# LLRF05

Europe/Zurich
CERN

Description

### Low Level Radio-frequency Workshop 2005

Sophisticated Low Level RF systems are needed in modern particle accelerators to deal with the characteristics of state of the art RF accelerating structures and their power sources, and to meet unprecedented levels of performance. The goal of the LLRF05 workshop is to share experience between linac and synchrotron projects (SNS, J-PARC, ILC, LHC etc.) and to discuss the best engineering practice.

This four day workshop will be the 15th in the series of mini-workshops under the auspices of the ICFA beam dynamics panel, and specifically will be the second in a series on low-level RF techniques, initiated at Jefferson Lab, USA in 2001.

Participants
• Alessandro Fabris
• Alessandro Ratti
• Alfred Blas
• Alireza Nassiri
• Alun Watkins
• Andree Buechner
• Andrew Butterworth
• ANDREW MOSS
• Andrew Young
• Angela Salom Sarasqueta
• Anton Rohlev
• Axel Winter
• Bernhard Zipfel
• Bo Hong
• Brian Chase
• CHRISTOPHE JOLY
• Claudio Rivetta
• Curt Hovater
• Damien TOURRES
• Dan Van Winkle
• Dariusz Makowski
• Dayle Kotturi
• Dmitry Teytelman
• Doug Horan
• Edmond Ciapala
• Elena Shaposhnikova
• Elmar Vogel
• Fernand RIBEIRO
• Flemming Pedersen
• Francis Perez
• Frank Ludwig
• Fu-Tsai Chung
• Fulvio Schiappelli
• Fumihiko Tamura
• Guenter Moeller
• Guido Koch
• Harald Klingbeil
• Hengjie Ma
• Henning-Christof Weddig
• herve lebbolo
• Holger Schlarb
• Hooman Hassanzadegan
• J. Michael Brennan
• James Rose
• Jaques Cherix
• Jaroslaw Szewinski
• Jean-Francois Genat
• Jean-Luc BIARROTTE
• Jean-Luc Vallet
• Joachim Tuckmantel
• John Dusatko
• John Fox
• John Molendijk
• John Musson
• Josef Frisch
• Joseph Orrett
• Jörn JACOB
• Karol Perkuszewski
• Kazunori AKAI
• Kenneth Fong
• Kevin S. Smith
• Kirk Davis
• Krzysztof Czuba
• Krzysztof Pozniak
• Larry Doolittle
• Lidia Ghilardi
• lili Ma
• Makoto Tobiyama
• Manuel Brönnimann
• MArco DI GIACOMO
• Marco Mauri
• Maria Elena Angoletta
• Mark Champion
• Mark Crofford
• Mark Prokop
• Markus Hoffmann
• Martin Kumm
• Massamba DIOP
• Matthias Hoffmann
• Maurice Piller
• Michael Laverty
• Michel Luong
• Olivier Le Dortz
• Olivier Piquet
• Patricia Shinnie
• Paul Joireman
• Philippe Baudrenghien
• Piotr Pucyk
• Rajesh SREEDHARAN
• Richard Abbott
• Rocco Paparella
• Roger Kaplan
• Roland Garoby
• Ryszard Romaniuk
• Sergey Belomestnykh
• Sergey Ivanov
• Shinichiro Michizono
• Stefan Simrock
• Stephen Smith
• Steve Myers
• Steve Herb
• Steven Hancock
• Sungil Kwon
• Thomas Bohl
• Thomas Schilcher
• Tom Hayes
• Tom Powers
• Tomasz Czarski
• Tomasz Jezynski
• Toshihiro MATSUMOTO
• Trevor Linnecar
• Uros Mavric
• Valeri Ayvazyan
• Valeri Lebedev
• Volker Schlott
• Wai Lau
• Waldemar Koprek
• Wojciech Cichalewski
• Wojciech Giergusiewicz
• Wojciech Jalmuzna
• Wolfgang Hofle
• Yeh Meng-Shu
• YOLANDA GOMEZ MARTINEZ
• Yoon Kang
• Yubin Zhao
Support
• Sunday, 9 October
• 18:30 20:00
Welcome drink and Registration 1h 30m Glass Box in Restaurant 1, Bldg. 501 (Restaurant 1)

### Glass Box in Restaurant 1, Bldg. 501

#### Restaurant 1

Bldg. 501
• Monday, 10 October
• 08:15 08:45
Registration 30m Bldg. 40, Central Area

### Bldg. 40, Central Area

#### CERN

• 08:45 10:00
Opening Session Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Dr. Thomas Bohl

Convener: Dr Trevor Linnecar (CERN)
• 08:45
Administrative details 5m
Speaker: Dr Trevor Linnecar (CERN)
• 08:50
LLRF Workshop Introduction 10m
Introduction talk
Speaker: Dr Steve Myers (CERN)
• 09:00
Workshop to Workshop: Four Years of Low Level RF Development* 30m
The first LLRF workshop was held at Jefferson Lab in April 2001. Since then, many of the questions posed at that workshop have been answered while new issues have arisen. Two topics that resonated at that the first workshop were issues of Analog vs. Digital, and Self Excited Loop (SEL) control vs. Generator Driven Resonator (GDR) control. A number of talks were given on both subjects, and useful, if not passionate, discussions ensued. Digital LLRF has matured since the first workshop, and with the increasing number of installed digital LLRF systems it has become the obvious choice for moderate to large accelerators. In the debate between SEL and GDR control, the consensus was that the choice is application driven. Each has specific advantages depending on the application. In the past four years a number of LLRF systems have been installed, and new accelerators have been proposed that raise new challenges for the LLRF community. This talk will give a review of the first workshop, discuss some of the work that has gone on since that time, and look at the issues that we are concerned about today. *This work was supported by DOE contract No. DE-AC05-84ER40150.
Speaker: Mr Curt Hovater (Jefferson Lab)
• 09:30
KEKB RF System 30m
The KEK B-Factory (KEKB) is a high-luminosity asymmetric energy electron-positron collider to support physics research programs on CP-violation and other topics in B- meson decays. The RF system for KEKB was designed to cope with difficulties arising from high current stored beam. It has two types of innovative heavily damped cavities for stabilizing coupled-bunch instabilities due to higher-order modes and the accelerating mode. Several feedback loops are implemented including the direct RF feedback and the zero and -1 mode longitudinal coupled-bunch oscillation dampers. To protect the Belle detector and beam-line hardware components from unstable beams caused by RF trips, a fast beam abort triggering and monitoring system has been developed. New RF stations are being constructed for the crab cavities to be installed next year for further increase of the luminosity. We describe the design features, operating status and future plans of the KEKB RF system.
Speaker: Prof. Kazunori Akai (KEK)
• 10:00 10:30
Coffee break 30m
• 10:00 10:30
Poster Session Preparation: Posters on Display on Tuesday and Wednesday Bldg. 40, Room 40-S2-C01

### Bldg. 40, Room 40-S2-C01

#### CERN

Posters should be put up during the Monday coffee break. Help and sticky tape will be provided. They will remain in place on Tuesday and Wednesday and can be viewed on these days except during the working group sessions. Authors are expected to be present in front of their posters for one hour after the first part of Talks Session 2 (i.e Tuesday, 09h50 to 10h50), to discuss and explain their work. At other times registrants are invited to contact the authors directly. Posters should be taken down on Wednesday evening or at the latest on Thursday morning.

