Method for full-digital random sampling

摘要

For the signal under acquisition which varies monotonically before and after the trigger time, a method for full-digital random sampling employs first sampled data before the trigger time and first sampled data after the trigger time to fit a curve, and obtains an intersection point of triggering level and the fitted curve, then, calculates the time interval between sampled data after the trigger time and the intersection point in the end, reconstructs the original signal, i.e. the signal under acquisition by a time interval of each acquisition. Thus, an analog trigger circuit and a time measurement circuit of conventional random sampling system can be eliminated, that simplifies the circuit design of data acquisition system and decreases its hardware complexity. Moreover, the higher sampling rate for the signal under acquisition is attained, and more waveform details are obtained.

主权项

1. A method for full-digital random sampling, the method comprising the following steps:
(1). Acquisition of Sampled Data
acquiring sampled data of a signal by an analog-to-digital converter (ADC) of a random sampling system, the signal under acquisition varying monotonically before and after a trigger time, and in a monotonic section of the signal, a sampled data can be obtained before and after the trigger time respectively;sending a sampling command randomly by a controller of a data acquisition system to open a writing enable (wen) of a FIFO (First In First Out) for storing sampled data from the ADC;when attaining a pre-triggering depth Lp of the FIFO:
opening a reading enable (ren) of the FIFO to maintain a volume of sampled data with depth Lp;detecting whether the sampled data is satisfied with a triggering condition; andif the sampled data satisfies the triggering condition, closing the reading enable (ren) of the FIFO; andwhen attaining a storage depth Ls;
closing the writing enable (wen) of the FIFO such that a range of sampled data in the FIFO is [dn−Lp+1,dn+Ls−Lp], where 1≦Lp≦Ls, the trigger being between a sampled data dn and dn+1, where n is the position of a first sampled data before the trigger and n+1 is the position of the first sampled data after the trigger; (2). Time Interval Calculation
fitting a curve based on the sampled data dn and dn+1, obtaining a intersection point TrL of a triggering level AT and the fitted curve, and calculating a time interval tL between sampled data dn+1 and the intersection point TrL; (3). Waveform Reconstruction
according to step (1) and step (2), repeating the process of acquisition and time interval calculation, and obtaining multiple pluralities of sampled data dij and time intervals tLi of each acquisition, where i is a acquisition serial number, iεN, N is a number of acquisitions, j is sampled data sequence number;interlacing the multiple pluralities of sampled data dij by ascending order of the time interval tLi, and storing the interlaced multiple pluralities of sampled data dij; and,reconstructing the signal under acquisition with the interlaced multiple pluralities of sampled data dij.