No abstract

No abstract

No abstract

h SPECTRAL ANALYSIS IS one of the most widely used methods in several areas such as geophysics, oceanography, medical sciences, etc. It provides information about different frequency components present in a signal that are otherwise not observed in time domain. In electronics, it is extensively used to test dynamic parameters of a wide range of semiconductor circuits and systems, such as analog-to-digital converters (ADCs).ADCs are one of the most widely used mixedsignal circuits. Sufficient spectral performance of ADCs is critical for high-speed and high-resolution applications such as communications [1], [2]. To ensure high accuracy in spectral test, the IEEE standards [3], [4] recommend the test setup to satisfy a list of stringent requirements.As ADC resolution and speed become high, the IEEE requirements have become challenging. Furthermore, satisfying these requirements inevitably increases test time and test cost. Relaxing these requirements not only decreases test time/cost but also facilitates on-chip test capability. Several methods have been proposed in the past that relax one or more of these requirements. However, it is not clear which method suits what applications.We provide a comparative study of four methods that relax stringent conditions for spectral test, based on several practical criteria. This enables the user to select the most suited method for his/her situation.Ideal DFT-based spectral test Figure 1 shows the ideal discrete Fourier transform (DFT)-based spectral test setup. The IEEE recommended requirements include the following.

No abstract