Analysis of Dynamic Element Matching (DEM) in Pipelined ADCs

Abstract- Dynamic element matching (DEM), typically used to algorithm dynamically selects different DAC unit elements enhance the linearity of the digital-to-analog converter (DAC) to represent a given digital input code at different times. in a multi-bit Ay, modulator, can also be utilized to linearize This technique, typically utilized to linearize the DAC of a the sub-DAC in a multi-bit pipeline stage of a pipelined analog- multibit Al modulator, can also be applied to the sub-DAC to-digital converter (ADC). In this paper, an analytical.. approach to estimate the effect of DEM on the spurious-free of a multibitpipeline stage [6,7].dynamic range (SFDR) of a pipelined ADC is presented. A In a multibit pipeline stage, zero-order DEM can closed-form expression for the 3rd-harmonic power and, hence, randomize the linearity error due to the sub-DAC the SFDR of a pipelined ADC is derived. The proposed nonidealities and, hence, spreads the signal harmonics analysis shows that performing DEM in the first stage of a without correcting for the linearity errors [7,8]. Therefore, pipelined ADC reduces the 3rd-harmonic power in the ADC the SFDR of the ADC is improved, while its signal-to-noise-output spectrum by a factor of (2m4-), where m is the number and-distortion ratio (SNDR) is not increased because its of effective bits in the first pipeline stage. Behavioral Monte- noise floor is not decreased. Although it has been Carlo simulations are presented to demonstrate the accuracy of demonstrated that DEM reduces the power of the signal the proposed analysis. harmonics due to the sub-DAC nonidealities [6-8], no exact I. INTRODUCTION relation has been presented for the harmonic power befor