Long Term Measurement of Bandgap Voltage and System Level Integral Non-Linearity Drift

Abstract

All modern precision measurement systems require voltage references that are stable across all variables, namely temperature and time. Most of these references are made through bandgap voltage circuits. While it is impossible to eliminate long term drift in these circuits it is industry standard to collect long term drift data, such that customers can know when their references may drift out of specification. While this is done for voltage references there are complicated system level performance metrics, such as integral non-linearty, that may also drift in time but which we do not collect data for. Most long term drift measurement systems are only equipped to deal with analog voltage measurement and are unable to measure these other system level performance metrics. This thesis details the development of a new long term drift measurement system architecture that can support these system level measurements while also preserving the ability to measure analog data such as voltage reference drift. The architecture was designed with ease of configurability in mind where many types of measurements and product lines can be supported with the necessary development of proper sub-boards.