• 10:30 13:00
Opening Session Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Dr. Thomas Bohl

Convener: Dr Trevor Linnecar (CERN)
• 10:30
Bunch by bunch feedback systems for KEKB 20m
Transverse bunch-by-bunch feedback systems for curing coupled-bunch instabilities have been working well since the early stages of the commissioning of the rings. To meet requirments of 90 degrees of phase shift, suppression of the static component and adjustable digital delay, a high-speed digital filter system with two-tap FIR functionality has been developed. Beam diagnostic systems which are part of the bunch feedback systems, such as bunch current monitors, betatron tune measurement systems, and bunch oscillation recorders, have been playing important roles in the tuning and understanding of the rings, enabling stable high-luminosity operation. The performance of these systems is reported here. Recently, a decrease of the luminosity with increase of the vertical feedback gain has been found. Possible explanations, including the stability of the feedback systems and the effect of residual noise, are also discussed.
Speaker: Dr Makoto Tobiyama (KEK Accelerator Laboratory)
• 10:50
LLRF developments at CERN 30m
Speaker: Dr Flemming Pedersen (CERN)
• 11:20
Operation and performanc of PEPII LLRF 20m
The PEP-II Low Level RF systems incorporate multiple feed back loops to control the cavity impedance as well as maintain the operating point regulation. The system incorporates several recording and playback buffers which can be used to both setup and monitor the operating point of the system. In addition, the recording buffers aid in the diagnostics of hardware failures within the system and are recorded at every beam abort. An overview of the PEP-II LLRF system will be discussed. In addition, in depth information regarding both the operational and diagnostic modes of the system will be presented. Examples of both modes of operation will illuminate some of the complexities of operating a large complicated LLRF system as well as demonstrate techniques to solve problems to root cause.
Speaker: Mr Dan Van Winkle (SLAC)
• 11:40
Klystron Linearizers for PEP-II 20m
The RF systems in PEP-II use direct and comb loop feedback techniques to minimize the cavity fundamental impedance driving low-mode coupled bunch instabilities. The effectiveness of these techniques are strongly dependent on the linearity and dynamic behavior of the klystron amplifier in the feedback path. This short talk will summarize the impedance control techniques in PEP-II, and review measurements which highlight the impact of klystron saturation on low mode instabilities. Technical options to linearize the klystron and our specific implementation are described. Results from full power test stand measurements and beam tests are presented.
Speaker: Dr John Fox (Stanford Linear Accelerator Center)
• 12:00
SNS Linac System 30m
The Spallation Neutron Source (SNS) project, which is scheduled for completion in Spring 2006, will utilize 100 RF systems for acceleration and bunching of the H- beam. Ninety-six of these systems drive the Linac, which was successfully commissioned with beam in Aug-Sep 2005. The remaining four RF systems are presently being tested in preparation for the Ring commissioning run, scheduled for Winter 2005-2006. In this paper the SNS RF systems will be described along with some operational results and plans for a power upgrade project, which will require an additional 38 RF stations.
Speaker: Mark Champion (Oak Ridge National Laboratory / Spallation Neutron Source)
• 12:30
LLRF Developments at the BNL Collider-Accelerator Department 30m
The RF group within the BNL Collider-Accelerator Department (BNL C-AD) supports the development, operations and maintenance of RF systems for three operational synchrotrons (AGS Booster, AGS and RHIC), and is also responsible for RF system development for a number of upgrades and projects both planned or in progress. Active projects include the C-AD LLRF Upgrade (upgrade of AGS Booster, AGS and RHIC LLRF systems), the Electron Beam Ion Source (EBIS) project (a high charge state heavy ion injector for the AGS Booster), and the Super Conducting Energy Recovery Linac (SC-ERL) experiment (a very high current SC-ERL prototype and SC RF photo- cathode gun). Planned projects include the Rare Symmetry Violating Processes (RSVP) upgrade (1E14 protons per pulse, 2.5Hz AGS upgrade), and a possible Super- Neutrino Driver (high current SC proton linac for AGS injection). With very limited LLRF engineering resources available to support all these efforts, the decision was made to focus development efforts at designing a common digital hardware platform, modular in nature, which could then be configured as necessary to satisfy the unique requirements of the various LLRF systems. We believe such a system offers a number of significant benefits, including: maximum leveraging of very limited engineering resources; reduction in the number of unique hardware components we need to produce, operate and maintain; flexibility in individual LLRF system configuration; flexibility of future upgrade options; and, greatly reduced risk from component obsolescence. Our talk and poster will describe the architecture as currently envisioned, lessons learned from the RHIC, Booster and SNS Ring LLRF designs, the various LLRF system requirements, and the factors effecting design decisions and tradeoffs.
Speaker: Mr Kevin S. Smith (BNL)
• 13:00 14:00
Lunch 1h
• 14:00 15:30
Talks Session 1 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Dr. Steve Hancock

Convener: Dr Larry Doolittle (LBNL)
• 14:00
Cornell Digital LLRF System 30m
A new digital LLRF system has been developed recently at Cornell University. The system is suitable for a wide variety of accelerator applications as it provides great flexibility, high computational power and low latency. It features very fast feedback and feed-forward controls, a state machine for automatic start-up, calibration and trip recovery, CW and pulsed modes of operation, quench detection, and cavity frequency control. The system uses in-house developed digital (based on FPGA and DSP) and RF hardware. The first generation system was installed, commissioned and has been in operation at the CESR storage ring for more than a year. The achieved field stability surpasses requirements. The other future application of this system is in the Energy-Recovery Linac presently under development at Cornell. The superconducting cavities in the ERL will operate at a high loaded Q factor and will be required to satisfy very stringent requirements on the cavity field amplitude and phase stability. To prove that the RF field in a high loaded Q cavity can be stabilized, and that Cornell's newly developed digital control system is able to achieve this, the system was connected to a high-loaded-Q cavity at the JLab IR-FEL. Excellent cw field stability was achieved. The operational experience with the digital LLRF system in CESR and results of experiments at JLab are presented. The second generation of the digital LLRF system is under development. It has somewhat different architecture to satisfy requirements of the Cornell ERL. The ERL LLRF system is briefly discussed.
Speaker: Dr Serguei Belomestnykh (Cornell)
• 14:30
Digital LLRF feedback control system for the J-PARC linac 20m
Twenty high power klystrons will be installed in the J-PARC linac. The rf fields are required to be stable less than +-1% in amplitude and +-1deg. in phase. The digital feedback (FB) system using FPGAs are adopted so as to satisfy these requirements. External monitors using wave-detectors and mixers are utilized for confirming the stability. The measured stabilities are less than +-0.15% in amplitude and +-0.15 deg. in phase during the 500s flat-top. A tuner control system was tested and 18- hour FB operation was also carried out.
Speaker: Dr Shinichiro Michizono (High Energy Accelerator Research Organization (KEK))
• 14:50
JLAB LLRF System 20m
Speaker: Tom Powers (Jefferson Laboratory)
• 15:10
Modularised Low Level RF for a L-band Klystron test stand at SLAC 20m
Speaker: Andrew Young (SLAC)
• 15:30 16:00
Tea break 30m
• 16:00 17:25
Talks Session 1 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Dr. Steve Hancock

Convener: Dr Larry Doolittle (LBL)
• 16:00
Fermilab LLRF Software: Architecture and Development 20m
Fermilab’s Main Injector, Recycler and Tevatron synchrotrons Low Level RF systems perform a wide variety of tasks in support of accelerator goals, colliding beam HEP and neutrino production. In this paper, we will focus on the architecture and features of the LLRF software, the development practices we use and the lessons we have learned in the process. These specific features are presented within the framework of a standard control model. The LLRF systems are called upon to perform a variety of different tasks including data acquisition, beam acceleration, machine to machine beam transfer, feedback control, data transfer, and user interface. To do this, we implement separate processes that run with different periods, ranging from a few microseconds to minutes. These systems must also implement a flexible toolset that allows an end-user to reconfigure the system, combine the tools in unforeseen and useful ways, and request new features. The systems must be highly reliable, and both fault detecting and fault tolerant. In addition, changes to any system must minimally affect beam operation of all systems. In order to meet these requirements we use a variety of techniques. We employ a set of size C VXI modules built in-house from a standard base design, which includes, SHARC DSP, ALTERA FPGA, and VXI interface. The remainder of the module is customized to perform specific functions, such as data acquisition, frequency control, or RF switching. The common DSP platform enables significant savings in software design and implementation by allowing us to develop, test and deploy a single solution to a common system problem across multiple VXI modules. In order to meet system specific goals, such as acceleration or beam transfer, cooperation between modules is required. We implement an object-based methodology which provides a common interface to data and services provided by a VXI module. System level software employs these module objects by requesting services to fulfill system tasks and exposing module specific data to the accelerator control system. The system level software also implements the interface to support user requests for system reconfiguration. The interface allows the user to reconfigure a single parameter, such as feedback loop gains, or the user can specify a specific timed scenario of operations for the system to perform. Error logging and reporting are tightly integrated into the system software to enable debugging and fault diagnosis.
Speaker: Dr Paul Joireman (Fermilab)
• 16:20
Beam based feedback for control 20m
Speaker: Dr Holger Schlarb (DESY)
• 16:40
Tutorial on Optimal Controller 45m
The designer of an RF control system will be aiming for the best possible control of the cavity fields that is possible. Optimal control deals with the problem of finding a control law for a given system such that a certain optimality criterion is achieved. In control theory the control that minimizes a certain cost functional is called the optimal control. The problem formulation usually also contains constraints. For rf control the goal is usually to regulate the accelerating field to the required stability (or better if possible) with the available rf power. Contraints usually include robustness against parameter variations, rf power limititations, minimizing the trip rate (i.e. maximum availability of the accelerator), maximizing the lifetime of the components, fast recovery from faults, and other operational aspects. Important for the quality of field control are also the cavity field detectors, the beam diagnostics, the actuators available for control, the perturbations to be controlled and the choice of the controller. The controller usually applies fast control to the klystron drive (low level rf) and slow and fast control to the cavity frequency tuners. Design choices for fast rf control include amplitude and phase versus IQ-control, analog versus digital, self-excited loop versus generator driven and individual cavity control versus control of the vector-sum. A combination of feedforward for repetitive errors and feedback for stochastics errors is used. Slow drifts are often corrected by beam based rf feedback. The tutorial will present the definition of optimality in control theory and considerations for optimal control for accelerators. Design choices will be discussed and compared. A concept for the optimal controller will be developed.
Speaker: Dr Stefan Simrock (DESY)
• Tuesday, 11 October
• 08:15 08:30
Registration 15m Bldg. 40, Central Area

