Histogram Measurement of ADC Nonlinearities by Using Incomplete-Scale Sine Waves

Abstract

To compute nonlinearities, the histogram method, in which the input sine-wave signal has a range exceeding the full scale, is usually employed. In some cases, however, it is required to determine the characteristics of an analog-to-digital (ADC) system for the incomplete-scale sine wave. A method of measuring static ADC errors for the incomplete-scale input sinusoid, which is a modification of the histogram method for a signal exceeding the full scale, is developed. In the standard histogram method, the overlapping of the range of the input voltages corresponding to the full scale is necessary to ensure the fallout of all possible codes at the ADC output in the presence of gain and zero offset errors. The magnitude of the offset in the standard method is determined based on the number of samples that fall outside the operating range of the ADC. In the proposed method, the conditional overlapping of the range is used to determine static ADC errors. The computation of static ADC errors is reduced to calculating the amplitude of the output signal based on the histogram of advantageous ADC codes. The code transition levels are computed; and the offset and gain errors are determined. The proposed method is tested by modeling in MATLAB. The accuracy of the method is consistent with the expected values of the input errors. The proposed method makes it possible to compute the conversion errors of ADC systems with blocks that limit the amplitude of the input ADC signal.