A method for modelling and design of built in self test circuit for data converters

Loading...
Thumbnail Image

Date

item.page.authors

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Integrating highly complex mixed-signal circuits onto a single chip has made testing more tedious and expensive. Data Converters (DCs) are the backbone of mixed-signal circuits, which bridge the gap between the analog and digital worlds. The performance of the mixed-signal circuits is decided by the accuracy of the Data converters. Hence modelling and performance evaluation of data converters are important and inevitable in the design of mixed-signal systems. Machine learning facilitates the development of accurate data converter models that capture the static characteristics of converters, leading to improved performance and efficiency. Machine learning algorithms can analyze large datasets to identify analog and digital patterns, enabling accurate prediction of converter behavior in various conditions. Six regression algorithms in supervised learning have been used to model a 10-bit integrated DAC in the LPC2148 microcontroller and an 8-bit discrete DAC0800 with reference Voltage (Vref) set to 3.3 V and 5 V. Four classification algorithms in supervised learning have been used to model a 10-bit integrated ADC in the ATmega 328 microcontroller with Vref set to 3.3 V and 5 V. To validate this modelling methods, Same six regression algorithms have been used to model an ideal Digital to Analog Converters (DACs) and same four classification algorithms is used to model an ideal Analog to Digital Converters (ADCs) with resolution up to 12-bits and Vref up to 12 V. newline

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced