ADC Noise Shaping for Higher SNR With Reduced Bit Output
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analog to digital converters (ADCs) face a challenge in maintaining a high signal-to-noise ratio (SNR) when using fewer bits than the number of output bits, due to increased quantization noise, which is problematic in scenarios where bandwidth limitations or regulatory restrictions require reducing the number of output bits.
Innovation Solution
A processing block is introduced in the ADC that shifts frequency components of the error signal representing the least significant bits outside a desired band of interest, using a noise-shaping filter to enhance the SNR by reducing noise within that band, thereby improving the representation of input signals with fewer output bits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If fewer bits are used to represent digital codes than the ADC output bits, then bandwidth requirements are reduced, but signal-to-noise ratio deteriorates due to increased quantization noise
Solution Approach 1:
The patent converts the harmful quantization noise into a beneficial form by shaping its spectral distribution. The noise-shaping filter redistributes the quantization noise energy from the band of interest to out-of-band frequencies, transforming the harmful in-band noise into out-of-band noise that can be easily filtered away, thereby improving SNR while maintaining fewer bits for transmission
Solution Approach 2:
The patent changes the spectral distribution parameter of the quantization noise through noise shaping. By applying a noise-shaping filter with specific transfer function characteristics, the noise spectrum is modified to concentrate noise energy at frequencies outside the band of interest, while maintaining low noise levels within the band, thus improving measurement precision without increasing bit depth
2Measurement precision
If quantization noise is reduced by using more output bits, then signal-to-noise ratio improves, but bandwidth consumption increases
Solution Approach 1:
Instead of simply increasing bit depth to reduce quantization noise, the patent applies noise shaping to convert the inherent quantization noise into a spectrally distributed form where most noise energy lies outside the band of interest. This allows achieving high SNR within the band without proportionally increasing the number of bits required for representation and transmission
Solution Approach 2:
The patent applies local quality by differentiating the treatment of different frequency regions. The noise-shaping filter is designed to provide noise attenuation specifically within the band of interest while allowing noise energy to exist outside this band. This localized noise suppression achieves high measurement precision in the relevant frequency range without requiring uniform noise reduction across all frequencies, thereby reducing bandwidth requirements
Data Source
AI summary
Improving signal-to-noise ratio (SNR) when using fewer bits than the number of output bits of an ADC as digital representation of the strength of the samples of an input signal. In an embodiment, an ADC generates digital values of H bits by sampling an input signal at corresponding time instances. An error signal representing the (H-N) least significant bits of the H-bit digital values is processed to determine respective filtered values, which are respectively added to the corresponding ones of the H-bit digital values. The (H-N) bits of the resulting added values are dropped to generate N bit values. The N bit values thus generated may have improved SNR at least in a band of interest, as desired.