Detection And Implementation Of Sweep Signal In Electrical Interconnection Based On All Phase FFT Algorithm

In the semiconductor manufacturing process,the problem of incomplete signals caused by breakpoints in board level interconnections is very serious.In order to ensure the integrity of signals in interconnected lines,it is often necessary to determine the location of breakpoints through detection methods.Therefore,this paper mainly focuses on the swept continuous wave(FMCW)detection method,using the all phase FFT algorithm to detect breakpoints.Among them,the all phase FFT algorithm has linear invariance compared to the FFT algorithm,allowing the signal resolution of MHz level to reach a signal frequency of 1Hz.This paper will design and implement the above detection device.Firstly,the relevant theories of PCB board level interconnections were elaborated,and the physical model and transmission characteristics of board level interconnections were introduced;Introduced phenomena such as mutual interference and ground bounce in interconnection lines and their solutions;Introduced the calculation of parasitic capacitance,resistance,inductance and other related parameters in the RLGC model,as well as the calculation of S parameters for network ports such as single port,two port,and four port.Secondly,the design and implementation of a device for collecting and filtering swept frequency signals in board level interconnection lines are carried out.Firstly,the FPGA chip and ADC sampling chip are selected,and the corresponding PCB board is designed to collect the sweep signal.The collected signal is filtered in FPGA.Use the software Vivado to implement the filtering function of FPGA.The digital filtering function is realized by saving the filtering coefficients obtained from the IIR type filter designed in Matlab as a CSV file and transferring them to the IIR IP core in Vivado.Thirdly,the system of all phase FFT is optimized and implemented.Firstly,the idea of all phase preprocessing and the related processes of three types of preprocessing:windowless,single window,and double window are described.Through comparison,the method of double window all phase preprocessing is selected,and on this basis,the method is optimized to reduce the distance between the main and side lobes.After building an optimized dual window all phase FFT system model in Simulink for simulation verification,this function is implemented in Vivado.Finally,experimental analysis was conducted on the scanning signal detection device in the interconnection line of the plates and poles.Build a physical detection device for the interconnection line of the board and pole.The relevant parameters in the PCB interconnection line length detection experiment were determined.Through relevant experimental data,it can be shown that the error of this experiment is 10.09 mm for the maximum interconnection line length,0.28 mm for the minimum interconnection line length,and the relative error percentage is below 1%.The difference between the maximum and minimum interconnection line lengths of all phase FFT and FFT processed data in this experiment is 7.78 mm and 0.13 mm,respectively.The relative proportion of the difference is about 1.4%.In summary,this article focuses on the swept frequency signals in board level interconnection lines and applies the all phase FFT algorithm to the processing of swept frequency signals.By designing processing processes such as acquisition,filtering,and all phase FFT,and conducting experimental tests,the detection and implementation of the sweep signal in the board level interconnection line are completed,and the position of the interruption point in the board level interconnection line is determined to ensure that the accuracy error of the position is within 1%.