Low Level RF Workshop 2013

30 Sept 2013, 08:45 → 4 Oct 2013, 20:00 US/Pacific

Larry Doolittle (LBNL), Steve Smith (SLAC)

We are pleased to announce that the 2013 Low Level RF workshop will be held from Tuesday, October 1, to Friday October 4, 2013 at the Granlibakken Resort in Lake Tahoe and hosted jointly by the Lawrence Berkeley National Laboratory and the SLAC National Laboratory. Sophisticated and reliable Low-Level RF systems are essential to control RF structures and their power sources in modern particle accelerators. The goals of the LLRF2013 workshop are to share our experiences, to present the status of our work, and to discuss recent developments and future prospects in this field.

Monday, 30 September

08:45 → 12:30 SLAC Tour

Meet Kavli Auditorium just inside the front gate. Contact Steve Smith

10:00 → 14:30 LBNL Tour

Contact Alex Ratti, two groups begin at 10:00. One at ALS building 6, the 2nd group at building 71 -meet in the building 71 2nd floor lobby.

17:00 → 19:00 Registration Open

19:00 → 20:30 Dinner - Granhall

20:30 → 21:30 Registration

Tuesday, 1 October

07:30 → 08:30 Registration Tuesday Morning

08:30 → 08:35 Welcome Speakers Doolittle/Smith

08:35 → 10:19 Session1: Lab Status/Activities/Highlights

08:35 KEK Lab Talk : KEK LLRF SYSTEMS

The various projects in KEK are in upgrade or construction stage. STF is a test facility of the superconducting RF. The beam operation was performed for the first time in the STF. Compact Energy Recovery Linac(cERL) is the test facility for the future 3-GeV ERL plan. The construction of injector of cERL was finished. The beam commissioning of the injector was performed for 2 months before this summer. The SuperKEKB is now under construction. The injector is also being upgraded. The high power test by using new LLRF system was performed. The digital feedback boards being employed in cERL, STF, and SuperKEKB are the almost same. Good performance was obtained at the each system. For J-PARC, the injection energy to RCS will be upgraded to 400 MeV for increase beam intensity. The new ACS cavities are installed and operated at a frequency of three times higher than the current one. I will summarize the status of the LLRF systems of the current progress projects at KEK.

Speaker: Dr Takako Miura (KEK)

08:52 Status of LLRF technology at SPring-8

SPring-8 campus has three major accelerator complexes, which are the 8 GeV SPring-8 storage ring, the 1 GeV New-SUBARU storage ring, and the X-ray free-electron laser, SACLA. The individual accelerators have own independent LLRF systems to stably drive their high-power rf sources. These LLRF systems almost are independently working each other among SPring-8, New SUBARU and SACLA. The biggest future plan at the SPring-8 campus to expand ability for user experiments is an upgrade of the SPring-8 storage ring to a diffraction-limited ring, which is called SPring-8 II. A major change of the LLRF systems will be enforced by the SPring-8 II plan. An injection of an electron beam from SACLA to SPring-8 II ring is going to be necessary, because of a small dynamic aperture of the diffraction-limited ring. This means the injected electron beam must have a very small emittance of less than 1 mm mrad, which is already established by SACLA. From this relation, the LLRF systems of SACLA and the ring should be strongly coupled. A reference rf signal and a timing pulse should control both the SACLA and ring LLRF systems, respectively. However, the rf frequency of SACLA does not have any simple integer relationship to that of SPring-8. Therefore, we presently started design work for the synchronization system between the SACLA linac and SPring-8 II ring, and replacement from the present LLRF system for SPring-8 to SPring-8 II. The idea of the synchronization system is fully digital phase control architecture by using a 10 MHz accurate time reference signal to drive main signal oscillators of both the linac and ring and to realized about a 1 ps temporal accuracy to inject electron beam into the very small dynamic aperture of the ring. The replacement of the ring LLRF system will be done by using state-of-art technology, such as advanced TCA, to obtain further controllability than that of the present ring. This presentation introduces overall future plan of SPring-8 campus and some details of the synchronization system and the replacement of the ring LLRF system.

Speaker: Dr Yuji Otake (XFEL Research and Development Division, RIKEN SPring-8 Center, RIKEN)

09:09 JLAB LLRF activities

JLAB is currently commissioning ten new C100 cryomodules, a major part of the 12 GeV Energy Upgrade Project. For each of eighty cavities, a new LLRF system has been installed and now is undergoing rigorous tests. A new 400 MHz LLRF system for Compact X-ray Source based on inverse Compton scattering has been developed, tested and delivered along with MO (master oscillator) and HPA (high power amplifier). A new 199.6 MHz LLRF & Resonance Control has been developed and delivered for Wisconsin Superconducting RF Electron Gun. This system has been successfully tested and (as will be presented during the poster session) this Wisconsin RF Gun has delivered its first beam. LLRF team provided new systems for JLAB Vertical Test Area as well as a number of 1497 MHz LLRF chassis working as a test-bed for old-style CEBAF cryomodules re-commissioning.

Speaker: Tomasz Plawski (Jefferson Lab)

09:21 Toward the next generation: LLRF activities in LBNL

LLRF systems are required to integrate information from many sources: at least the RF system, cryomodule system, interlock system, beam based feedback system, laser system, timing and synchronization system. Success is providing a stable RF field to interact with the beam in the accelerator. Multiple LBNL groups are working together to develop hardware and programming for LLRF systems that can meet the stability and integration requirements for the next generation of accelerators. In this workshop, we summarize the ongoing and evolutionary progress at LBNL, from system architecture and engineering implementation, as well as projects that apply these concepts, including APEX and SPX.

Speaker: Gang Huang (LBNL)

09:33 LLRF Status and Future Projects at Fermilab

Fermilab is currently in transition from the Energy Frontier to Intensity Frontier Physics and is coming out of a 1.5 year shutdown of the Main physics program. This shutdown provided upgrades to allow 700kW of beam power on target for neutrino physics, requiring a new 53 MHz LLRF system for the Recycler Ring where beam slip stacking will take place. The Advanced Superconducting Test Accelerator (ASTA), has recently seen first beam from the photocathode gun and cool down of the SRF cryomodules is in progress. Plans for this user facility are discussed. Initial LLRF designs for Project X and the Project X Injector Experiment (PXIE) are in progress with several possible High Level RF design concepts under evaluation.

Speaker: Mr Brian E. Chase (FNAL)

09:50 ORNL Lab Talk: Operational Experience with the SNS Low Level RF Systems

The Spallation Neutron Source has been in operation since first beam on target in 2006. Over the years we have made incremental improvements to the RF systems to correct problems encountered and to better support operation of the accelerator. Several of these improvements will be highlighted along with some of the operational statistics for the accelerator. We have also made a major effort to add to the testing infrastructure at the SNS site to include our recently commissioned vertical test area, plasma processing system, and infrastructure to test the new RFQ scheduled to arrive in the first quarter of 2014.

Speaker: Mark Crofford (Oak Ridge National Laboratory)

10:02 Cern Highlights 2011-2013 (Lab Talk)

An overview will be given over the CERN LLRF highlights of the last two years with an outlook into the future. This includes operation and optimization of the LHC LLRF system for luminosity production of the LHC run I. Moreover, preparations are under way for the upgrades of the LHC injectors and the LHC itself with an ambitious plan of increasing LHC luminosity both for protons and ion operation. Amongst the new accelerators that require dedicated LLRF systems are the low intensity HIE-Isolde superconducting Linac for radioactive ions, the new H- Linac, “LINAC4” and an anti-proton decelerator project “ELENA” that has been launched. Testing of accelerating structures at 12 GHz for a future linear collider (“CLIC”) continues and activities for LLRF for superconducting cavities is ramping up with the demand of sophisticated feedback systems for crab cavities for LHC as well as general cavity testing for future accelerators.

Speaker: Dr Wolfgang Hofle (CERN)

10:20 → 10:40 Break

10:40 → 11:48 Session 2: Lab Status/Activities/Highlights

10:40 Overview on the LLRF and MicroTCA developments at DESY

In 2010, the decision was taken to base the European XFEL LLRF controls hardware on the new emerging crate standard MTCA.4. First proof-of-principle experiments to operate a cryogenic accelerator module equipped with 8 cavities have been successfully carried out in autumn 2011. The final board design choices, several improvements and architectural adaptions have since been made allowing now for European XFEL LLRF series production. Meanwhile, several accelerator and test facilities such as the free electron laser FLASH, the Cryo Module Test Bench (CMTB), the Accelerator Module Test Facility (AMTF) and the normal conducting 3.0 GHz Relativistic Electron Gun for Atomic Exploration (REGAE) have been equipped and operated based on MTCA.4. In addition, DESY has received funding from the Helmholtz Association through the Validation Fund .MTCA.4 for Industry. to further develop the MTCA.4 crate standard and to foster its usage in scientific applications and to industrial markets.

Speaker: Holger Schlarb (DESY)

11:02 BNL Lab Talk: Recent Applications and Performance of the RHIC LLRF Platform Across the Collider-Accelerator Complex

The RF systems of the Collider-Accelerator complex at BNL present a significant and diverse array of LLRF control challenges: multiple hadron species, fast cycle to cycle configuration changes, large frequency sweeps, precision wide dynamic range cavity control, beam control feedback loops, complex RF bunch manipulation gymnastics, machine to machine synchronization, system protection, diagnostic data, etc. The RHIC LLRF Platform was developed to provide a common, modular, flexible and scalable LLRF digital control platform to address this variety of application and performance demands. Initial applications of the platform from 2010-2011 included the first phases of the RHIC LLRF upgrade, a new LLRF system for the RHIC Electron Beam Ion Source and implementation of the RHIC Spin Flipper Controller. During 2012-2013, the RHIC LLRF system was further expanded and had significant new functionality added, a new LLRF system was developed and commissioned for the R & D Energy Recovery Linac, the AGS LLRF system was upgraded and commissioned, and the first phase of the Booster LLRF upgrade was completed. More applications are in development stages. Here we provide an overview of these most recent applications, emphasizing the platform features which provide for flexibility, scalability and ease of integration.

Speaker: Kevin Smith (BNL)

11:19 Overview of the LLRF Activities at SLAC

The Linac Coherent Light Source (LCLS) utilizes the last 1 km of the SLAC Linac to serve as the electron beam source for the free electron laser (FEL). The FEL provides the x-ray pulses from several to a few hundred femtoseconds. To achieve this goal, the RF system is required to be temporally stable below 100fS in several critical klystron stations. A lot of R&D and upgrades of the LLRF systems have been completed. These upgrades include the Linac source Master Oscillator, Master Amplifier, new design of the LCLS Reference System, new Phase and Amplitude control and detecting system, and the solid state sub booster (SSSB) amplifiers used as klystron pre-amplifiers. LCLS’s successful operation has enabled the investment in light-source and accelerator related research facilities at SLAC. Several LLRF systems have been provided for these facilities such as Accelerator Structure Test Area (ASTA), X-Band Test Area (XTA), FACET XTCAV and recently the X-Band Transverse Deflector for Femtosecond Electron/X-ray Pulse Length Measurements at LCLS. The LLRF team has also been working on beam induced RF phasing, intra-pulse feedback, and more compact and robust control with MicroTCA based LLRF system. This paper will give an overview of all the LLRF activities at SLAC.

Speaker: Bo Hong (SLAC)

11:31 The SwissFEL LLRF system - concept and realization

PSI prepares the construction of an X-ray free-electron laser facility, SwissFEL, which has received full funding by end of 2012. The baseline design consists of a 5.8 GeV normal conducting linear accelerator operated with a repetition frequency of 100 Hz and two FEL lines providing X-rays with pulse lengths in the fs range and wavelengths down to 0.1 nm. The generation of high-brightness beams with a very compact accelerator layout imposes tough stability requirements on the RF system and hence on the LLRF system. In this contribution the concept of the SwissFEL LLRF system will be introduced. It has to control the accelerating fields of RF standing and traveling wave structures at S-, C- and X-band frequencies and cope with vector sum control of up to four cavities including phase modulation schemes for the operation of barrel open cavity RF pulse compressors. The modular hard- and software structure of a scalable modern digital processing platform based on FPGAs, multicore CPUs running real-time Linux, and FMC standards will be presented. A prototype LLRF system for C-band has been developed and first experiences will be given.

Speaker: Mr Roger Kalt (Paul Scherrer Institut)

12:00 → 12:30 Session 3: Systems

12:00 Progress in Low Level RF System at FLASH Facility

The Free-Electron Laser in Hamburg (FLASH) is a user facility delivering femtosecond short radiation pulses in the wavelength range between 4.2 and 44 nm using the SASE principle. Currently the major extension of the facility is ongoing. The modifications include a new experimental hall to double the number of user stations and an additional variable-gap undulator in a separate tunnel to be able to deliver two largely independent wavelengths to two different user stations simultaneously. The electron beam will be switched between the present fixed-gap undulator line and the new variable gap undulator. In the meantime the FLASH low level RF system is being upgraded to MicroTCA-based system to replace the VME-based system and serve as a test bench for the European X-ray Free Electron Laser LLRF system. We will present details on the new FLASH LLRF system setup. The benefits and measurement results of the newly installed system will be given. We will also show the first preliminary results demonstrating simultaneous RF operation for multi beam-lines at the FLASH facility.

Speaker: Dr Valeri Ayvazyan (DESY)

12:30 → 14:00 Lunch - Granhall

14:00 → 15:35 Session 4 Systems

14:00 Update on XFEL LLRF system development and production

Due to its scale and tight specifications, the European XFEL poses real challenges in the field of low-level radio-frequency (LLRF) control. The implementation of this semi-distributed control system on the microTCA for physics (MTCA.4) platform triggered numerous designs of electronic boards and helped refine the usage of MTCA.4 technology for physics and control systems. This contribution presents an overview of the MTCA.4-based LLRF system for the European XFEL with an emphasis on key hardware and software components. The current production status, installation planning, quality management control and preliminary results are also presented.