### Bldg. 40, Central Area

#### CERN

• 08:30 09:50
Talks Session 2 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Dr. Elena Shaposhnikova

Convener: Mr Curt Hovater (JLAB)
• 08:30
The LHC Low Level RF 20m
The LHC RF includes eight 400 MHz super conducting cavities per ring. Each cavity is independently powered by a 300 kW klystron via a circulator. The challenges are: very high beam current (more than 1A RF component) and very low RF noise (emittance growth time in excess of 25 hours). To achieve that, the Low-Level RF comprises the following sub-systems: • We have one Cavity Controller per cavity. It is meant to provide adequate control of the voltage seen by the beam and to keep the power demanded at acceptable levels. It includes a Klystron Polar Loop (that keeps the gain and phase constant from the RF modulator input to the cavity main coupler input), an RF Feedback Loop (that reduces the effects of the cavity impedance) and a Tuner Loop (that maintains the cavity at a tune that minimizes the power transients due to the passage of batches and gaps). • We have one Beam Control per ring. It includes a Phase Loop and a Synchronization Loop that locks the two beams to a common reference during the acceleration ramp. This loop can be replaced by a Radial Loop for commissioning and machine developments (de-synchronization of the two rings). • The RF Synchronization implements the bunch into bucket transfer from the SPS into each LHC ring. • Finally a Longitudinal Damper (one per ring) is planned to reduce emittance blow-up due to filamentation following phase and energy errors at injection. At start-up it will act via the main 400 MHz cavities. It tries to damp both dipole and quadrupole modes.
Speaker: Dr Philippe Baudrenghien (CERN)
• 08:50
CERN LEIR LLRF 20m
From 2008 onwards the LHC physics programme will include, in addition to proton- proton collisions, running periods with heavy-ion collisions. The injector complex has to be adapted in order to satisfy LHC ion requirements. In particular, a Low Energy Ion Ring (LEIR) has been built at CERN and will be commissioned by early/mid 2006. LEIR will be equipped with a new digital beam control and cavity servoing system, based on VME modules and daughtercards carrying Analog Devices floating point DSPs and Altera FPGAs. The system will include, among others, frequency program, phase and radial loop capabilities, vector sum and dual harmonic cavity servoing. Several features will be implemented by software, e.g. phase rotation and signals delay, timing events and reference functions generation. Finally, full digital diagnostics information will be provided. This system will also serve as a pilot project for its migration to the other accelerators of the PS complex, such as the Antiproton Decelerator, the PS Booster and PS. This talk outlines the system capabilities and the new schemes used for their implementation. Some test results obtained by a scaled-down version of the system deployed in the PS Booster will also be given.
Speaker: Maria Elena Angoletta
• 09:10
CERN LINAC LLRF 20m
A VME based control system has been developed and built at CERN for the servo loops regulating the field in linac accelerating structures. It is an all-digital system built on a single VME card, providing digital detection, processing, and modulation. It is foreseen to be used, in different versions, for the needs of both present and future CERN hadron linacs and is currently being used in the energy ramping RF chain of the CERN Heavy Ion Linac (Linac 3). The design principle and the experimental results are described.
Speaker: Tony Rohlev (CERN)
• 09:30
Femtosecond Stable Timing and Synchronisation Systems 20m
Speaker: Dr Volker Schlott (PSI)
• 09:50 10:20
Coffee break 30m
• 09:50 11:20
Poster Session with Author Participation: Poster Discussion

Authors of posters are expected to be available next to their posters during this time to discuss and explain their work.

