Signature-based testing for adaptive digitally-calibrated pipelined analog-to-digital converters

This paper presents a new methodology for testing the adaptive digitally-calibrated pipelined ADCs. By the using the uncalibrated output codes, the health of the ADC under test can be reliably examined. We propose using ¿the error signature¿ which is obtained by integrating the absolute difference between the uncalibrated and exact output codes as a response to applying an on-chip generated test stimulus. In contrast with the existing approaches, our method has three achievements. First, it avoids the margin degradation masking that might be caused by the adaptive calibration engine, and hence our test results are more reliable. Second, it relaxes the need for highly sophisticated ATEs and requires shorter test time. Third, the results can be efficiently used for device binning. The method is scalable and can be applied to low as well as high resolutions ADCs. The results demonstrate the effectiveness of our method..