Houjun Wang

Research on a method to improve the waveform acquire rate of DSO

The waveform acquire rate is an important index to judge Digital Storage Oscilloscopes data acquisition performance. This paper introduces a method to improve the waveform acquire rate of the digital oscilloscope with a parallel coprocessor framework. Firstly, the relation between waveform acquire rate and dead time and the relation between waveform acquire rate and transient signal are analyzed from the perspective of probability theory. Then the design and implementation plans are explained in detail. Finally, this technology is applied into the design of a certain type of oscilloscope, and the relative testing results are provided.

Analysis on multiple-component synchronization of ultra-fast time-interleaved analog-to-digital conversion systems and its novel parameterized hardware solution.

Parallelism-based technique of time-interleaved analog-to-digital conversion (TIADC) has become an effective solution for the higher sampling rate acquisition system to acquire non-repetitive waveforms. With the increase of sampling frequency, the indeterminacy of combining sequence of sampled data among multiple components has become a highlighted barrier for the reset operation of high-speed acquisition systems, and this is especially obvious for the ultra-fast TIADC systems. In this paper, we clarify the root of the problem in multiple-component synchronization (MCS) caused by such reset operation. Also we propose a novel and reliable hardware solution to precisely condition each reset signal, including three key circuit design parameters, i.e., the best time interval, required edge uncertainty, and the minimum delay precision. Besides, the designing scheme and debugging procedures are presented in detail in a generalized platform of this system type. Finally, in order to demonstrate the feasibility, parametric materialization and testing verification are gradually accomplished in a 20 Giga Samples Per Second (GSPS) TIADC system composed of four 5 GSPS ADC components. The results show that the proposed method is feasible and effective for ensuring the combined determinacy of multiple groups of sampled data and solving the MCS problem. In comparison with other existing solutions, it adopts some simple logic components more easily and flexibly, and this is significant for the development of congeneric systems or instruments featuring the MCS.

Research on the correction method of three-dimensional non-uniform errors in the parallel acquisition system

Parallel alternative sampling is an effective way to break through the speed limit in analog-digital (A/D) conversion and improve the digitalization rate. A normalized correction algorithm is proposed through the analyses of the mathematical model of three errors of parallel alternative acquisition system; the error of inter-channel sampling time is then obtained by estimating the non-uniformity parameters of time on condition that the offset and gain errors of the system are solved; finally, the time non-uniform errors is corrected through the high-precision programmable clock delay network. The correction method proposed in the thesis, after testing and verifying, is proved to be with a high real-time performance and have reduced the design difficulty of hardware and costs and improved the real-time performance of the system.