| X-ray spectroscopy is one of the most important means of information acquisition in the fields of nuclear radiation,X-ray fluorescence and nuclear technology applications.In the 1990s,with the continuous development of manufacturing processes and nuclear electronics,the rate of the energy resolution of detectors is constantly improving.Since the 21st century,the nuclear electronic system has gradually evolved from analog circuits to digitalization.The energy resolution of various common detectors is approaching to its inherent energy resolution,while the application field of nuclear pulse signal processing is constantly expanding.But during its rapid development,it also encountered technical bottlenecks.Semiconductor detectors are widely used for their high energy resolution,such as HPGe detectors and FAST-SDD detectors.However,the switch reset type preamplifiers used in semiconductor detectors are frequently reset,resulting in more generation of mutation pulses,which are not identified or repaired,will exist as false peaks in front of the omnipotent peak in the final energy spectrum,seriously affecting the fine analysis of the energy spectrum.If the false peak is just superimposed on the characteristic peak of a weak element,the identification and the complex caculation content of the characteristic peak will be deeply weakened.All the research contents in this paper are focused on the distorted pulses generated by the preamplifier circuit and the false peaks caused by the distorted pulses.In order to obtain a more accurate spectrum and a reliable count rate,this paper proposed two nuclear pulse processing techniques,which is based on pulse shape discrimination,digital pulse shaping and pulse pile-up.The main contents are as follows:1.Aiming at the problem of false peaks in the measured spectra caused by distorted pulses generated by frequent reset of switch-reset preamplifiers in semiconductor detectors,a pulse elimination technique is proposed to deal with the distorted pulse.The core algorithm of pulse elimination technique is to discuss the rising time and the jumping time,and analyze the influence of different rising time and jumping time on the rejection results.Finally,it is concluded that when the rising time of triangulation is later than the jumping time of pulses,such pulses will be rejected by the pulse discrimination unit as wrong pulses.In the research process of pulse elimination technology,simulation and experiment are carried out.2.The object of pulse repair technology is still distorted pulses generated by reset of switch reset preamplifier,but the method of identifying distorted pulses is different from pulse elimination technology.Considering the characteristic that all sampling points in the pulse mutation part instantaneously jump to zero,the zero-judgment method is used to locate the distorted pulses,and all zero sampling points are iteratively repaired by the value of the previous sampling point.The theoretical basis of repairing is to select a seven-order successive approximation method which has good repairing effect and is easy to implement in the FPGA.After repairing,the negative exponential pulse sequence is triangulated.3.In the experiment,FAST-SDD detector is used to independently develop the X-ray spectrometry measuring device.Taking 55Fe standard source,one kind of rock sample and a self-made iron,tin and strontium sample as the measuring objects,the effects of pulse elimination and pulse repair techniques on the treatment of distorted pulses are verified respectively;the influence of two pulse processing techniques on peak area and counting rate was validated with 238Pu standard source;the probability of distorted pulses in samples with different counting rates was analyzed;the stability of pulse repair technology was analyzed.The experimental results show that:(1)Both pulse rejection and pulse repair techniques can effectively deal with abrupt pulses and eliminate the false peaks caused by abrupt pulses.(2)Taking 238Pu standard source as the measurement object,the increase rate of the four characteristic peaks in the low energy section obtained in each measurement fluctuates greatly through peak area analysis.The increase rate of peak area of characteristic peaks in high energy band tends to be stable,while the increase rate of peak area sum of all selected elements is stable at about 2.65%after restoration.(3)If the loss of counting rate caused by mistake elimination in low energy segment,the probability of producing distorted pulses is about 5.03%for Fe-Sn-Sr sample with count rate of 8.07×103/cps and 2.99%for 238Pu standard source with count rate of 5.251×104/cps.(4)The distribution probability of each element in each confidence interval is approximately equal to the theoretical outline of the interval.The measurement results conform to the normal distribution and the whole measurement system is stable and reliable. |