parallel. The same basic modulator design is used across all channels, but with configurable parameters that can be adjusted per channel to create different noise transfer functions. This universal design approach reduces complexity compared to implementing completely separate decorrelation mechanisms for each channel.
Solution Approach 2:
The patent decorrelates quantization noise by changing parameters within the sigma-delta modulator for each channel, such as quantizer threshold values, feedback coefficients, or integrator gains. By modifying these parameters across channels, each channel produces different noise transfer characteristics without requiring additional noise injection hardware or complex processing, thereby achieving decorrelation with minimal complexity increase.
4Device complexity
If a sigma-delta modulator is used to reduce DAC bit depth, then the device complexity and power consumption are reduced, but the sampling frequency must be increased to maintain good over-sampling factor
Solution Approach 1:
The patent employs dynamic sigma-delta modulation techniques where the modulator adapts its operation to maintain effective over-sampling. The system dynamically adjusts the modulation index or feedback parameters based on the input signal characteristics, allowing the use of lower fixed sampling frequencies while still achieving the noise shaping and filtering benefits typically associated with high over-sampling ratios. This dynamic adaptation reduces the required sampling frequency compared to static approaches.
Data Source
AI summary
A multichannel transmit and/or receive system, each channel includes a DAC and a sigma-delta modulator the transfer function of which is expressed thus:OUT(z)=IN(z)·FTS(z)+Q(z)·FTB(z),where OUT is the output signal of the sigma-delta modulator, IN is the input signal of the sigma-delta modulator, FTS is the transfer function of the input signal, Q is the quantization noise and FTB is the transfer function of the quantization noise, the second terms of the transfer function of the sigma-delta modulator only being distinct from one another for two channels Vi, Vj, in order to decorrelate the quantization noise of distinct channels, the first term of said transfer function for channel Vi being equal to the first term of said transfer function for channel Vj.