Research And Implementation Of Spectrum Analysis Function In Mixed Domain Analyzer

With the rapid development of modern electronic information technology,electronic signals are presenting the characteristics of complication and diversification.In order to meet the testing requirements of complex wideband time-varying electronic signals,measuring instruments are required to analyze not only the time-domain characteristics of signals,but also the characteristics of other domains,including frequency-domain and time-frequency-domain of signals.Mixed domain analysis is becoming increasingly prominent in the field of modern electronic testing.Based on the mixed domain analyzer platform with analog bandwidth of 8GHz and sampling rate of 20 GSPS,the author researched and developed a spectrum analysis function for signals,which can overcome the limitation that traditional oscilloscopes can only process and analyze signals from the perspective of time domain.In addition,a maximum spectrum analysis bandwidth of 8GHz and a high-frequency resolution of10 Hz are realized in this paper,where detailed spectrum analysis of high-frequency narrow-band signals can be carried out.Specific research contents are as follows:1.The problem that it is difficult to realize high resolution under high sampling rate is solved by increasing frequency resolution through combination of digital down frequency conversion technology and FFT.Digital down frequency conversion consists of two parts,i.e.frequency conversion and extraction filtering.Frequency conversion moves the frequency band of target signal to low frequency.Extraction filtering realizes down-sampling and aliasing suppression of high-speed data streams.Where,the Cascade Integrator Comb filter and half-band filter cascaded structure is used for extraction filtering.The sampling rate can be changed by setting the extraction ratio of CIC filter and the number of cascades of the half-band filter,realizing multi-rate signal processing.In this paper,the designed maximum extraction ratio is 2000000,and a frequency resolution up to 10 Hz can be realized.Moreover,a spectrum analysis bandwidth of 8GHz can be obtained by performing FFT calculations directly after frequency conversion of time-domain data.2.Real-time analysis and processing of frequency-domain data is realized.Because the FPGA master clock used in this paper is 250 MHz,in order to realize real-time spectrum analysis under 20 GSPS,the extraction filtering module is divided into two parts:80-channel parallel and single-channel serial.The parallel CIC and parallel half-band filter structure is designed and implemented to solve the time sequence problem of parallel filtering through hierarchical operation and multiphase structure.3.The time-frequency analysis function is realized through short-time Fourier transform,which can perform joint time-frequency analysis against complicated unstable signals.The discrete data after digital down-conversion processing is overlapped and outputted,and is multiplied by a certain length of window function sequence.Then,the short-time spectrum of signal is obtained through FFT calculation of windowed data.At the same time,the time resolution of STFT analysis results can be adjusted by changing the length and type of window function.Finally,on the mixed domain analyzer platform,various types of unstable signals are used to test the above spectrum analysis function and related indexes,and the correctness of the spectrum analysis function is verified.