Accurate and robust spectral testing with relaxed instrumentation requirements

Abstract

Analog-to-Digital Converters (ADCs) are becoming increasingly common to be involved in most systems in Integrated Circuits (ICs). Thanks to the rapid growth of modern semiconductor technology, the performance of the data converters becomes better and better. One of the difficulties being faced is to be able to accurately and cost-effectively test the continually better performance ADCs. The conventional test method for ADCs can be difficult to implement accurately and cost effectively due to the stringent requirements, such as achieving coherent sampling, requiring high purity test stimulus, avoiding signal clipping, maintaining stationary test environment and minimizing clock jitter. To relax these necessary conditions and to reduce test cost while achieving high accuracy, several new algorithms are introduced in this paper, to perform accurate spectral and linearity test of the ADC without requiring precise test instruments. Extensive simulation results performed in MATLAB first validated the accuracy as well as robustness of the proposed algorithms. Furthermore, measurement results using different devices and test platforms have verified many of these algorithms. For the first time, the accurate testing can be performed without the need for stringent requirements of conventional test, which offers much more flexibility as well as lower cost. Combined with their easy, cost-effective setup and high accuracy, these algorithms are readily available for precision ADC characterization and can be implemented into Built-in-Self-Test (BIST). solutions for production and bench testing.