Data Acquisition And Treatment Of The Multiple Non-exponential Decay

The multiple non-exponential decay is a universal phenomenon in nature. Its time span is so broad and lifetime distribution is so complex, that the traditional methods are not applicable. Using the method of Lg(t) derivative to process data of the multiple decay, we can intuitively display the distribution of lifetimes. The basic characters of the Lg(t) derivative peaks of the single and multiple decay, such as the peak shape and the full width at the half maximum (FWHM), place a key in the data analysis of the work. When the width of lifetime distribution is very broad, the time span of the K(t)-Lg(t) figure will be very wide also. Therefore, in order to display the multiple decay, the time span of the data should to be wide enough. In this work, we design and facture a data acquisition system which is based upon the PC bus. In the system we use an AD9432 ( a 12-Bit, 105MSPS A/D converter ), a LTC1603 ( a 16-Bit, 250KSPS A/D converter ), an IS61LV25616AL-10 ( a 256K×16 asynchronous static RAM, its high-speed access time is 10 ns ), and a high-speed address counter (composed of five 74F163A counters with counting frequency of 100MHz ). Arranging the timing of the write enable of SRAM, the encode clock of AD9432 and the outputs of high-speed address counter in reason, the outputs of AD9432 will be stored in SRAM accurately, and realize high speed data buffer storing with counting frequency of 100MHz. The simple interface circuit is based upon the ISA bus, which can realize the start control of high-speed ADC, the sampling control of slow-speed ADC and the postmortem transfer of buffer data. The system accomplished data acquisition over a 10th order time span, from 10 ns to 100 s. In order to make sure the outset of the high-speed data, we acquired a single decay with the lifetime of 100 ns, by comparing its Lg(t) derivative peak with theoretical peak we can calibrate the outset of the high-speed data. The shape of the Lg(t) derivative peak is susceptible to the errors of experimental data, and is unrelated to the initial decay value N00. Compared with the method of Lg[N00-N(t)], the method of Lg(t)derivative is more accurate. A simple and flexible signal source of multiple decay was designed with the linear adding of many RC charging signals. Two suits of multiple decay data which are acquired independently over a 10th order time span indicate that, the shape of the Lg(t) derivative peak is coherent with its theoretical peak unitarily. The Lg(t) derivative peaks come from its experimental data. Its corresponding theoretical peaks come from the theoretical multiple decay formula with parameters measured from each RC component and each measured amplitude of single decay. The mean value of the relative error between experimental data and theoretical data is about 6%, which indicates that the data from the acquisition system has better reliability.