Speaker: Julien Branlard (DESY)

14:25 Recent Progress of LLRF System for SACLA

The X-ray free-electron laser, SACLA, at SPring-8 emits a steady laser with an intensity of 300 uJ at 10 keV, which strongly supported by a highly accurate and stable LLRF system. This LLRF system realizes about a 20 fs short-term temporal stability. However, SACLA still requires operator trimming once a 30 minutes to the rf phase and amplitude of injector cavities. Hence the long-term stability of the laser is not sufficiently stable. The target long-term stability must be within about 300 fs at least, which calculated by simulation. One of causes of the drifts is a temperature movement of +/- 0.1 K around enclosures, where LLRF instruments like in-phase/quadrant (IQ) modulator/demodulators are installed. This temperature movement also affects the optical length of phase-stabilized fibers. A temporal value corresponding to this length change is about several-hundred fs ~ several ps for the 1 km optical fiber. To mitigate these kinds of the effects, we installed precise temperature controllers with a temperature stability of 0.01 K for the injector enclosures and are presently installing optical fiber length controllers. This precise temperature controllers use heaters and the pulse width modulation (PWM) method to manipulated an AC heater current for realizing a 0.01 K stability. The optical fiber length controller uses Michelson interferometory to stabilize the whole fiber length by feedback control. The precise temperature controller greatly helps to increase laser intensities from 300 uJ to over 400 uJ and its stability. The installation of the fiber length control system will be finished around the end of this year. We expect this system also reduces the drift. This presentation describes recent progress of our LLRF system development including the improvements mentioned above.

Speaker: Dr Yuji Otake (XFEL Research and Development Division, RIKEN SPring-8 Center, RIKEN)

14:50 LLRF system for the HIE-Isolde

The HIE-ISOLDE project is a major upgrade of the ISOLDE and REX-ISOLDE (radioactive nuclear beams) facilities at CERN. The most significant improvement will come from replacing most of the existing REX accelerating structure by a 40 MV superconducting linac based on 32 independently phased superconducting quarter-wave resonators (cavities). The new linac will raise the energy of post-accelerated beams from 3 MeV/u to over 10 MeV/u. The resonators operate at a frequency of 101.28 MHz and at a relatively high Q, providing an operational bandwidth of only a few Hertz. The resonators are tested and conditioned at their intrinsic Q providing bandwidth only of a fraction of a Hertz. A new, fully digital LLRF system is being developed to operate the cavities at 0.2°/0.2% field accuracy. The very narrow resonator bandwidth introduces specific problems in cavity conditioning, measuring their parameters and the operation. A mechanical tuning system with a fraction of a Hz resolution and low microphonics is being developed to tune the cavity to the desired frequency. Lorentz force detuning of the tuning plate makes the cavity power up sequence, fast set point changes, as well as recovery from a sudden field loss very challenging. The cavity needs to be started up with a self-excited loop, undergo the mechanical resonant frequency tuning and be glitch-lessly handed over to the generator driven mode. The system design is presented along with the challenges and first results obtained on a cold cavity.

Speaker: Dr Daniel Valuch (CERN)

15:15 Crab Cavity system for the LHC high-luminosity upgrade. Proposed LLRF

In collision, the LHC has so far been operated at 4 TeV/c per beam with up to 0.4 A DC current (per beam) distributed in 1320 bunches at 50 ns spacing. The future is 7 TeV/c, up to 1.1 A DC current and 25 ns spacing (HiLumi LHC). The roadmap to HiLumi LHC includes the installation of Crab Cavities operated at the fundamental accelerating frequency (400.8 MHz), with six cavities per beam at the two high luminosity experiments ATLAS and CMS. The LLRF must reduce the cavity impedance of the deflecting mode, without injecting excessive RF noise on the betatron bands that would result in transverse emittance growth and decrease the luminosity lifetime. Also important is the perfect pairing of crabbing and un-crabbing kicks on each side of an experiment so that the bunch tilt does not propagate to the rest of the trajectory. Following a cavity trip, the field in the companion cavities will follow to minimize the losses till the beam dump has reacted (300 microseconds maximum). The LLRF must also cope with a complex operational scenario. During filling and ramping the cavities will be detuned to make them transparent to the beam. On flat top, the detuning will be reduced, with zero total demanded voltage (probably using counter-phasing between cavities). The generator power needed to compensate the beam loading will be used to guide the beam centering using dedicated corrector magnets. Finally the desired crabbing field will be applied.

Speaker: Dr Philippe Baudrenghien (CERN)

15:50 → 16:10 Break

16:10 → 16:55 DSP Tutorial I: Basic Theory - Presenter Tim Berenc, ANL

Convener: Tim Berenc (Argonne National Laboratory)

16:55 → 17:55 Session - Other

16:55 Standardization of EPICS Device Support for data acquisition

Every control system has to deal with a large number of input/output devices which offer a similar kind of capabilities. For example, all data acquisition (DAQ) device offer sampling at some rate, which in many cases is configurable. Here, an attempt to standardize such interfaces. The Nominal Device Model (NDM) is a model which proposes to standardize the EPICS interface of analog and digital input and output devices, as well as image acquisition devices (cameras).

