Font Size: a A A

Reseach And Implementation Of Mismatches Calibration Algorithm In TIADC System

Posted on:2016-08-10Degree:DoctorType:Dissertation
Country:ChinaCandidate:K J YangFull Text:PDF
GTID:1108330473456096Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
Data acquisition system is widely used in electric systems such as communication system, instruments and Radar. To meet the requirement of capture and process of complicated wide-band signal, it is significant to research data acquisition system with higher performance. Sample rate is one of the important indicators in data acquisition system. When sample rate gets higher, the bandwidth which can be analysed becomes wider, and the sample interval between two sample points becomes shorter which implies the original signal can be reconstructed better. The sample rate of data acquisition system is decided by the conversion rate of ADC(analog-to-digital converter). In recent years, the conversion rate of ADC has been enhanced greatly with the devolepment of manufacturing technique of intergrated circuits. However, the conversion rate is still too low to the requirement of real systems. Under the condition of nowadays, time-interleaved technique is the most suitable method to increase sample rate and ensure the performance of system. We can break through the limitation of the conversion rate of single ADC chip and realize ultra-high-speed data acuiqsition sytem by using of this technique.An ADC system based on time-interleaved technique is called TIADC(Time-interleaved ADC) system. Although the sample rate is increased by time-interleaved technique, the performance of TIADC system is decreased by offset error, gain error and timing skew error. These errors are introduced by the difference between multiple ADC chips and the nonuniformity of sample interval. The mismatched will result in distortion in time-domain and spectrum spurs in frequency-domain. Accordingly, the SNR(signal-to-noise ratio) is decreased. I dedicate myself to the design and realization of ultra-high sample rate digital oscilloscope based on TIADC. Thus, the research of this dissertation is focus on how to estimate and correct the errors in TIADC system. Combining with the research projects granted by National Nature Science Foundation of China, this dissertation mainly does some thorough researches on the following aspects:1. According to the structure and principle of TIADC, the system model of TIADC is constructed. Then, according to the realization type of TIADC system in engineering, the mismatch resources of real TIADC system are analyzed. The output spectrum with mismatches is derived. Especially, the mismatches in TIADC are analysed by a random input, which is different with the traditional analysis with sine wave input. The effect of mismatches on system performance is given to guide the design of real TIADC system.2. Two mismatch estimation methods are proposed. In the sine-fit based estimation method, three-parameter sine-fit algorithm is utilized to obtain the offset, gain and phase of each channel in TIADC. Furthermore, the mismatches are calculated by the obtained offset, gain and phase. Compared with the estimation method, which utilizes interpolation to obtain standard output and use the least square algorithm to calculate mismatches, proposed by existing liteture, no interpolation is need. Accordingly, there is less calculation in our method. In the other method named statistic-based estimation method, the average, summation of absolute value and product of sampled data are utilized to calculate the mismatches of TIADC. There are few floating-point calculations in the algorithm, and it can be realized very easy in a real system.3. The Lagrange interpolation is employed to correct timing skew, a filter structure which can be realized in a real system is given, and the coefficient of interpolation filter is derived. A timing skew adaptive calibration method based on product of sample data is proposed, the calculation of adjustment step of timing skew is more easier than existing adaptive method. From the frequency domain, a FFT-based timing skew adaptive calibration method is proposd. In this method, the success-failure method is used to obtain the timing skew range, then, quadratic interpolation is utilized to calculate the adjustment step of timing skew. The convergence speed of our method is much better than most of existing adaptive calibration method.4. Based on the TIADC structure, a 20 GSPS digital oscilloscope is developed. Considering the requirement of system, a low jitter sample clock circuit whose phase is adjustable is designed. The data synchronization problem between multiple ADC chips in TIADC system is also researched intensively. The cases which result in synchronization problem are analyzed, and the solution to solve data synchronization problem is given. The aforementioned estimation and correct method are used to calibrate the system, thus the system performance is improved. To solve the real-time data process problem in a high-speed TIADC system, a seamless acquisition method is proposed.
Keywords/Search Tags:adaptive calibration, seamless acquisition High-speed data acquisition, time-interleaved sample, TIADC, sine fit
PDF Full Text Request
Related items