Accurate Spectral Testing With Impure Source and Noncoherent Sampling

Analog-to-digital converters (ADCs) are becoming increasingly more common to be involved in most systems with integrated circuits. One of the difficulties being faced is to be able to accurately and cost-effectively test the continually higher performance of ADCs. A part of this test is being able to assess the dynamic linearity of the ADC through dynamic spectral testing. The standard test method for ADCs can be difficult to implement accurately and cost-effectively due to the stringent requirements. This paper develops an algorithm that relaxes the requirements on the linearity of the test signal and of the need to achieve coherent sampling. The standard test requires that the input signal linearity be about 20 dB purer than the ADC under test along with always maintaining coherent sampling. This algorithm will reduce the purity requirement by allowing the test signal to be less pure than the ADC under test while also completely removing the need for coherent sampling. The proposed method will be able to test a 16-b ADC with a target total harmonic distortion (THD) of around -95 dB using a test input signal that has a THD as bad as -50 dB. A test signal with this level of purity would be easy to design as well as be easily accessible in most systems that are being designed. The accuracy and robustness of these methods are demonstrated through simulation and measurement results.