Speaker: Dr Stefan Simrock (ITER)

17:20 LLRF Using SoC FPGAs in a Multiple Development Tool Environment

The shrinking silicon geometries in FPGA technology provides the designer today with greatly increased computing and logic resources in a compact low cost package. Application class floating point hard processor cores with a number of high speed peripherals are tightly integrated with FPGA logic on a single chip. The new FPGA architectures come with several high level design and debugging tools that promise to reduce the development effort and time for complex systems. Multiple tools such as, QuartusII, Qsys, DSP builder, Nios EDS, SoC EDS from Altera, Matlab and Simulink from Mathworks, DS-5 development environment from ARM, Linux or VxWorks operating system kernel builders for the ARM, need to be used to design with these new FPGA chips. In the context of a system design for a LLRF system, these new tools and FPGA’s are evaluated for their ease of use and the new system architectures that could be adopted for future systems. A complete LLRF system is designed with an Altera 28-nm CycloneV SoC chip with 110k logic elements and dual 800 MHz ARM processors and the tools and new architecture options are explored.

Speaker: Philip Varghese (Fermi National Accelerator Laboratory)

17:55 → 18:00 Announcements

18:00 → 19:00 Welcome Reception

19:00 → 20:30 Dinner Granhall

Wednesday, 2 October

08:30 → 08:34 Announcements

08:35 → 09:20 Keynote - Applications of X-Ray Light Sources

Speaker: John Byrd (LBNL)

09:20 → 10:20 Session 1: Hardware

09:20 A Leading-Edge Hardware Family for LLRF and Diagnostics in CERN's Synchrotrons

A leading-edge hardware family, evolution of that successfully deployed in CERN’s Low-Energy Ion Ring (LEIR), is under development at CERN to address the low-level RF (LLRF) needs of synchrotrons in the Meyrin site. It will be deployed in 2014 in the CERN’s PS Booster and in the medical machine MedAustron. It will be then retro-fit to the LEIR machine to standardise the LLRF implementation. It will also be used for the LLRF as well as longitudinal diagnostics implementation for the new Extra Low ENergy Antiproton (ELENA) Ring, a new synchrotron that will be commissioned in 2016 to further decelerate the antiprotons transferred from the CERN’s Antiproton Decelerator (AD). The requirements for the LLRF as well as for the diagnostics systems are very demanding owing to the revolution frequency swing, dynamic range and low noise required by the cavity voltage control and digital signal processing to be performed. This talk gives an overview of the main building blocks of the hardware family and of the associated firmware and IP cores.

Speaker: Mr John Molendijk (CERN)

09:45 RF Backplane for MTCA.4 Based LLRF Control System

The Low Level RF (LLRF) control system developed for linear accelerator based Free Electron Lasers (FEL) require real-time processing of thousands RF signals with very challenging RF field detection precision. To provide a reliable, maintainable and scalable system a new development of the LLRF control based on MTCA.4 architecture was started in DESY for FLASH and European-XFEL. In contrast to standard RF control systems realized in 19" modules, we could demonstrate setup with field detection, RF generation, RF distribution, DAQ system and the high-speed real-time processing entirely embedded in the MTCA.4 crate system. This unique scheme embeds ultra-high precision analog electronics for detection on the Rear Transition Module (RTM) with powerful digital processing units on the Advanced Mezzanine Card (AMC). To increase system reliability, maintainability and reduce performance limitations by RF cabling network, we developed and embeded in the MTCA.4 crate an unique RF Backplane (uRFB) for RTM cards. This backplane is used for distribution of high-performance Local Oscillator (LO), RF and low-jitter clock signals together with low-noise analog power supply to analog RTM cards in the system. In this paper we present the architecture of the MTCA.4 crate with the uRFB, the RF Backplane design and successful laboratory test results of the LLRF control system demonstrating the performance of our development.

Speaker: Dr Krzysztof Czuba (ISE, Warsaw University of Technology)

10:20 → 10:40 Break

10:40 → 11:25 DSP Tutorial II: Real-life Implementation

10:40 DSP Tutorial II : Real-life implementation in FPGAs

The tutorial will demonstrate the integrated flow from system design to implementation of real-time DSP applications on FPGAs. The objective is to show developers with limited or no FPGA design experience that they can quickly take a DSP algorithm from concept to verification. Traditional RTL developers can also benefit from this design methodology to cut down on verification time and perform implementation trade-off analysis. The steps taken to validate a simple application using XILINX System Generator with MATLAB / Simulink and hardware-in-loop verification will be shown. An overview of our current designs at BNL using this method and results achieved will be discussed.

Speaker: Mrs Geetha Narayan (BNL)

11:25 → 12:15 Session 2: Hardware

11:25 MTCA.4 Fast Digitizer

This contribution describes the design of an Advanced Mezzanine Card (AMC) in the MTCA.4 standard suited for direct analog-to-digital conversion of high-frequency signals up to 2.7 GHz with a maximum ADC clock frequency of 800 MHz. Signal conversion is performed using the undersampling technique. This card was designed for the needs of the LLRF and other control and measurement systems of the FLASH and XFEL accelerators. The designed module consists of eight very-high-speed ADC channels, four high-speed and precision DAC channels, a powerful FPGA unit, fast SRAM memory, along with special power supply and diagnostic circuits. The AMC digitizer work in pair with various project-specific Rear Transition Modules (RTMs). This paper describes such issues as system organization allowing acquisition of data at such high rates as circuit synchronization by high-quality clock signals, as well as performance and usage of direct sampling of high frequency signals.

