Defining and Testing Dynamic Parameters in High-Speed ADCs, Part 1 - AN728

The first part of this article series discusses commonly known definitions most crucial for high-speed data converters (in this case analog-to-digital converter or short ADCs) used in communications, instrumentation and data acquisition applications. The purpose of this article is to help the reader gain a better understanding of common parameters such as signal-to-noise ratio (SNR), signal-to-noise-and-distortion (SINAD), total harmonic distortion (THD) and spurious-free dynamic range (SFDR). In the second part of this article series (see "Dynamic Testing of High-Speed ADCs" for further reading), these parameter definitions are put to the test by measuring them in real-world test scenarios. Additional Information: Dynamic Testing of High-Speed ADCs, Part 2 Dynamic specifications for ADCs are very important in high-speed applications such as digital communications, ultrasound imaging, instrumentation, and IF digitization. The following discussion provides a definition and a mathematical foundation for each parameter, offers useful techniques for evaluating the dynamic performance of high-speed ADCs, and explains how the dynamic parameters correlate with ADC performance. Part 1 of this twopart discussion covers the definition of these specifications: ● Signal-to-noise ratio (SNR) ● Signal-to-noise and distortion ratio (SINAD) ● Effective number of bits (ENOB) ● Total harmonic distortion (THD) ● Spurious-free dynamic range (SFDR) ● Two-tone intermodulation distortion (TTIMD) ● Multi-tone intermodulation distortion (MTIMD) ● Voltage standing-wave ratio (VSWR) In explaining how to measure these parameters, Part 2 provides insight into the practical aspects of dynamic performance testing. Note that some specifications allow more than one approach for measurement and even for definition. Thus, the test techniques of Part 2 represent one method and are not mandatory. Any of the methods described can be extended or altered as necessary to suit the application at hand. When testing high-speed A/D converters, one emulates the operation of a spectrum analyzer used to quantify linearity in analog circuits. For this instrument and for the test procedure, dynamic specifications are usually expressed in the frequency domain, using the Fast Fourier Transform (FFT). In both cases, the data output represents the magnitude of this FFT. As an example (Figure 1), consider the FFT plot for an 80Msps, 10-bit ADC designed and optimized for ultrasound imaging and the digitization of baseband/IF frequencies. Such FFT plots contain impressive amounts of information and can be generated quickly. However, to make use of an FFT, one must understand how its parameters are defined. Figure 1. 8192-point FFT plot for the MAX1448. Signal-to-Noise Ratio (SNR) For a waveform perfectly reconstructed from digital samples, SNR is the ratio of an rms (root mean square) fullscale analog input to its rms quantization error (AQUANTIZATION[rms] = ALSB/ = AREF/[2 × ]). The rms value of a sine wave is one half its peak-to-peak value divided by , and quantization error is the difference between an analog waveform and its digitally reconstructed replica, which is characterized by a staircase-shaped transfer curve. The difference function resembles a sawtooth wave that oscillates once per sample between the levels +1/2LSB and -1/2LSB (LSB being least-significant bit). The difference function's rms value is its peak value (1/2LSB) divided by . For an ideal N-bit converter, SNR is defined as