• 09:50
LCLS LLRF System 3m
The Linac Coherent Light Source (LCLS) project [1] at SLAC uses a dense 15 GeV electron beam passing through a long undulator to generate extremely bright x-rays at 1.5 angstroms. The project requires electron bunches with a nominal peak current of 3.5kA and bunch lengths of 0.020mm (70fs). The RF stability required by the bunch compressors is tighter than what is currently required to run experiments. This paper discribes the upgrades to the RF monitoring and control system required to meet the 100fS phase stability requirements.
Speaker: Dayle Kotturi (SLAC)
• 09:53
A Cavity Synchronization System For Heavy Ion Synchrotrons Based on DSP, DDS and FPGA Technology 3m
A cavity synchronization system has been realized which allows the synchronization of the gap signals of different cavities. The system is designed in such a way that the cavities may run at different harmonics. In future it will also be possible to synchronize the cavity with the beam, i.e. to realize closed-loop beam phase control. In order to fulfill these requirements, the overall system is based on different scalable modules which can flexibly be used in completely different applications. The key modules are the following ones: * DSP System: This subsystem is used for high-precision phase and amplitude detection and fast closed-loop control algorithms. It includes analog preprocessing in the IF range, ADC and DAC modules, suitable digital interfaces and comfortable diagnostics features. * FPGA Interface Board (FIB): This module provides an interface to the central control system and several other interfaces which are used as a standard in the GSI synchrotron RF group. The interface protocols are realized by an FPGA which allows one to implement routing, control and signal processing applications. * DDS unit: This module is used for the generation of RF master signals and can also be used as actuator in a closed loop RF control. Furthermore, it will be used in future to realize the functionality which is currently performed by an analog offset LO (local oscillator). The DDS module is realized by combining an FIB with a dedicated piggyback PCB. These key modules are described here. Furthermore, the status of the following applications is reported: * Digital cavity synchronization for single-harmonic operation * Dual harmonic operation Finally, an outlook to other applications is given.
Speaker: Dr Harald Klingbeil (GSI RF Department)
• 09:56
RF System Modeling and Simulation for the SOLEIL Synchrotron 3m
An analysis of beam stability in the SOLEIL synchrotron with two different basic systems (direct and amplitude/phase feedback) was carried out during the preliminary design phase in 1999. Since then, on the one hand, the beam energy was pushed from 2.5 GeV to 2.75 GeV, which led to a change of several other machine parameters such as the harmonic number, accelerating voltage, relative beam loading factor and external coupling factor; on the other hand, an analog LLRF system combining one fast direct feedback and one slow amplitude/phase feedback was approved for the machine commissioning. Therefore, a new simulation for the optimization of the LLRF system parameters appeared necessary. It additionally takes into account different features (loop delays, bandwidth limitation, extra power budget, possible implementation of a comb-filter, etc.), which were ignored in the preliminary analysis. A comparison with a fast digital I/Q LLRF system, currently under development, is also presented with a Matlab and Simulink based simulation tool, which is more versatile than the formerly used Fortran based code.
Speakers: Mr Massamba Diop (Synchrotron SOLEIL) , Mr Michel Luong (CEA Saclay)
• 09:59
LLRF system within the framework of the R&D of supraconducting SPOKE cavities 3m
Within the framework of the current European research programs EUROTRANS and EURISOL on High Intensity Proton Accelerators, and particularly on the R&D on superconducting SPOKE cavities, a Low Level Radio Frequency Digital system is developed at IPN Orsay in collaboration with LPNHE Paris, both IN2P3-CNRS laboratories. Due to Lorentz's forces, mechanical vibrations or RF power perturbations, the amplitude and phase of the electromagnetic wave inside the cavities need to be controlled. Other goals are a better reliability, a high level of integration and a fast response time of the feedback control system. Digital techniques should allow to meet all of these goals and provide an improved flexibility compared to analog techniques, with the integration of the main algorithms and functions into an FPGA. The main design options and some preliminary results are presented.
Speakers: Mr CHRISTOPHE JOLY (CNRS/IN2P3 IPN Orsay) , Mr HERVE LEBBOLO (CNRS/IN2P3 LPHNE Paris) , Mr JEAN-FRANCOIS GENAT (CNRS/IN2P3 LPHNE Paris) , Mr JEAN-LUC BIARROTTE (CNRS/IN2P3 IPN Orsay) , Mr OLIVIER LE DORTZ (CNRS/IN2P3 LPHNE Paris)
• 10:02
The Low Level Radio Frequency system for the High Intensity Proton Injector (IPHI) 3m
Within the framework of European research programs on High Intensity Proton Accelerator, IPHI (High Intensity Proton Injector) is a 3MeV 100mA CW proton injector prototype. It consists of the SILHI ECR source (100keV), a 3MeV Radio Frequency Quadrupole (RFQ) and a high energy beam line for beam quality analysis. Two 352MHz klystrons inject 1200kW RF power into the RFQ. The Low Level Radio Frequency system (LLRF) controls the amplitude and phase of the accelerating voltage inside the RFQ for CW or pulsed mode operation, and also the RFQ resonant frequency. It consists of seven feedback analog loops controled by PXI acquisition cards, a distributed I/O system based on FieldPoint modules and an Ethernet communication with the supervision system EPICS. RF power network and LLRF system with some results are presented
Speaker: Mr CHRISTOPHE JOLY (IN2P3-CNRS IPNO Orsay)
• 10:05
Low Level RF System in KEK-STF 3m
At the electron linac of KEK-STF (Superconducting RF Test Facility), an accelerating electric field of ±0.1% in amplitude and ±0.1degree in phase is required for Low-Level RF (LLRF) system. Digital feedback (FB) system is adopted for flexibility of the FB and feedforward (FF) algorism implementation to accomplish these requirements. In order to carry out the efficient testing of the control system, rf system modelling with MATLAB/Simulink library is utilized for the investigation of the control method and cavity simulator using a FPGA board has been developed.
Speaker: Mr Toshihiro MATSUMOTO (KEK)
• 10:08
CONTROL AND LOW LEVEL RF SYSTEM OF THE SOLEIL SYNCHROTRON 3m
In the SOLEIL storage ring, two cryomodules, each containing a pair of 352 MHz superconducting cavities, will provide the maximum power of 600 kW, required at the nominal energy of 2.75 GeV with the full beam current of 500 mA and all the insertion devices. They will be supplied with liquid helium from a single cryogenic plant and each of the four cavities will be powered with a 190 kW solid state amplifier consisting in a combination of 315 W elementary modules (about 750 modules per amplifier). The low electronic system that will be used in the first phase consists in “slow” amplitude, phase and frequency loops, complemented with a direct RF feedback. A fast digital, FPGA-based, I/Q feedback is currently under development, that should be implemented later on. The control of the whole system is insured by several PLCs and a µcontroller, which monitors the amplifier parameters through a multiplexing system. The PLCs are linked to the SOLEIL TANGO framework via Ethernet. The control and low level RF system is described and the first operational/experimental results are reported in this paper.
Speaker: Mr Fernand RIBEIRO (SYNCHROTRON-SOLEIL)
• 10:11
Evaluation of Libera as field control module 3m
Libera is a product family targeting instrumentation and controls applications on particle accelerators. So far three members have been introduced and very well accepted by the accelerator community. Libera's hardware architecture presents a universal platform that has all the hardware interfaces to convert signals from analog to digital and vice versa. In between there is a big FPGA that offers abundant computing power for loop control. This article presents a possible application of Libera as a field control module in a LLRF control system. It first describes in details the main hardware building blocks. The article continues with a description of simulations and discussions of results of a mathematical model of a feedback control system comprising of a basic klystron, RF cavity, cable of certain length and field control module. Conclusion discusses the applicability of Libera as a LLRF field control module.
Speaker: Mr Uros Mavric (Instrumentation Technologies)
• 10:14
Analysis of a digital beam phase control system 3m
At GSI a closed loop beam phase control is planned, which will be used to damp coherent dipole oscillations of particle bunches. The system is based on a DSP System for high-precision phase and amplitude detection, which was also developed at GSI based on commercial DSP, ADC and DAC modules and is also applied for cavity synchronisation. A special digital filter with variable pass band is used to convert the beam phase signal to an adequate correction signal for the accelerating RF-voltage. In order to get a proper correction signal the digital filter eliminates noise but it amplifies phase oscillation signals in the matched bandwidth of the synchrotron frequency of the individual accelerated particles. The filter blocks slow variations of the beam phase to allow changes of the synchronous phase without fixing it to a predefined value. Anyhow the implementation of this digital filter seems to be straightforward and the realisation does not spend much computational power. Taking into account limited computing power and given signal processing delays as well as noise on analogue cables the loop stability limit is determined by simulation. Influence on beam loss and longitudinal phase space plots can be given. In future the beam phase control will reduce emittance blow up during accelerating in SIS18/12.
Speaker: Dr Bernhard Zipfel (FH-Fulda / GSI)
• 10:17
FPGA Phase Detector Implementation 3m
An optimized digital phase detector for a 106 MHz superconducting cavity was designed using state machine logic. Transitions of the feedback and reference inputs trigger corresponding changes in the state of the detector. The state machine has been implemented using a high speed Xilinx field programmable gate array. The resulting design incorporates two phase detectors (one for the phase loop and one for the tuning loop), two frequency counters, and a subractor. The counters and subtractor are used to determine the frequency error for initial tuning of the cavity. The design and some initial test results are presented.
Speaker: Mr Michael Laverty (TRIUMF)
• 10:20
Multiplexed Pulsed RF Reference/Cavity Signals at the Spallation Neutron Source 3m
The Spallation Neutron Source (SNS) low-level RF (LLRF) control system currently utilizes separate channels to process 50 MHz down-converted pulsed Cavity IF and CW Reference IF signals. A new scheme creates pulsed Reference RF signals which are multiplexed with pulsed Cavity RF signals near the front-end and linear accelerator (linac) RF cavity structures. The multiplexed pulsed RF Reference/Cavity signals are transported to each LLRF control system via a coaxial cable and processed in a single channel. Implementation details and initial results are discussed.
Speaker: Mr Maurice Piller (Spallation Neutron Source)
• 10:23
SIMCON hardware, from 2.0 to 3.1 3m
Hardware development status for DESY UV-FEL were presented. The SIMCON stands for the microwave, resonant, superconductive accelerator cavity simulator and controller (embracing the hardware and software layers). The current version of the SIMCON is 3.1 which is a considerable step forward from the previous 8-channel version 3.0 which was released at the beginning of 2005 and was made operable in April. The following main differences were implemented in SIMCON 3.1 It is a stand alone VME board (whereas SIMCON 3.0 was modular). It has IP and multiple gigabit optical I/O. It has 10 ADC channels. It bases on the Virtex II Pro 30 chip with two embedded Power PCs and with the DSP blocks. The architecture idea and block diagrams of the PCB for SIMCON 3.1 were presented.
Speaker: Mr Wojciech Giergusiewicz (Warsaw University of Technology faculty of Electronics and Information Technology)
• 10:26
The on-line nonliearities measurements of the VUV-FEL accelerator RF- stations high power chain. 3m
As every high power amplifier also the pulse microwave 10 MW klystrons (in the VUV- FEL accelerator) have nonlinear output power vs. input power characteristic near to the device saturation point. This undesirable behavior of the tube amplifier may cause lower efficiency of the close loop RF field regulation in LLRF control system. In order to provide a solution for existing distortion compensation it is necessary to examine each of the RF-power station. The main goal is then designing and establishing an on-line measurement system and appropriate procedures that would give comprehensive knowledge about distortions introduced by the klystron and power preamplifiers and will be transparent for regular accelerator operation. In this paper the problems concerning appropriate amplifiers characterization as well as propositions for on-line nonlinearities monitoring will be discussed. Also the currently used solutions for the high power amplifiers linearisation will be presented and a proposition for VUV-FEL RF-station high power chains distortion compensation will be described.
Speaker: Mr Wojciech Cichalewski (Technical University of Lodz)
• 10:32
DECOUPLING PI CONTROLLER DESIGN FOR A NORMAL CONDUCTING RF CAVITY USING A RECURSIVE LEVENBERG-MARQUARDT ALGORITHM 3m
This paper addresses the system identification and the decoupling PI controller design for a normal conducting RF cavity. Based on the open loop measurement data of an SNS DTL cavity, the open loop system’s bandwidths and loop time delays are estimated by using batched least square. With the identified system, a PI controller is designed in such a way that it suppresses the time varying klystron droop and decouples the In-phase and Quadrature of the cavity field. The Levenberg- Marquardt algorithm is applied for nonlinear least squares to obtain the optimal PI controller parameters.
Speaker: Mr Sungil Kwon (LANL)
• 10:35
ON-LINE GAMMA RADIATION AND NEUTRON FLUENCE MONITORING 3m
Radiation present in an accelerator poses a real threat to electronic devices and systems placed in the main tunnel. Radiation in the accelerator tunnel is produced as a result of the electron beam's interaction with high-Z materials. Because of the Total Ionizing Dose effect gamma radiation is responsible for a long term degradation of all devices installed in the accelerator's chamber. Neutrons are responsible for Single Event Effects that generate malfunctions in digital systems and they could result in repairable or hard damage of the devices. Therefore gamma radiation and neutrons monitoring is strictly recommended to avoid the unwanted breakdown of the control system. The paper presents the system dedicated for gamma radiation and neutrons monitoring in a linear accelerator in real-time. The presented detector is intended to be used in the newly installed X-Ray Free-Electron Laser X-FEL at DESY research centre in Hamburg. Two different detectors were used to monitor gamma radiation and neutrons. The radiation sensitive transistor RADFET is responsible for gamma radiation measurement whereas SRAM-based detector was used to measure neutron fluence in the tunnel. The radiation-selective sensors are connected to the microcontroller-based read-out system. The system was built with the usage of redundant elements to assure radiation tolerance. Measured data are gathered in a database, thus gamma radiation and neutron fluence history is accessible. These parameters help to predict damage of electronic systems that are placed in the tunnel. We have conducted a few experiments with the system at DESY. The devices were exposed to a neutron field from an Americium-Beryllium neutron source 241AmBe. The systems were installed in two accelerators: VUV-FEL as a prototype of X-FEL accelerator and Linac II. The results of the operation within a few months are discussed in the paper.
Speaker: Mr Dariusz MAKOWSKI (Department of Microelectronics and Computer Science, Technical University)
• 10:38
A Review of the Advanced Photon Source 350MHz Low-Level RF Systems 3m
The status and performance of the 350MHz low-level rf systems and hardware used at the Advanced Photon Source will be presented. Technical descriptions of how this hardware is utilized to provide rf control include rf source generation and distribution, cavity resonance, gap voltage amplitude and phase control, and parallel-klystron operation.
Speaker: Mr Doug Horan (Argonne National Laboratory)
• 11:20 12:55
Talks Session 2 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Dr. Elena Shaposhnikova