Speaker: Mr Samer Bou Habib (DESY / WUT-IES)

11:55 Standardization of ITER Instrumentation and Controls based on Hardware Catalogs

The engineering challenge of the ITER procurement model is to maintain the schedule of-over 170 Plant System delivery, commissioning and integration in order to assure cost-effective construction with timely transition to the operation phase. The standardization of the components used in construction plays a vital role to reach these objectives. The CODAC Division is addressing the Instrumentation and Controls (I & C) standardization through an ITER baseline document, called Plant System Design Handbook (PCDH) which collects together the requirements and the mandatory procedures. The PCDH does not define required hardware specifications but rather addresses the mandatory industrial standards; the most important from the I/O point of view being the PCI Express and Ethernet interconnect. The PCDH Document is supported by numerous satellite documents which are each addressing the specific requirements or procedures by linking them with real-life I & C implementations using common engineering terms. The ITER I & C hardware catalogs are PCDH satellite documents which collect together Commodity off the Shelf (COTS) I & C components which are validated and recommended by the CODAC Division. The Slow Controller catalog is listing validated PLC devices which are designated to be used in control loop in 10 ms range. The Fast Controller catalog lists numerous PCI Express and Ethernet enabled I & C devices in PXIe, ATCA and MTCA.4 form factors to deal with fast I & C problems covering the application range all the way up to the ITER Diagnostics Plant Systems. Third catalog is listing the network equipment to be used within the I & C systems.

Speaker: Dr Stefan Simrock (ITER)

12:15 → 12:30 Topical Discussion, Trends in FPGA development tools

12:30 → 14:00 Lunch -Granhall

14:00 → 17:00 Afternoon Outing - free time

17:00 → 18:55 Start Bus service to Banquet

Bus departs to Gar Wood 5:00, 5:10 & 5:20 (1 -29 passenger runs back and forth until 8:30) Ref: Northstar

19:00 → 21:00 Working Dinner at GarWoods

21:00 → 21:50 Bus Service (Return)

Buses depart 8:30, 9:00 and last bus leaves 9:30

21:50 → 22:00 Finish Bus service return

1 large bus (40 passenger) finishes service at 9:30

Thursday, 3 October

08:30 → 08:35 Announcements

08:35 → 09:20 Industry Talk, "Realizing the Potential of High-Performance Data Converters," Tom Linnebrink

08:35 Realizing the Potential of High-Performance Data Converters

Modern analog-to-digital converters (ADCs) and digital-to-analog converters (DACs) have been designed to provide very high performance in terms of instantaneous bandwidth, signal-to-noise ratio, and linearity. Realizing their inherent performance in a system requires careful engineering: Details matter! This paper will focus primarily on the application of high-resolution and/or high-sampling-rate ADCs, and to a lesser degree, DACs, including direct conversion converters. Topics will include power and ground distribution, analog/RF signal paths, digital signal paths, timing, and special considerations for direct conversion. Standards for measuring ADC and DAC performance will also be mentioned.

Speaker: Tom Linnenbrink (Teqnovations)

09:20 → 10:20 Session 1: Operations

09:20 A LLRF implementation to Extend the Present LHC RF System to High-Lumi Beam Currents

The LHC RF/LLRF system is currently setup to achieve extremely stable RF voltage (in amplitude and phase) to minimize transient beam loading effects. The available klystron forward power would limit the LHC performance to nominal beam current with the present operational scheme. For beam currents above nominal (and possibly earlier), the cavity phase modulation by the beam (transient beam loading) will not be corrected, but the strong RF feedback and One-Turn Delay feedback will still be active for RF loop and beam stability in physics. To achieve this, the voltage set point should be adapted for each bunch. This work presents the theoretical background, simulation studies, test-bench evaluation, and measurements with beam in the LHC of a new iterative algorithm that adjusts the voltage set point to achieve the optimal cavity phase modulation which minimizes klystron forward power requirements.

Speaker: Dr Themistoklis Mastoridis (California Polytechnic State University)

09:50 Cavity autorecovery with beam

ALBA is a 3GeV synchrotron light source located in Barcelona and operating with users since May 2012. The RF system of the SR is composed of six cavities, each one powered by combining the power of two 80 kW IOTs through a Cavity Combiner (CaCo). At present, there are several RF interlocks per week. The redundancy given by the six cavities makes possible the survival of the beam after one of these trips. In these cases, the cavity has to be recovered with the circulating beam. An autorecovery process has been implemented in the digital LLRF system in order to recover the faulty RF plant after a trip without affecting the beam. The stages of this automatic process, how to adjust the system and future upgrades will be also presented.

