A High-Performance Procedure for Effective Number of Bits Estimation in Analog-to-Digital Converters

This paper deals with a procedure for high-accuracy fast estimation of the effective number of bits ( ENOB) of an analog-to-digital converter (ADC). According to this procedure the ADC output sine-wave parameters are determined through the interpolated discrete Fourier transform (IpDFT) method. One criterion for the selection of the window that will be used in the IpDFT method is provided. In addition, mathematical expressions for the accuracy of both the sine-fitting procedure based on the IpDFT method and the proposed ENOB estimation method are derived, and a lower bound on the number of acquired samples that ensures accurate ENOB estimates with a high confidence level is proposed. In particular, it is proved that the uncertainty of the proposed ENOB estimation procedure is almost the same as the theoretical lower bound for the variance of any unbiased ENOB estimator. Finally, the accuracy of the proposed procedure and the algorithms suggested in the existing standards for ADCs testing are compared through both computer simulations and experimental results. In addition, the processing times required by each considered method are compared, therefore proving the advantage of the proposed procedure in terms of the required computational burden. © 2006 IEEE.