Convener: Mr Curt Hovater (JLAB)
• 11:20
Phase noise measurements in TRIUMF ISAC 2 cryomodule 20m
The phase noise spectra of the superconducting cavities of the ISAC 2 cryomodule and of the test cryostat were measured under various conditions. Analysis of the results show that the major contributors to this phase noise are microphonics, including those from the superconducting cavities, the cryomodule and the power amplifiers, each of which can be identified with their unique signatures. The results also show the effectiveness of the phase stabilizing loop in suppressing these noise components. These measurements have become a useful diagnostic tool to identified the various noise sources and verify the effectiveness of their suppression.
Speaker: Dr Kenneth Fong (TRIUMF)
• 11:40
Noise Effects on Stored Beams in a Proton Synchrotron 20m
The presentation will start with a few general remarks on a theory of beam motion under noises (inside and beyond RF buckets, non-linearity, stationary and periodically non-stationary noises, drift fluxes, etc). Existing difficulties would be outlined and ways to solve them sketched out. Availability of ready-to-use proven techniques and tools will be commented. Experimental verification of the key noise diffusion issues in the MD runs of U70 PS of IHEP-Protvino (flattening bunches with RF noise gymnastics, stochastic slow extraction) will be reviewed in short. Major emphasis will be put on the problems most applicable to the LHC: specifics of beam noise-driven halation under a non-negligible energy loss due to synchrotron radiation of high-energy protons (diffusion plus or minus drift due to the SR). An attempt to explain (most of) coasting-beam halo phenomenology in the HERA-p ring will be made and projected onto the LHC case.
Speaker: Dr Sergey Ivanov (Institute for High Energy Physics (IHEP))
• 12:00
Bunch Lengthening in Tevatron due to RF noise 20m
If not suppressed the RF system noise can significantly affect the luminosity of hadron collider. It causes the bunch lengthening with consecutive decrease of the beam intensity. Presently, the RF noise is the second leading reason of bunch lengthening in Tevatron yielding to the intrabeam scattering at the store beginning. The noise is mainly created by microphonics in RF cavities and is strongly suppressed by local feedback system. The report discusses the beam-based and direct measurements of the RF noise in Tevatron. The beam measurements are based on observing particle diffusion in satellite bunches to the center of the distribution. Such choice allows one to reduce effects of the intrabeam scattering and the beam-beam-effects. The measurements yielded the spectral density of RF phase of about 4e-11 rad^2/Hz. The report also discusses the theory of bunch lengthening required to compare the direct and beam-based measurements.
Speaker: Mr Valeri Lebedev (FNAL)
• 12:20
The LCLS LLRF Control System 20m
The LINAC Coherent Light Source requires an RF stability of 0.1% rms in amplitude, 100 fs in phase for a 850 ns fill time for the S-band structure, 125 fs in phase for a 100 ns fill time for the X-band structure and 30 ns rise times for the RF gun pulse shaper. This paper describes the design of a new VME/EPICS-based control system for the LCLS low level RF system that will monitor and control the RF phase and amplitude so that these requirements can be met with the use of beam based feedback. The challenges with measurement of RF phase as seen by the beam are also discussed.
Speaker: Dayle Kotturi (Stanford Linear Accelerator Center)
• 13:00 14:00
Lunch 1h
• 14:00 15:30
Working Group 1: Synchrotrons/LHC: WG Session 1 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Convenor for this working group is Dr. Mike Brennan and the Scientific Secretary Dr. Philippe Baudrenghien.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Dr Mike Brennan (BNL)
• 14:00
Development of LLRF control system for J-PARC RCS 7m
The J-PARC Rapid cycling synchrotron (RCS) requires a very stable and precise low level RF (LLRF) control system to handle the ultra-high proton beam currents. The MA-loaded low-Q cavities for the RCS are driven by the superposition of the dual-harmonic RF signal for both the acceleration and the longitudinal bunch shaping. We employ a full-digital system based on direct digital synthesis (DDS). The multi-harmonic RF signals generated by the DDS are easily synchronized without PLLs. We describe the design and the structure of the LLRF blocks. The common feedback loops are used for the stabilizing the beam orbit and phase as well as the RF voltages of the cavity. The heavy beam loading effect is compensated by using the beam feedforward method. We also present the test results of the recently manufactured modules.
Speaker: Dr Fumihiko Tamura (JAREI)
• 14:14
Experiments of Low Level Control Based on Analog IQ Technique 7m
Abstract Low level control (LLC) system based on IQ technique will be constructed for RF cavities at Shanghai Synchrotron Radiation Facility (SSRF). There are two paths for LLC to go through. One is based on analog IQ, another adopts all digital technique. Some experiments of IQ control loop in our laboratory have been done, which shows that the suppression 20dB of modulation in amplitude and phase and the loop width over 1KHz are achieved. Key word: IQ technique Amplitude modulation phase modulation Loop width
Speaker: Mr Yubin Zhao (Shanghai Institute of Applied Physics.CAS)
• 14:21
LHC RF Noise Simulation 20m
Speaker: Joachim Tuckmantel (CERN)
• 14:00 15:30
Working Group 2: List future applications - state of development, review works in progress Bldg. 40, Room 40-5-A01

### Bldg. 40, Room 40-5-A01

#### CERN

The Convenor for this working group is Mr. Mark Champion and the Scientific Secretary Mr. Tony Rohlev.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Mr Mark Champion (ORNL/SNS)
• 14:00 15:30
Working Group 3: Session 1a: RF Modelling Bldg. 40, Room 40-4-C01

### Bldg. 40, Room 40-4-C01

#### CERN

The Convenor for this working group is Dr. Stefan Simrock and the Scientific Secretary Dr. Andy Butterworth.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations 39,42,50 will be presented in a joint session with working group 4 and are described there. Again the convenors will decide the timetable for this.

