Dithered Waveform Averaging for ADC Nonlinearity Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques fail to effectively mitigate correlated noise and distortion in the digitization of repetitive waveforms, particularly due to analog-to-digital converter (ADC) nonlinearity, which affects the characterization of repetitive waveforms in applications like optical pulse shaping and telecommunications.
Innovation Solution
The use of continuous dithered waveform averaging, where a time-varying dither signal is applied to the analog repetitive signal to decorrelate distortions, combined with a synchronized averaging process to produce a digital averaged signal, reducing correlated noise and improving fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional waveform averaging is used to reduce noise, then uncorrelated noise is reduced, but correlated noise and distortion from ADC nonlinearity cannot be mitigated
Solution Approach 1:
The patent applies dynamic dithering by continuously varying the offset voltage during the averaging process. Instead of using static offsets, a time-varying dither signal is added to the input signal before ADC conversion. This dynamic approach ensures that correlated distortions from ADC nonlinearity are randomized across multiple waveform captures, allowing them to be averaged out along with noise, thereby improving both measurement precision and reliability
Solution Approach 2:
The patent changes the parameter of the offset voltage from static to time-varying. By modulating the dither signal amplitude and frequency, the system transforms the ADC input signal in a controlled manner that decorrelates distortion products. This parameter change enables the system to maintain high measurement precision while effectively mitigating correlated noise that would otherwise persist through conventional averaging
2Reliability
If dithering is applied to decorrelate distortions, then correlated noise is reduced, but system complexity increases
Solution Approach 1:
The patent introduces a dither signal as an intermediary element between the input signal and the ADC. This intermediary signal serves as a mediator that randomizes the quantization error without directly affecting the underlying waveform information. The dither acts as a controlled disturbance that enables distortion decorrelation while maintaining signal integrity, reducing reliability concerns without requiring fundamental system redesign
Solution Approach 2:
The patent generates the dither signal as a separate, independent copy that is added to the input signal path. Rather than modifying the ADC or input signal source, a duplicate dither waveform is created and combined with the original signal. This copying approach isolates the complexity to a dedicated dither generation module, making the system easier to implement and maintain while achieving correlated noise reduction
3Measurement precision
If extensive averaging is performed to improve signal-to-noise ratio, then measurement accuracy improves, but measurement time increases
Solution Approach 1:
The patent applies preliminary dithering to the signal before the averaging process begins. By pre-modulating the signal with a dither signal, the system prepares the input in a way that reduces correlated distortions upfront. This preliminary action ensures that subsequent averaging converges faster to the true waveform, improving signal-to-noise ratio more quickly and reducing the total measurement time required
Solution Approach 2:
The patent employs periodic dithering where the dither signal is applied in synchronized cycles with the input waveform. This periodic modulation creates a regular pattern of distortion randomization that accelerates the convergence of the averaging process. By aligning the dither period with the signal repetition rate, the system achieves high signal-to-noise ratio with fewer average cycles, thereby reducing measurement time
Data Source
AI summary
Methods and devices for digitizing an analog repetitive signal using waveform averaging are described. An example method includes generating a time-varying dither signal, receiving the analog repetitive signal comprising multiple instances of a waveform, wherein each waveform has a waveform duration, wherein an average of the time-varying dither signal over multiple waveform durations is substantially zero, and wherein the time-varying dither signal varies over each waveform duration, generating a timing alignment, combining each waveform with the corresponding portion of the time-varying dither signal over each waveform duration to produce an analog output signal, converting the analog output signal to a digital output signal, and producing, based on the timing alignment, a digital averaged signal based on averaging the multiple instances of the waveform in the analog output signal, wherein the timing alignment is used to align the multiple instances of the waveform in the analog output signal.