Speaker: Ms Angela Salom (CELLS)

10:20 → 10:40 Break

Ref: group p

10:40 → 11:25 DSP Tutorial Part III: Avoiding Resource Overutilization

Speaker: Lawrence Doolittle (LBNL)

11:25 → 12:15 Session 2: Operations

11:25 Multiharmonic beam loading compensation in the J-PARC synchrotrons

Beam loading compensation is a key for acceleration of high intensity proton beams in the J-PARC synchrotrons, the rapid cycling synchrotron (RCS) and the main ring (MR). In both rings, magnetic alloy (MA) loaded rf cavities are employed to achieve very high accelerating voltages. The Q values of the MA cavities of the RCS and the MR are set to 2 and 22, respectively, so that the frequency responses cover the frequency sweep of the accelerating rf to follow the velocity changes of the proton beams without tuning bias loops. Furthermore, the wide-band (Q=2) cavities of the RCS are driven by dual-harmonic rf signals for bunch shaping to mitigate the space charge effects. On the other hand, the wake voltage in the RCS cavity consists of not only the accelerating harmonic, but also the higher harmonics. The higher harmonic components of the wake are source of the bucket distortion. In case of the MR, the cavities are driven by the single harmonic rf, however, the neighbor harmonics are within the cavity frequency response. The neighbor harmonics are source of the periodic transient effects and they are a possible source of coupled bunch instabilities. Therefore, multiharmonic beam loading compensation is necessary for both of the RCS and the MR. We employ the rf feedforward method. We developed multiharmonic feedforward systems for the RCS and the MR. We present the configuration of the feedforward system. The system uses the I/Q demodulation and modulation technique and it works essentially as a tracking bandpass filter. Although the system is rather simple, the commissioning of the feedforward is not trivial. We developed the commissioning methodology of the feedforward. We present the commissioning methodology, the commissioning results, and the beneficial effects of the feedforward compensation in the beam operation. The feedforward compensation is now indispensable for high beam power operation of the RCS and the MR, at several hundred kilowatts.

Speaker: Dr Fumihiko Tamura (J-PARC Center)

11:50 Beam-Based Feedback for the European XFEL

Pump-probe and seeding experiments at free-electron laser facilities such as the European XFEL are highly dependent on precise regulation of bunch arrival time and bunch compression. Combined RF field and beam-based measurement and feedback allow to fulfill these requirements. This contribution shows the beam-based feedback implementation for XFEL and current achievements and limitations at FLASH. Furthermore, the large amount of accelerating modules within the FEL main linac requires a beam energy management. An outline of slow and fast beam energy feedbacks, which will further improve the beam performance, is given.

Speaker: Mr Sven Pfeiffer (DESY)

12:15 → 12:30 Short presentations from student posters

12:30 → 12:45 Group Photo

12:45 → 14:00 Lunch - Granhall

14:00 → 18:00 Posters, Exhibitors

18:00 → 19:00 Exhibitors and Sponsors Reception - Granhall

19:00 → 20:30 Dinner - Granhall

Friday, 4 October

08:30 → 08:35 Announcements

08:35 → 09:20 DSP Tutorial Part IV: System Simulation From Design Through Commissioning, Stefan Simrock, ITER

08:35 DSP Tutorial IV: System Simulation From Design Through Commissioning

09:20 → 09:29 In Memoriam: Ron Akre

09:30 → 10:20 Session 1: SRF & Piezo

09:30 Performance of CW superconducting cavity at ERL test facility

The 35 MeV compact Energy Recovery Linac (cERL) is a test facility of the future 3-GeV ERL project at KEK. At present, the RF system for a buncher cavity and total of 3 injector cavities were constructed. A FPGA-based LLRF system was established to stablize the RF field. The 0.1% and 0.1° stabilization goals can be achieved by the closed-loop operation. A gain scanning experiment for determining the optimal gains was carried out to improve the performance of the system. Furthermore, main system parameters (i.e. loop delay and bandwidth) were identified by modern system identification method. According to the current closed- loop experiment, The optimal RF stability was 0.01% RMS for amplitude and 0.02° for phase for the injector cavities while the corresponding beam energy stability was 0.006% RMS. In this presentation, we describe the current status and performance of the RF system at cERL.

Speaker: Dr Feng Qiu (KEK)

09:55 RF Operational Experience with High QL CW SRF Cavities

The CEBAF 12GeV energy upgrade cavities have a QL of 3x107 and operate at gradients of 20 MV/m. Operation of these cavities is challenging due to their microphonic sensitivity, cavity to cavity mechanical coupling and Lorentz force detuning. Before our last shutdown, 16 of these cavities (two cryomodules) were operated for six months in the CEBAF LINAC. During this run field control, fault recovery and start up routines were optimized and improved. This presentation discusses these issues including SEL to GDR, fault recovery, and resonance control.

