In many accelerators, particles are accelerated with radio frequency (RF) electromagnetic waves inside cavities. The quality of the acceleration directly depends on the stability of the frequency of the RF wave, which is provided as a pulse of a thousandth of a second. An electron is accelerated in an electric field of a sine curve. When temporal timing changes, the bunch size of the electron also changes, as the accelerated energy changes. In the ILC, RF stability must be controlled around 0.3%, and achieving this level of stability, a feedback system to stabilise the frequency using digital technology is essential for the Low Level RF generator (LLRF).

Zhe Qiao Geng worked on software for radio frequency control.

In Comparing RF control to sound, it corresponds to continue making a constant volume of sound at a constant tone. For example, regular tuning is necessary for a piano in order to get a constant tone and volume of sound. On the other hand, RF control can keep making a sound while it is being tuned.

Zhe Qiao Geng, ILC- IHEP member, has been studying software about feedback in LLRF for the Superconducting Linac Test Facility since 2005. In August 2005, Geng stayed at KEK for three months for the first time and worked on the development of software for the cavity simulator.

This cavity simulator study will determine the characteristics of circuitry which electrically simulate a cavity on a digital integrated circuit called, Field Programmable Gate-Array (FPGA). It will also confirm the performance of feedback instead of the actual superconducting cavity.

More than ten years ago, scientists performed a feedback study on one of the analog circuits, which used a resistance or condenser. That circuit responded quickly enough, but lacked the flexibility of being able to change the feedback parameters for the precise tuning of the beam. The recent development of high-speed Digital Signal Processor (DSP) technology paved the way to developing a custom-designed semiconductor chip that modifies the shape of the electric wave digitally. It is possible to change a parameter of feedback freely, highly increasing the system expandability, but it was inferior to feedback for the analog circuit in terms of stability. Thus, the time-current characteristic suffered from the input of a digital signal when using DSP. (In feedback, if parts from reception to output are high-speed, it becomes more stable.)

Starting in 2001, FPGA, with its high-speed capability of input and output for digital signals, came to be used in high-speed digital circuits. By using FPGA, scientists can get approximately equal feedback and stability from analog circuits. Therefore its stability is the same as the analog circuit.

In addition, it was possible to easily install various additional capabilities by an internal program because it was a digital circuit. FPGA is the device which came to be used mainly on such digital communication as cell-phones, digital television, and high-speed processors in the past five years. The development of these devices is connected to the development of recent digital technology in low-level RF control.

Geng, who visited KEK again one year later, was in charge of approximately three jobs for this study. Firstly, Geng worked on creating the fast-interlock program, using a commercially available FPGA board. This program stops the apparatus immediately by receiving the interlock signal if something goes wrong, and displays the status. Secondly, he worked on the development of the program for the monitor, using the high-speed FPGA board, which had a high-speed Analog to Digital Converter (ADC) and Digital to Analog Converter (DAC). (ADC and DAC is the exchange circuit between analog and digital.) "If we can monitor it on web, it can reduce costs. I think this is a useful and important study," Geng said. Thirdly, he also worked on the development of the Graphical User Interface (GUI).

"Geng did a very fine job," said Shigeki Fukuda, who has worked with him for more than four years on RF source collaboration for remodeling BEPC-II. "I want to keep a collaborative relationship with him for many years to come," Fukuda said.

On 25 October, Geng left KEK and reflected on the past three months. "I am very glad that I was able to study in such a splendid facility and with the researchers at KEK", he said. Shinichiro Michizono, who supervised Geng, said, "the controlling circuit which used FPGA will widely be used. This development will also be useful for LLRF for the ILC."

The leader of the ILC group at IHEP, Jie Gao, is very happy to have such a strong collaboration with KEK. "We are very satisfied with the actual IHEP-KEK collaboration on ILC by exchanging visiting researchers, especially young scientists like Geng in LLRF, and many others, such as Zhang Guo Zong in SC R&D, Ji Yuan Zhai in SC coupler, Xiao Ping Li in positron source, and Jun Hui Yue in instrumentation, who are actually at KEK."

-- Nobuko Kobayashi