FPGA-based Digital Multi-channel System

Thanks to the great advantages of digital signal processing technology and the development of high-speed ADC technologies, digital multi-channel systems have more advantages than analog multi-channel systems. Digital multi-channel systems(MCA), due to the stability and flexibility, have a higher pulse processing precision and a higher pulse passing rate, and the method of amplitude extraction can be further optimized. Programming of digital multi-channel systems is more convenient.We can change and adjust the pulse processing algorithm or parameters more flexibly. Besides, the digital circuit has a smaller volume, and is more accurate and widely-used than the traditional analog circuit. Baseline fluctuation is one of the major problems affecting the energy resolution of the digital multi-channel systems, which can be eliminated by method of baseline subtraction. However, random high-frequency noises always superimpose on the baseline, as a result, the baseline discrimination becomes very difficult. Taking the method of maximum and minimum searching, the pulse peak and the baseline can be determined rapidly, and the random high frequency noises can be eliminated. The energy resolution and count rate of the spectrum system can be improved.In this thesis, the design is accomplished on the ARM + FPGA digital multi-channel system, of which the analog front-end circuits include the preamplifier and linear amplifier using the CR-110 and CR-200 from Cremat respectively. As to the digital multi-channel system, the controller of ARM is the TQ2440, and the digital processing chip FPGA is the EP2C8Q208C8 N manufactured in Altera.The current pulse produced in the detector is firstly transmitted to the pre-amplifier circuit, and then is shaped by the Gaussian-shaping circuit. The Gaussian pulse signal is sampled by a high-speed ADC, and then is transmitted to the FPGA to perform the threshold discrimination, peak detection, amplitude acquisition and storage. The amplitude data is transferred to ARM in serial, and then is transferred to the upper computer via the USB cable, thus the energy spectrum be drawn on the interface of the host computer. Communication between FPGA and ARM is carried out by means of bus transmission, and the communication between ARM and PC is realized through the USB cable. ARM drive the FPGA device and the USB device. The edition and generation of the program of the driver module are completed in the Linux operating system in the host computer. In this thesis, the digital programming is accomplished using HDL Verilog language of the Quartus II software. Circuit simulation software is the Modelsim6.6 from Mentor Graphics. As to the amplitude acquisition algorithm, there are three choices: peak to peak algorithm, average baseline algorithm, fixed baseline algorithm. And the influence of the three algorithms on energy spectrum resolution is analyzed. In the experiment study, we use Am241 radiation source to irradiate the cadmium zinc telluride(CZT) detector. We analyzed the spectrum change and the corresponding principle when linearly increasing bias voltages are applied to the detector respectively.The system adopts a very good interface mode. ARM can control the FPGA, and the acquisition time is set on the interface of a PC. The operation of the entire system is controlled by controlling FPGA chip, thus the system is quite flexible. When inputting different preamplifier signals with different baseline values, the threshold of the algorithm can be modified to obtain the spectrum. It can eliminate the baseline subtraction in the traditional analog circuit. This digital multi-channel system is very practical, and is beneficial to the implement of high-performance spectrum analysis systems.