Speaker: Mr Ramakrishna Bachimanchi (Jefferson Lab)

10:20 → 10:40 Break

10:40 → 12:00 Session 2: SRF & Piezo

10:40 Development and Test of Digital LLRF Control Procedures and Techniques in Scope of ILC

In order to operate the superconducting cavities at the International Linear Collider (ILC) near their maximum gradients, cavity input (PK) and cavity loaded Q (QL) have to be controlled individually (PKQL control). In this scope a fully automated PKQL operation procedure was developed and demonstrated at cavity gradients of 16 MV/m and 24 MV/m with Q_L values of 9e6 and 3e6 at the linear electron accelerator at the Superconducting RF Test Facility (STF) at the High Energy Accelerator Research Organization (KEK). During a long-time operation with beam (6.4 mA, 615 μs) the vector sum gradient and phase stabilities during the beam transient were ΔA/A_RMS= 0.009% and Δϕ_RMS=0.009° with cavity gradient stabilities of ΔA/A_(cav1,RMS)=0.041% and ΔA/A_(cav2,RMS)=0.031%. Since in ILC the cavity gradient spread will be±20% around 31.5 MV/m the required range of loaded Q values is 3e6 to 1e7. High loaded Q operation at QL = 2e7 with a 6.1 mA beam was demonstrated at STF. The stabilities were ΔA/A = 0.008%RMS and Δφ = 0.014°RMS. Furthermore a near klystron operation within 5% of saturation, which is an ILC requirement, was performed at STF with a 6.2 mA beam. The stabilities were ΔA/A = 0.010%RMS and Δφ = 0.009°RMS. An FPGA-based klystron linearization algorithm for amplitude including an in the I and Q plane circular limiter was developed, implemented, and successfully tested at New Muon Laboratory (NML) at Fermi National Accelerator Laboratory (FNAL).

Speaker: Mr Mathieu Omet (Graduate University for Advanced Studies)

11:05 Application of Active Disturbance Rejection Control in Superconducting Radio Frequency Cavities

In this talk, the previous results on the mitigation of microphonics using Active Disturbance Rejection Control (ADRC) is briefly reviewed first to show the benefit of this new control method, followed by the introduction of ADRC in more details. Then the application of ADRC at Facility for Rare Isotope Beams (FRIB)/ National Superconducting Cyclotron Laboratory (NSCL) to other problems, such as beam loading are addressed. Recent work on the implementation of ADRC in FPGA and future work on improving tuner control will be discussed as well.

Speaker: Dr Shen Zhao (Facility for Rare Isotope Beams, Michigan State University)

11:30 Piezo characterization, tests and operation at FLASH

The superconducting cavities operated at high Q level need to be precisely tuned to the RF frequency. The TESLA cavities at FLASH accelerator are tuned using slow (step motors tuners) and fast (piezo tuners) driven by the control system. The goal of this control system is to keep the detuning of the cavity as close to zero as possible even in the presence of disturbing effects (Lorentz force detuning and microphonics). The fast frequency tuner makes use of a pair of multilayer piezoelectric (piezo) actuators to drive its fast detuning compensation action. The presentation covers the discussion on piezo operating environment and parameters (the choice of unipolar or bipolar mode of operation), the piezo's parameters important for the control system, piezos characterization and long term test results. The piezo operation at FLASH will be also reported.

Speaker: Dr Mariusz Grecki (DESY)

12:00 → 12:15 Topical Discussions, Piezo Control math

12:16 → 12:30 Topic Discussion, Quirks of various high power RF sources

12:30 → 14:00 Lunch

14:00 → 14:05 Announcements

14:05 → 14:55 Session 3: Phase Calibration

14:05 Properties of a Distortion-Compensating Phase Calibration Processor

When designing a downconverting RF measurement system, there is a perennial tradeoff choosing the signal level at the mixer. Too low, and Johnson noise at its output limits the precision of the measurement. Too high, and distortion generated in the mixer (characterized by its IP3) causes systematic errors in the measurement. Superimposing calibration tones (also known as pilot tones) on the input, as is necessary for cable drift compensation, is thought of as aggravating the situation. Here, we present and analyze a digital-signal-processing (DSP) mechanism that can characterize and correct distortion of a two-tone (pilot tone plus unknown signal) system. This opens up the possibility of running mixers with higher signal strengths than before.

Speaker: Lawrence Doolittle (LBNL)

14:30 XFEL MO and RF Phase Reference Distribution

One of the most important requirements for the European XFEL RF system is to assure a very precise RF field stability within the accelerating cavities. The required amplitude and phase stability equals respectively dA/A <3E-5, dphi<0.01 deg @ 1.3GHz in the injector and dA/A<1E-3, dphi <0.1 deg @1.3GHz in the main LINAC section. Fulfilling such requirements for the 3.4 km long facility is a very challenging task. We describe the proposed architecture of the RF Master Oscillator and the Phase Reference Distribution System designed to assure high precision and reliability. A system of RF cable based interferometers supported by femtosecond-stable optical links will be used to distribute RF reference signals with required short and long term phase stability. We also present test results of prototype devices performed to validate our concept

Speaker: Dr Krzysztof Czuba (ISE, Warsaw University of Technology)

14:55 → 15:00 Closeout