Convener: Dr Stefan Simrock (DESY)
• 14:00
Superconductive cavity driving with FPGA controller 7m
The LLRF – Low Level Radio Frequency cavity control system is still under development in order to regulate accelerating fields of the superconductive resonators. The first testing of the cavity driving applying the FPGA (Field Programmable Gate Array) technology have been carried out in DESY Hamburg. The experiments focused attention to the general recognition of the cavity features and projected control methods. The electrical model of the resonator is taken as a consideration origin. The calibration of the signal channel is considered as a key preparation for an efficient cavity driving. The identification of the resonator parameters is confirmed as a proper approach for the required performance: driving on resonance during filling and field stabilization during flattop time with reasonable power consumption. The feed-forward and feedback modes were applied successfully for the CHECHIA cavity and ACC1 module driving. Representative results of experiments are presented for different levels of the cavity field gradient.
Speaker: Mr Tomasz Czarski (Warsaw University of Technology Institute of Electronic Systems)
• 14:07
LANSCE-R Low Level RF Control System 7m
The Los Alamos Neutron Science Center (LANSCE) proton accelerator is scheduled for refurbishment starting in FY06. A new low level RF (LLRF) system is part of the refurbishment plan since the existing LLRF system is analog-based and requires significant setup and maintenance time. Both field and resonance control aspects of the current system do not have the flexibility to meet future performance requirements. The LANSCE accelerator provides both H+ and H- beams and due to the various user requirements there are a number of different beam pulse types varying in timing and current. In order to meet user needs, LANSCE must simultaneously transport both H+ and H- in the accelerator. These requirements have motivated the development of a new LLRF system based on soft radio technology. The new system will include field control using feedback and adaptive feed forward techniques, an upgraded resonance controller with frequency agility to improve startup and fault recovery times and a high power amplifier pre-compensation controller for improved cavity fill times and amplifier efficiency. Among the challenges with implementing the new system are interfacing with existing subsystems of the accelerator.
Speaker: Mark Prokop (Los Alamos National Laboratory)
• 14:14
VHDL programming and software for LLRF control system SIMCON and results from tests 7m
The talk presents firmware and software for new LLRF control system called SIMCON (SIMulator and CONtroller). The system is based on FPGA technology and the firmware is written in VHDL. The simulator of superconductive cavity and the controller of 1 and 8-cavity accelerating module was made in FPGA. The firmware and software was tested in CHECHIA and in ACC1 module in TTF2 in DESY with pretty results.
Speaker: Mr waldemar koprek (Warsaw University of Technology, DESY)
• 14:21
Dynamic Modeling and Simulation of LER-HER PEP II Rings 7m
This paper presents the dynamic model and simulation of the beam-cavity interactions in both LER and HER rings at PEP II. The motivation for development of this toll is to explore the stability margins and performance limits of PEP II LLRF systems at higher currents and possible upgraded RF configurations. The simulation is a time-domain model, so that nonlinear elements in the klystron and processing can be included. The ring current is represented by macro-bunches, with time structure consistent with the time resolutionof the simulation and necessary modal frequency resolution to explore coupled-bunch modes within the RF system bandwidth. The program has been validated with data collected from PEP II and SPEAR. Measured transfer functions of the cavity and control system are compared with the simulation results to show the agreement between both results. Measured grow rates for the beam in different conditions at PEP II and SPEAR rings are compared with the preliminary grow rate predictions using the simulation.
Speaker: Dr Claudio Rivetta (SLAC)
• 14:49
Digital Low Level RF Control System for the DESY TTF VUV-FEL Linac 7m
In the RF system for the Vacuum Ultraviolet Free Electron Laser (VUV-FEL) Linac each klystron supplies RF power to up to 32 cavities. The superconducting cavities are operated in pulsed mode and high accelerating gradients close to the performance limit. The RF control of the cavity fields to the level of 1e-4 for amplitude and 0.1 degree for phase however presents a significant technical challenge due to the narrow bandwidth of the cavities which results in high sensitivity to perturbations of the resonance frequency by mechanical vibrations (microphonics) and Lorenz force detuning. The VUV-FEL Linac RF control system employs a completely digital feedback system to provide flexibility in the control algorithms, precise calibration of the accelerating field vector-sum, and extensive diagnostics and exception handling capabilities. The RF control algorithm is implemented in DSP firmware and DOOCS (Distributed Object Oriented Control System) servers. The RF control system design objectives are discussed. Hardware and software design of the DSP based RF control are presented.
Speakers: Dr Stefan Simrock (DESY) , Dr Valeri Ayvazyan (DESY)
• 14:00 14:45
Working Group 4: Platforms (L. Doolittle) Bldg. 40, Room 40-S2-C01

### Bldg. 40, Room 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 14:45 15:30
Working Group 4: ADC/DAC (C. Hovater) 40-S2-C01

### 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 14:45
Adaptive Resonance Control for LANSCE-R 10m
Resonance control of the low level RF control of the linac accelerator system achieves the match between the driving RF frequency and the cavity resonance frequency, the match between the cavity resonance frequency and the linac design operating frequency. First of all, these matches are obtained by adjusting the driving RF frequency to the cavity resonance frequency. When the driving RF frequency is far from the cavity resonance frequency, the frequency difference is estimated. This frequency error estimate programs the NCO whose output is used for the carrier frequency of the IQ modulator in the low level RF system. When the adjustment of the driving RF frequency to the cavity resonance frequency is successful and hence the frequency error estimate is within the design specification, the water resonance cooling control system (RCCS) begins to tune water system such that the cavity resonance frequency tracks the linac design operating frequency while the driving RF frequency continuously tuned to follow the cavity resonance frequency as well. As tunes proceed further successfully, the driving RF frequency and the cavity resonance frequency enter the region of frequency within the specified boundary of the design linac operation frequency. The driving RF frequency is locked to master oscillator which runs at the linac design operating frequency and only the RCCS tunes the cavity resonance frequency to follow the design linac operation frequency. In this paper, we propose an adaptive algorithm that estimates the frequency difference between the generator frequency and the cavity frequency, the tracking scheme of the generator frequency to the cavity frequency, the system level design of the hardware in the ALTERA DSP Builder Environment, implementation of the whole scheme on the ALTERA STRATIX II FPGA, and the system test on the LANSCE low power RF system test-stand.
Speaker: Mr Sungil Kwon (LANL)
• 15:30 16:00
Tea break 30m
• 16:00 17:30
Working Group 1: Synchrotrons/LHC: WG Session 1 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Convenor for this working group is Dr. Mike Brennan and the Scientific Secretary Dr. Philippe Baudrenghien.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Dr Mike Brennan (BNL)
• 16:00
An IQ-based low-level RF prototype for ALBA 7m
An analog Low-level RF prototype based on IQ demodulation has been designed and developed at CELLS which should regulate the amplitude of the cavity voltage with a stability of 1% and the phase with a stability of 1°. The LLRF is controlled and monitored by an industrial PC with cPCI data acquisition cards for the I/O signals. A series of tests have already been done to evaluate the performance of the open- loop system and the individual parts including the IQ mod/dem, the differential amplifiers and the electronic phase-shifter. To test the closed-loop system, a pill- box mock-up cavity has been designed and built at CELLS which later will be driven and controlled by the LLRF. A general overview of the deign and some test results will be presented.
Speaker: Mr Hooman Hassanzadegan (CELLS)
• 16:07
Digital LLRF for Alba - Conceptual design 20m
Speaker: Ms Angela Salom Sarasqueta
• 16:27
ALBA RF Sytem 20m
Speaker: Dr Francis Perez
• 16:00 17:30
Working Group 2: Lessons learned from recent developments Bldg. 40, Room 40-5-A01

### Bldg. 40, Room 40-5-A01

#### CERN

The Convenor for this working group is Mr. Mark Champion and the Scientific Secretary Mr. Tony Rohlev.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Mr Mark Champion (ORNL/SNS)
• 16:00 17:30
Working Group 3: Session 1b: RF Software concepts (Control Systems, Architecture of LLRF, Subsystems, etc.) Bldg. 40, Room 40-4-C01

### Bldg. 40, Room 40-4-C01

#### CERN

The Convenor for this working group is Dr. Stefan Simrock and the Scientific Secretary Dr. Andy Butterworth.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations 39,42,50 will be presented in a joint session with working group 4 and are described there. Again the convenors will decide the timetable for this.

