Abstract
Analog-to-digital converter (ADC) characterization is usually performed using stationary stimuli like sine waves. However, the use of a non-stationary stimulus, besides providing testing conditions closer to those found in real applications, can lead to interesting improvements in ADC testing speed, This kind of signal needs proper processing techniques in order to extract useful information. In this paper we propose the use of joint time-frequency analysis (JTFA) for this purpose. The basic principles of the technique, and how it can be used in ADC testing are presented. in particular, a method for characterising an ADC on its entire bandwidth using a single stimulus is described.

Abstract
The estimation of the harmonic content of an ADC output is fundamental to evaluate its suitability to perform according the requirements specified for an application. The use of the traditional frequency analysis leads to a large hardware overhead due to the amount of on-chip processing being involved, or to a large quantity of data to be transferred in case the processing is performed in a tester. This paper presents an algorithm capable of estimating the harmonics with similar accuracy but with the advantage of being more suitable for a BIST implementation, since it requires a reduced number of on-chip operations, and that only a small set of values has to be supplied outside the chip for further processing. It relies, on the fact that harmonics generated by an ADC are mathematical related with the polynomial coefficients of its transfer function. ADC offset and gain errors can also be measured.

Abstract
A procedure is proposed to estimate an analogue-to-digital converter's signal-to-noise plus distortion ratio using the histogram method. The procedure provides results that are in close agreement with the ones obtained with the spectral analysis and sinewave fitting methods. It is shown that the errors obtained by using former implementations of the histogram method are due to not considering the input stimulus probability density function, and it is shown how these errors can be rectified.