Pipelined ADC Gain Correction via Digital Error Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pipelined analog-to-digital signal converters face conversion errors due to finite amplifier gain, which are not effectively addressed by existing techniques, as they often introduce noise and fail to maintain accuracy across variations in temperature, fabrication process, and supply voltage.
Innovation Solution
The system employs controllers with gain estimators and back-end adjustors to process digital error signals and control voltages, adjusting digital codes and amplifier gains to enhance conversion accuracy, and uses piece-wise and polynomial approximations to address nonlinear amplifier gains varying with signal amplitude.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing error reduction techniques are applied to address finite amplifier gain, then conversion accuracy is improved, but noise is introduced into the signal processing path
Solution Approach 1:
The patent introduces a digital error signal as an intermediary that carries correction information without affecting the main analog signal path. The digital error signal is generated by comparing the actual output with the expected output and is used to adjust the input signal digitally, thereby avoiding direct manipulation of the analog signal path that would introduce noise.
Solution Approach 2:
The patent replaces analog signal processing techniques with digital signal processing. Instead of using analog circuits to correct amplifier gain errors (which would introduce noise), the system uses digital error signals and digital-to-analog converters to achieve the same correction goal without degrading the signal quality.
2Measurement precision
If existing error reduction techniques are applied to address finite amplifier gain, then conversion accuracy is improved, but accuracy is not maintained across variations in temperature, fabrication process and supply voltage
Solution Approach 1:
The patent implements a feedback mechanism where the digital error signal is continuously generated based on the difference between expected and actual converter outputs. This feedback loop automatically adjusts for variations in temperature, fabrication process, and supply voltage by dynamically correcting the input signal to maintain accurate conversion across all operating conditions.
Solution Approach 2:
The patent makes the error correction dynamic rather than static. The digital error signal and the adjusted digital code are continuously updated based on current operating conditions, allowing the system to adapt to changing parameters such as temperature and supply voltage in real-time, thereby maintaining accuracy across varying conditions.
3Productivity
If pipelined systems break down total bits into cascade of low resolution converters, then high resolution at high sampling speeds is achieved, but conversion errors are introduced due to finite amplifier gain
Solution Approach 1:
The patent divides the high-resolution conversion task into multiple low-resolution converter stages connected in a pipeline, with each stage handling a portion of the total bits. This segmentation enables high sampling speeds by allowing parallel processing across stages while the digital error signal mechanism corrects the cumulative accuracy issues that arise from using multiple finite-gain amplifiers.
Data Source
AI summary
Converter system embodiments are formed with signal-processing stages which include successive signal converters and a preceding signal sampler wherein all but a last one of the stages provides an output signal to a succeeding one of the stages and all of said signal converters generate a corresponding digital code. The system embodiments generally address a selected one of the stages and include controllers which are configured to process, at a process rate less than the system's sample rate, a digital error signal and the back-end digital code of back-end ones of signal converters that succeed the selected stage to thereby adjust at least one of the back-end digital code and a control voltage in the selected stage to enhance the accuracy of the system digital code. Once the processes of these embodiments have been applied to the selected stage, they may be successively applied to preceding stages. System embodiments are also directed to nonlinear amplifier gain by including approximations (e.g., piece-wise and polynomial approximations) of amplifier gain so that they are better directed to an amplifier gain that is appropriate for the signal that is being processed through the amplifier.