Convener: Dr Stefan Simrock (DESY)
• 16:00
Measurement Receiver for LLRF Control System 20m
Speaker: Mr Tomasz Filipek (Warsaw University of Technology)
• 16:00 16:45
Working Group 4: DSP topics (B. Chase) Bldg. 40, Room 40-S2-C01

### Bldg. 40, Room 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 16:07
Software development for FPGA based cavity controller and simulator SIMCON 10m
The FPGA based cavity simulator and controller provides the features and performance which is unique in todays control devices used in LLRF system. The software which is provided for the hardware operation can be used also for algorithms development. It consists of two control enviroments DOOCS and Matlab based. The first one is dedicated for the regular device operation during the experiment. The second one is the main testing tool used during firmware algorithm development. Both systems use unified communication layer designed for FPGA based devices. A possible applicaions of the software environment have been presented.
Speakers: Mr Jaroslaw Szewinski (Institute of Electronic Systems Warsaw Univ of Technology / Deutsches Elektronen Synchrotron DESY) , Mr Piotr Pucyk (Institute of Electronic Systems Warsaw Univ of Technology / Deutsches Elektronen Synchrotron DESY)
• 16:45 17:30
Working Group 4: Summary 40-S2-C01

### 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 18:30 22:00
Banquet Evening: meeting point @CERN in front of Bldg 39. The bus will leave at 18:45.
• Wednesday, 12 October
• 08:30 10:20
Talks Session 3 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Mr. Joachim Tuckmantel

Convener: Mr Roland Garoby (CERN)
• 08:30
Fermilab High Power Test Facility 30m
The SMTF collaboration is constructing a broad-based high-power test facility (HPTF) for superconducting RF modules at Fermilab. The purpose of HPTF is to test pulsed and CW superconducting modules and support systems with and without beam to facilitate progress in SRF. Slated projects are International Linear Collider 1.3 GHz modules, Proton Driver 325 MHz modules and Cornell CW modules. Fermilab is collaborating with DESY to provide Low Level RF control systems for these projects. In addition, SNS will provide a second LLRF system which may be used as a passive monitor or as the system controller. Design of the Master Oscillator, LLRF and Simulation work are covered. An overview of some of the recent LLRF system upgrades is also presented to provide context to our plans and directions for HPTF
Speaker: Mr Brian Chase (FNAL)
• 09:00
RF for large heavily loaded rings: limiting factors and promising new developments 20m
Super B-factory designs under consideration expect to reach luminosities in the $10^{35}$ - $10^{36}$ range. The dramatic luminosity increase relative to the existing B-factories is achieved, in part, by significantly raising the beam currents stored in the electron and positron rings. In such machines beam loading effects drive the RF system design. The main effects are the synchronous phase transients due to the uneven ring filling patterns and the longitudinal coupled-bunch instabilities driven by the fundamental impedance of the RF cavities. A systematic approach to predicting such effects and for optimizing the RF system design will be presented. Existing as well as promising new techniques for reducing the effects of heavy beam loading will be described and illustrated with examples from the existing storage rings including PEP-II and KEKB. Work supported by U.S. Department of Energy contract DE-AC02-76SF00515
Speaker: Dr Dmitry Teytelman (SLAC)
• 09:20
Multichannel down-converter board for cavity field detection at the Tesla-Test-Facility 20m
Technical standard of Downconverters used for Tesla User Facility at DESY - Preview of Board schematic of 8 channel version of Downconverter - Behaviour of linearity, offset, crosstalk and noise parameters - Estimation of accuracy of phase measurement.
Speaker: Mr Guenter MOELLER (DESY)
• 09:40
Ultra-linear Receivers for Digital LLRF Control Systems 20m
Superconducting Accelerators, worldwide, are appealing to digital Low-Level RF (LLRF) control systems in order to achieve high-precision RF gradient and phase regulation, typically less than 0.1% and 0.01 degrees, respectively. Although mostly digital, these high-performance systems still rely on analog front-end receiver components for down-conversion, amplification, and pre-filtering. The linearity aspects of digital and numerical stages are undisputed, but in most cases, the effects of non-linear signal corruption within the analog RF components are not easily corrected, and can ultimately limit the system performance. Therefore, special design efforts are required to achieve ultra-linear performance, while controlling dynamic range, sensitivity, power consumption, and cost. This discussion presents some of the non-linear front-end parameters, and quantitatively relates them to system specifications. In addition, techniques used to predict, measure and quantify these effects are presented.
Speaker: Mr John Musson (Jefferson Lab)
• 10:00
Automation of Large Scale RF Systems 20m
Future accelerator projects such as the European X-FEL and the International Linear Collider (ILC) will require the operation of the order of 1000 to 10,000 cavities. The operation of these large scale RF systems must be highly automated to guarantee high performance and availability of the linacs. The automation must provide a framework in the accelerator control system such as sequencers or state machines which allow the implementation of operational procedures developed by experts or operators. The procedures include turn-on procedures, basic parameter settings, exception handling and self protection of the system. In addition, the automation should provide a procedure to continously optimizie machine parameters operating close to the performance limit where a human operator would be overstrained by the quantity of single subsystems to be operated and the complexity in interaction between the systems and influence on the beam quality at the end. It looks like due to the evolution of procedures, the lack of concept the large scale linear accelerator facilities nowerdays planned will run into problems. We describe a useful ansatz for an automation concept and a concept for a universal interface for implementing each of the single accelerator components to be automated.
Speaker: Dr Markus Hoffmann (DESY)
• 10:20 10:50
Coffee break 30m
• 10:50 12:50
Talks Session 3 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary of this session is Mr. Joachim Tuckmantel

Convener: Mr Roland Garoby (CERN)
• 10:50
Transient Microphonic and Ponderomotive Effects In Superconducting Cavities 20m
Transient microphonic and Ponderomotive effects have been observed and measured under a number of different operating conditions. Microphonic effects are those sources that are external to the cryomodule, or by devices such as piezo tuners and mechanical tuners. Ponderomotive effects are changes in frequency due to changes in rf field through the Lorentz force. This talk will give a brief summary of some of these effects. Of particular interest is recovery from an arc in the region of the cold window on a 5-cell CEBAF structure. Closed loop gradient control was employed during arc recovery experiments conducted in the CEBAF accelerator at Jefferson Lab. During this test, instabilities were observed in the cavity forward power signal, which were determined to be ponderomotive in nature. These ponderomotive effects were quantified using a cavity resonance monitor and a VCO_PLL RF system. Two types of ponderomotive effects were observed depending on the type of arc event. If the arc occurred in the vacuum space between the warm and cold windows, the transient frequency shift was about 75 Hz peak-to-peak. If the arc occurred on the cavity side of the cold window the transient frequency shift was about 400 Hz peak-to-peak. The background microphonics level for the tested cavity was approximately 30 Hz peak-to- peak. Other data relating to dynamic Lorentz force and piezo tuner transient effects will also be presented. Experimental results, analysis of the resultant klystron power transients, the decay time of the transients, and the implications with respect to fast reset algorithms will be presented.
Speaker: Kirk Davis (Jefferson Lab)
• 11:10
Precision low-noise field detectors 20m
For a precise detection of the cavity rf-field with a resolution of 0.01 degree (rms) in phase and amplitude smaller then 5 10^-5 (rms), we present a down- converter prototype. The down-converter is designed for a multichannel phase and amplitude readout using an intermediate frequency between 10-100MHz, which is sampled by an ADC. We give an overview of commercial available mixers, present noise sources and discuss how noise, linearity and long term drift limit the precision of field detection. In addition we study the beam jitter induced by these noise sources within the regulation. We acknowledge financial support by DESY Hamburg and the EUROFEL project.
Speaker: Mr Frank Ludwig (DESY)
• 11:30
The phase reference distribution system for the TESLA technology based projects 20m
The presentation will cover the actual development of the master oscillator (MO) and the frequency distribution system for the UVFEL and the XFEL projects. The MO design issues will be briefly described. The concept of a phase stable signal distribution system based on the RF and fiber-optic solutions will be shown. A long distance signal distribution subsystem with a feedback loop actively stabilizing signal phase will be described. Current distribution system status, performance measurements and plans for future development will be shown.
Speaker: Mr Krzysztof Czuba (DESY Hamburg)
• 11:50
Diagnostic System for Low Level RF Control System for VUV-FEL 20m
In order to provide a continuous work of the VUV-FEL and high stable RF field during pulse is necessary to monitor all parameters of the Control System. An advanced algorithms looking for a correlation between data from different subsystems of the LLFR, on-line measure field parameters , produce information about current condition and performance of the LLRF Control System. Diagnostic system requires some additional hardware – test signals, monitoring points for analog and digital signals. It is necessarily to integrate diagnostic system into control system to increase productivity and decrease cost of the system. If performance degradation is detected it is possible make action (e.g. calibrate system or check subsystem) before an error appear and do not break operation. This article describes the Construction and principles of the performance detection algorithms and requirements for hardware are presented.
Speaker: Tomasz Jezynski (Technical Univ. of Lodz)
• 12:10
Complex digital circuit design for LHC Low Level RF 20m
Modern Low-Level RF systems rely heavily on digital signal treatment. This evolution has been mainly enabled by the recent availability of powerful FPGAs. Current FPGAs can implement almost all of the signal-treatment requirements e.g. post-mortem and observation memory management, base-band network analyzer functionality, the remote interface to the control systems front-end computer and implement an optional DSP interface bus. The FPGA hardware design language descriptions required for these tasks can easily be entered or imported in the Visual-Elite mixed graphical textual design environment used for the LHC Low Level RF. Standard features are the support of standard VHDL text modules as well as block-diagrams, graphical state-machines, truth-tables and flow-chart hardware descriptions, however the main interest lies in the inherent code reusability between different projects and it's independence of target device architecture. This presentation will give a short overview of the design flow, the tools employed and some examples of implementations like: Serial programmable delay controller, IQ demodulator, base-band network analyzer, CIC and half-band decimation / interpolation filters, cross product and power product calculators and a CORDIC based calculator.
Speaker: Mr John Molendijk (CERN)
• 12:30
Characterization of SNS low-level RF control system 20m
A single high-density FPGA XC2V1500 plus 14-bit ADCs/DAC and 128Mb SRAM forms the digital hardware platform of SNS low-level RF control system. The carefully designed HDL implementation has limited controller latency under 150 ns (6 clock cycles) which allows the possibility for obtaining a rapid real-time feedback control. Given the typical 1 us external loop delay , a small signal control bandwidth over 100kHz has been demonstrated on a NC cavity with a classic SISO/P-I control only configuration. The large on-chip dual-port RAM and logic as well as the off-chip SRAM supports the implementations of more sophisticated dsp/feed forward algorithms required for pulsed super-conducting LINAC.
Speaker: Mr Hengjie Ma (SNS, Oak Ridge National Laboratory)
• 12:50 14:00
Lunch 1h 10m
• 14:00 15:30
Working Group 1: Synchrotrons/LHC: WG Session 2 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Convenor for this working group is Dr. Mike Brennan and the Scientific Secretary Dr. Philippe Baudrenghien.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Dr Mike Brennan (BNL)
• 14:00
Noise data in coast - RHIC 20m
Speaker: Dr Mike Brennan (BNL)
• 14:00 15:30
Working Group 2: Performance limits and challenges, problem areas in existing applications Bldg. 40, Room 40-S2-B01

