Uncertainty-Guided Online Closed-Loop Linearity Testing of Analog-to-Digital Converters (el. knyga) (skaityta knyga) | knygos.lt

Aprašymas

The characterization of static linearity in Analog-to-Digital Converters (ADCs) is an essential part of production testing but requires substantial measurement effort for high-resolution converters. This dissertation introduces a methodology that integrates measurement, estimation, and stimulus selection into a sequential, uncertainty-guided identification process. A parameterized model of the ADC transfer characteristic is recursively updated during the test. The current estimate is then used to adaptively select subsequent measurement points based on their expected information contribution. The model state is further exploited to accelerate transition-level localization. By performing estimation concurrently with data acquisition, the approach shifts the focus from minimizing the number of measurements toward optimizing the interaction between measurement and computation. The methodology supports different model representations and is suited to converter architectures with structured nonlinearity behavior, such as SAR ADCs. Simulation-based evaluations using realistic ADC models demonstrate accurate reconstruction of static nonlinearity characteristics while improving overall test efficiency.