In view of the enormous number of samples required by the conventional sine wave code histogram test in estimating the integral nonlinearity errors of high-resolution analog-to-digital converters

a novel sine wave code histogram test that uses the moving-average filters is proposed

and is called the moving-average method. In the proposed method

the integral nonlinearity errors are roughly estimated by the conventional sine wave code histogram test using fewer samples

and then a moving-average filter is applied to the estimated results to obtain the final accurate integral nonlinearity errors. The results of simulations and tests show that the moving-average method can reduce the number of samples by at least 90% to achieve almost the same accuracy as the conventional sine wave code histogram test

and hence save the time and the cost of tests.