Development And Preliminary Study Of Transverse Beam Profile Monitor System At The HLS â…¡ Storage Ring

In the electron storage ring, beam turn-by-turn transverse size and position will change in several or even dozens of turns in many cases such as beam injection phase or high beam current leading to beam instabilities. The existing beam position measurement system or bunch-by-bunch feedback system can only measure the turn-by-turn or bunch-by-bunch transverse beam position, but they can’t be used to measure the beam size directly. Transverse profile measurement systems based on synchrotron radiation usually use a CCD camera as a sensor, which has a maximal frame rate of several hundred Hertz. So these systems are indeed average measurements with multi-bunches in multi-turns, and can’t be used to measure fast beam transverse size. Therefore, this paper designs fast beam transverse size and position measurement system for HLS II storage ring based on a linear array multi-anode photomultiplier tube (MAPMT), and carries out some preliminary researches.We first investigate the status in the field of fast beam transverse size and position measurement. Foreign laboratories adopt Gaussian fitting method to extract beam size and position, which needs to obtain up to dozens of photocurrent signals from an array detector synchronously. The paper proposes a method of using the logarithm processing algorithm to extract beam size and position. The algorithm requires only four consecutive photocurrent signals of MAPMT and hugely reduces the cost and complexity of the system.In the logarithm processing algorithm, we define the ideal beam size signal and simulation results show it has a good linear relation with beam size. When the beam size varies within o=0.8-2mm, changes of beam position have an effect on normalized beam size signal less than 1%. Then we define the ideal position signal and simulation results show it has a good linear relation with beam position when beam position varies within δ-2.0-2.0mm. And changes of beam size have an impact on the position on signal sensitivity within 1%. Given the channel consistency among MAPMT anode channels, we put forward a method to calibrate the channel consistency with the beam size measured by a CCD camera and then modify the algorithm. The simulation results of modified beam size and position signal are in accordance with the ideal case.We redesign the magnification of light path to meet the need of the image size of MAPMT in both directions, and calibrate three achromatic lens to obtain the actual lens focal and positions of principal planes. Then we research and select the detector’s voltage-divider circuit and high voltage power supply. In addition, we design the photocurrent signal conditioning circuit, and the test results tell that bandwidth of low noise wideband amplifier is approximately 420MHz, gain errors between all the four channels and the second channel are less than 0.2dB, and peak signal errors among them are no more than 2%.In the low-speed scheme, we verify logarithm processing algorithm with average anode photocurrent and get the channel gain parameters. In the high-speed scheme, we analysis the jitter characteristics with interpolation and digital integral method with only one bunch filling in the storage ring. With swept frequency excitation, we get the vertical fractional tune using four photocurrent signals and measure the system resolution without excitation.In these measurement results listed above, we use high-speed oscilloscope to acquire data. As oscilloscope has a limited resolution and its sample rate isn’t the integer multiple of radio frequency of storage ring. Besides, the obtained data can only be analyzed off-line. So we develop a fast beam transverse profile monitor based on a commercial digital signal processor in the end. The natural logarithm can be deduced from CORDIC algorithm in hyperbolic mode, and we extend the input range of natural logarithm by increasing k≤0 iteration. Then we implement the logarithm processing algorithm in the LabVIEW FPGA, and the test results meet the requirement of the turn-by-turn processing speed and accuracy of the whole system very well.