Research On Measurement For Aperture Jitter Of ADC

With the rapid development and wide-spread application of computer technology, the technology of ADC as the bridge between the computer and the physical information to be processed is getting more and more important. Furthermore , in the state-of-the- art communication systems and in the developing radio soft-ware schemes , the using of ADC is pushed into the very front-end of the system to realize the direct sampling under the IF or RF sampling frequencies . This makes the neglected aperture jitter lower under picoseconds and sub-picoseconds when sampling at relatively lower frequencies to be the major limiting factor for the bad influence on the ADC sampling performance and makes it the focus of being paid attention to.The aperture jitter is the timming jitter occurred in the sampling instant caused by the sampling instant uncertainty and is one of the most important parameters for the ADC. But up to now there is no standardized and uniformed method to be used.The definition and the causation for the aperture jitter is introduced firstly in this article. It is mentioned here that dislike the external jitter, such as the clock jitter caused by the clock generator, the aperture jitter is caused by the inner SHA circuit in the integrated circuit of ADC.For the noise caused by aperture jitter is always among the total noises caused by all kinds of noise sources, so to get the insight for it , not only the non-jitter noise sources, but also the external jitter are discussed here.Different methods for measuring jitter are introduced and compared in details. Based on these knowledge ,a new method is proposed in this article .This method utilizes the technique of using a common sine wave generator for supplying the input signal and the sampling clock to realize the separation of the aperture jitter from the overall-jitter ,meanwhile using the line-fitting method to extract the jitter from the total noise including non-jitternoises according to the linearity relationship between input frequencies and output errors suggested in the derived formula by calculating deviation of 16384 samples sampled at the zero-crossing point of the input signal at different input frequencies .Thus the aperture jitter is obtained .At last, to verify the validation of this method, simulations are taken by using the AD6645 simulation model supplied by AD Company under the MATLAB circumstance, and comparisons are also executed by simulating the dual-histogram way proposed by Turkey Kuyel. The results prove that the measuring precision for the new method is better than the latter mentioned above, and is satisfying both in precision and in confidence.