Asynchronous Sample Rate Conversion via Analog Filter Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional sample rate conversion techniques require large amounts of memory to store filter coefficients, making them impractical for many applications, and introduce errors due to the need for high time-resolution and precise delay calculations.
Innovation Solution
Implementing sample-rate conversion using a discrete-time digital simulation of a continuous-time analog filter, which allows for high time resolution without the need for extensive memory storage, by simulating complex poles and using a composite tenth-order filter with a steep transition between pass-band and stop-band, reducing memory requirements and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If FIR filters are used for sample rate conversion, then filtering accuracy is improved, but memory requirements increase significantly
Solution Approach 1:
The patent changes the fundamental parameter of filter implementation from discrete-time FIR filters to continuous-time analog filter simulation. This parameter change allows achieving high filtering accuracy through analog continuous processing while avoiding the need to store large numbers of digital filter coefficients, thus resolving the contradiction between filtering accuracy and memory requirements
Solution Approach 2:
The patent substitutes the mechanical/digital FIR filter system with an analog filter simulation system. By using analog filter mathematics and continuous-time processing, the system achieves accurate filtering without requiring extensive digital memory storage for coefficients, thereby resolving the contradiction
2Manufacturing precision
If high time-resolution is achieved through multiple coefficient sets, then conversion accuracy is improved, but memory storage requirements increase
Solution Approach 1:
The patent changes the approach to time-resolution from using multiple discrete coefficient sets to using continuous-time analog filter mathematics. This allows achieving high conversion accuracy through continuous parameter adjustment in the analog domain without storing multiple sets of coefficients, thus resolving the contradiction between conversion accuracy and memory storage requirements
Solution Approach 2:
The patent uses mathematical models and simulations of analog filter behavior to replicate the functionality of high-resolution filtering without physically storing the corresponding digital coefficient data. This mathematical copying approach achieves accuracy without the memory burden of storing actual coefficient sets
3Adaptability or versatility
If asynchronous sample rate conversion is implemented, then adaptability to varying sample rates is improved, but system complexity increases
Solution Approach 1:
The patent implements a universal analog filter simulation framework that can handle various sample rate conversion scenarios (both rational and asynchronous) through a single continuous-time mathematical model. This multi-functional approach provides adaptability to varying sample rates without requiring separate complex systems for different conversion types, thus resolving the contradiction between adaptability and system complexity
Data Source
AI summary
The sample rate of a digital signal is converted by a digital simulation of an analog filter. The simulation can update the states of complex poles in the analog filter at arbitrary times using different techniques. One technique updates the states at variable rates. Other techniques update the states at a fixed rate in response to values of input or output samples that are modified to account for offsets between the times of the samples and the times the states are updated. The states may be updated by using interpolations of complex exponential functions. Values of the complex exponential functions may be obtained from a product of values obtained from multiple lookup tables.