Detection Method of Digital Beam Position Monitor Based on Histogram

Abstract

The digital beam position monitor processer (DBPM) is one of the important beam diagnosis equipments in particle accelerator and it will be massproduced during the construction of Upgrade of Beijing Electron Positron Collider (BEPCⅡ) and high energy photon source (HEPS), which brings great challenges to the testing work. The analog to digital converter (ADC) module is an essential component of DBPM and its conversion accuracy determines the resolution of the beam position measurement. Therefore, in order to ensure that these two projects meet the design targets, it is necessary to test the ADC performance. Its main observation object includes the phaselocked loop, sampling clock jitter, signaltonoise ratio (SNR), etc. To meet the batch test requirements of the selfdeveloped DBPM ADC, a method of testing the sampling performance based on the histogram was proposed. The histogram’s test principle is that the shape of the histogram can detect whether the sampling clock is phaselocked to the input signal, the boundary value of the histogram reflects the amplitude of the input signal to judge if the ADC is saturated, and the average and standard deviation of the data peak on the histogram can be used to calculate the clock jitter that conforms to the normal distribution. In this paper, the mathematical relation between the statistics of each data peak, clock jitter, and SNR was simulated by MATLAB in the situation of phaselocking. The separation algorithm was designed to obtain the data corresponding to each peak of the histogram and compute the clock jitter and SNR. An integrated test platform was built and experiments were carried out to verify the testing effect of the histogram method. The experimental results show that when the amount of sampling data is appropriate, the histogram method can complete the test quickly and accurately, and the measurement result is close to that of the highperformance phase noise analyzer. The clock jitter measurement error is less than 200 fs, the SNR measurement error is less than 2 dB, and the entire test duration is less than 1 min, which fulfills large volume testing demands. Compared with the manual instrument measure method, the histogram approach is simpler and lower cost to test with the same order of magnitude accuracy and has been implemented by programming to reduce labor intensity. The test program has been implemented in the automatic test platform and used to batch detection of DBPM products successfully.

Keywords

• digital beam position monitor processer
• analog to digital converter
• histogram
• clock jitter
• signal-to-noise ratio