### Bldg. 40, Room 40-S2-B01

#### CERN

The Convenor for this working group is Mr. Mark Champion and the Scientific Secretary Mr. Tony Rohlev.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Mr Mark Champion (ORNL/SNS)
• 14:00 15:30
Working Group 3: Session 2a: RF Software Feedback and Applications Bldg. 40, Room 40-4-C01

### Bldg. 40, Room 40-4-C01

#### CERN

The Convenor for this working group is Dr. Stefan Simrock and the Scientific Secretary Dr. Andy Butterworth.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations 39,42,50 will be presented in a joint session with working group 4 and are described there. Again the convenors will decide the timetable for this.

Convener: Dr Stefan Simrock (DESY)
• 14:00
Low-Level RF Station Operation for the VUV-FEL at DESY and other Linear Accelerators 20m
Speaker: Mr Wojciech Cichalewski (Technical University of Lodz)
• 14:20
Internal Interface 20m
Speaker: Mr Wojciech Jalmuzna (University of Technology, Warsaw)
• 14:00 14:30
Working Group 4: Discussion 40-S2-C01

### 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 14:30 15:00
Working Group 4: Receiver Technology (F. Pedersen) Bldg. 40, Room 40-S2-C01

### Bldg. 40, Room 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 14:30
Femtosecond optical synchronization systems for XFELs 7m
Speaker: Mr Axel Winter (DESY)
• 14:37
SNS LLRF Reference System 7m
Speaker: Mr Maurice Piller (ORNL)
• 15:00 15:30
Working Group 4: Master Oscillator / Distribution (F. Pedersen) 40-S2-C01

### 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

• 15:30 16:00
Tea break 30m
• 16:00 17:30
Working Group 1: Synchrotrons/LHC: WG Session 2 Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Convenor for this working group is Dr. Mike Brennan and the Scientific Secretary Dr. Philippe Baudrenghien.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Dr Mike Brennan (BNL)
• 16:00 17:30
Working Group 2: ILC session Bldg. 40, Room 40-S2-B01

### Bldg. 40, Room 40-S2-B01

#### CERN

The Convenor for this working group is Mr. Mark Champion and the Scientific Secretary Mr. Tony Rohlev.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that a few oral presentations will be presented in joint sessions of different working groups. Again the convenors will decide the timetable for this.

Convener: Mr Mark Champion (ORNL/SNS)
• 16:00 17:30
Working Group 3: Session 2b: RF Control Operation (Automation, Procedures, Reliability) Bldg. 40, Room 40-4-C01

### Bldg. 40, Room 40-4-C01

#### CERN

The Convenor for this working group is Dr. Stefan Simrock and the Scientific Secretary Dr. Andy Butterworth.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations 39,42,50 will be presented in a joint session with working group 4 and are described there. Again the convenors will decide the timetable for this.

Convener: Dr Stefan Simrock (DESY)
• 16:00 16:45
Working Group 4: Control of ferrite shifters (Y. Kang) Bldg. 40, Room 40-S2-C01

### Bldg. 40, Room 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• 16:00
Vector Modulation of High RF Power using Ferrite Phase Shifters 7m
In generation and distribution of high power RF for charged particle accelerators, a fan-out configuration to power many cavities using a high powered klystron is considered advantageous in saving construction and installation cost. High power fast RF phase shifters can be used to construct a vector modulator for independent control of RF amplitude and phase delivered to an accelerating cavity. A prototype vector modulator that employs two quasi-TEM mode coaxial ferrite phase shifters has been built and bench tested. The properties of the vector modulation system and characteristics of the phase shifters derived by simulations and measurements will be presented.
Speaker: Dr Yoon Kang (Oak Ridge National Laboratory)
• 16:45 17:30
Working Group 4: System Software / Verification (P. Joireman) 40-S2-C01

### 40-S2-C01

#### CERN

The Convenor for this working group is Dr. Brian Chase and the Scientific Secretary Dr. Edmond Ciapala.

The mini-oral presentations have been assigned to the different working groups, however the exact times of the presentations will be decided by the convenor at a later date. Note that oral presentations, 39,42,50 will be presented in a joint session with WG3. Again the convenors will decide the timetable for this.

Convener: Mr Brian Chase (FNAL)
• Thursday, 13 October
• 08:30 10:30
Closing session Bldg. 40, Room 40-S2-A01

### Bldg. 40, Room 40-S2-A01

#### CERN

The Scientific Secretary for this session is Maria Elena Angoletta.

Convener: Mr Mark Champion (ORNL/SNS)
• 08:30
Tutorial: Algorithms for pulsed digital RF control 45m
Ten years ago digital hardware offered for the first time the real time processing power required for particle accelerators RF control (LLRF). This opened the opportunity to implement control algorithms hardly realizable with analog electronics like adaptive feed forward or exception handling. In contrast to digital control used in automotive engineering, hi-fi and telecommunication sector the data rates to be processed are much higher and the latency has to be much shorter in digital LLRF. Compared to algorithms used in these sectors LLRF algorithms are in an initial state. After a short overview some present LLRF algorithms like digital RF feed back, RF feed-forward, cavity resonance control, exception handling and others will be explained illustrative of implementations at the DESY TTF.
Speaker: Dr Elmar Vogel (Unknown)
• 09:15
LLRF model extraction 45m
Speaker: Dr Dmitry Teytelman (SLAC)
• 10:00
Low-Level RF to Controls Interface: Future Perspectives 30m
Speaker: Dr Larry Doolittle (LBNL)
• 10:30 11:00
Coffee break and Workshop Photograph 30m
• 11:00 13:10
Closing session Bldg.40, Room 40-S2-A01

### Bldg.40, Room 40-S2-A01

#### CERN

The Scientific Secretary for this session is Maria Elena Angoletta.

Convener: Mr Mark Champion (ORNL/SNS)
• 11:00
Poster Session Summary 10m
Speaker: Dr Wolfgang Hofle (CERN)
• 11:10
Working Group 1: Summary 25m
Speaker: Dr Mike Brennan (BNL)
• 11:35
Working Group 2: Summary 25m
Speaker: Mark Champion (ORNL/SNS)
• 12:00
Working Group 3: Summary 25m
Speaker: Dr Steffan Simrock (DESY)
• 12:25
Working Group 4: Summary 25m
Speaker: Mr Brian Chase (FNAL)
• 12:50
Closing remarks 10m
Speaker: Dr Trevor Linnecar (CERN)
• 13:10 14:10
Lunch 1h
• 14:10 17:00
Visit to CERN machines