Testing of CapCav 2 LLRF Development with DESY

1
Testing of Cap-Cav 2LLRF Development with DESY

• Alexander Brandt, DESY, Hamburg
• Current and Future LLRF Developments at DESY
• DESY Low Latency Control Hardware
• Collaboration with FNAL-A0, SMTF and HPTF
  (present and future)

2
DESY LLRF Contributors

• DESY, Hamburg, Valeri Ayvazyan, Alexander Brandt,
  Gerhard Grygiel, Thomas Fröhlich, Olaf Hensler,
  Matthias Hoffmann, Bastian Lorbeer, Frank Ludwig,
  Günther Möller, Kay Rehlich, Stefan Simrock,
  Henning Weddig, ...
• WUT-ISE, Warsaw, Tomasz Czarski, Krzystof Czuba,
  Tomasz Filipek, Wojciech Giergusiewicz, Wojciech
  Jalmuzna, Pawel Kaleta, Waldemar Koprek, Karol
  Perkuszewski, Piotr Pucyk, Ryszard Romaniuk,
  Jaroslaw Szewinski, ...
• TUL-DMCS, Lodz, Wojciech Cichalewski, Piotr
  Cieciura, Mariusz Grecki, Tomasz Jezynski,
  Boguslaw Koseda, Dariusz Markowski, Pawel Pawlik,
  Przemsylaw Sekalski, Bartlomiej Swiercz, Marcin
  Wojtowski, ...
• IHEP, Protvino, Nikolay Ignashin, Sergej Sytov,
  ...
• INFN, Milano, Angelo Bosotti, Rocco Paparella
• KEK, Japan, Shin Michizono, Toshi Matsumoto
• Yerewan Institute, Gevorg Petrosyan, Ludwig
  Petrosyan
• ...

3
LLRF Developments at DESY VUV-FEL

• Running since 2004 (as successor of TTF1,
  1997-2003)
• SC linear injector, 6 undulator sections (down to
  6nm wavelength)
• Pulsed operation (2ms / 1-10Hz)
• 1 nc cathode, 48 sc cavities, 1 transverse
  deflecting cavity
• 5 rf stations (one klystron per 8-32 sc cavities)
• Field stability requirement 10-3 / 0.1
• Designed as user facility and for accelerator
  development
• Ideal testbed for LLRF developments!

VUV-FEL
4
LLRF Developments at DESY VUV-FEL
5
LLRF Developments at DESY XFEL, ILC

• FEL lightsource for sub-nm wavelength,
  commissioning 2012 at DESY
• 1000 cavities, 35 klystrons
• High field stability 10-4, 0.01
• User facility (high reliability)
• Demands for a low-maintenance, radiation
  resistive hardware

XFEL
ILC

• Future project, not yet scheduled (2012-2020?)
• 20000 cavities
• Collider experiment, therefore relaxed field
  requirements
• Very high degree of automation needed (e.g.
  global phase control, recovery automation)

6
Vecor Sum Control Challenges
VM
... (32)
1.3
ADC
DSP
DAC

• Latency in control loop limits feedback gain (and
  therefore field stability)
• ? build faster feedback hardware (Current
  revision SimCon 3.1 FPGA system, 200ns)
• Mechanical / electrical detuning limits
  performance and increases power demand
• ? build fast resonance frequency control system
  (piezo or magnetostrictive tuning)
• Calibration of signals determines precision
  (nonlinear effects)
• ? build transient detection hardware
• Reference and Distribution determines field
  stability
• ? build long range temperature stable reference
  system

7
Further Projects
8-channel downconverter boards (already in
operation)
Bubble Neutron Dosimetry system w/ automatic
readout s/w
Piezoelectric tuner installed at one cavity
Transient detection hardware (test setup)
Box equipped with several field detectors for the
rf gun
8
DESY's Low Latency Control Boards

• SimCon 2.1
• 2004
• 2xADCs, 2xDACs
• Virtex II FPGA
• Successfully tested at DESY 9-cell teststand
  (single cavity)
• Successfully tested at A0
• Optical Gigalink 2.0GB/s mezzanine card
• Measured latency 200ns160ns!

• SimCon 3.0
• 2004
• 8xADCs, 4xDACs
• Virtex II FPGA
• Successfully tested at DESY 9-cell teststand
• Vectorsum Test at VUV-FEL scheduled for May/June
• Optical Gigalink on board (3.125GB/s)
• Measured latency 200ns160ns!

• SimCon 3.1
• 10xADCs, 4xDACs
• Virtex II Pro FPGA with 2 PPC405
• 2x Optical Gigalink
• Schematics finished
• Prototypes expected in July

• SimCon x.x
• 128xADCs, 64xDACs
• Allows complex control algorithm
• Account for additional signals

• DSP C-67
• Successor of C-40
• (1997-2003)
• TI C-67 CPU
• In Operation for VS control since 2004
• 8 Gigalink channels
• No ADCs/DACs on board
• 3us

C-67 DSP Board
SimCon3.1 Block Diagram
SimCon3.0 Board
SimCon2.1 Board
9
LLRF Plans for A0/SMTF (Present / Short Term)

• Refer to Brian Chase talk
• What we have done already
• Commision SimCon 2.1 prototype (1 week in March
  05) at A0
• Achieved high gain (100) already
• Cleaned up DOOCS control system at A0
• What we plan (short term, May '05 until end '05)
• Again set up SimCon 2.1 at A0
• Improve calibration of system
• Test various algorithms
• Establish remote connection DESY ?? FNAL
• Update DOOCS control system at FNAL
• DESY support by G. Grygiel, O. Hensler, K.
  Rehlich
• DOOCS has interface to other systems, e.g. EPICS

10
Tests in March '05
Readout Panel Top I and Q (blue) resp.
Amplitude and Phase (red) of the Cavity Bottom I
and Q setpoint curves
Matlab Control Panel Feedback Gain in the order
of 100
11
LLRF Plans for A0/SMTF (Medium / Long Term)

• Refer to Brian Chase talk
• What we plan (medium term, starting from end '05)
• Continue support on the DOOCS control system at
  FNAL
• Provide further SimCon systems (SimCon 3.1 to
  arrive by the end of this year)
• Provide signal detection hardware
  (downconverters)
• Collaboration with FNAL-staff on algorithm
  development
• What we plan (long term)
• Provide and support further SimCon systems
• Continue support on the DOOCS control system at
  FNAL

12
Summary and Outlook

• Based on the successful experience of TTF1 (first
  machine with VS control) and VUV-FEL, DESY is
  currently in the process of developing the LLRF
  control for XFEL
• Development of LLRF control for XFEL already
  accounts for the requirements of ILC (automation,
  radiation hardness)
• Demonstrated single cavity control at A0 in March
  '05
• Established collaboration with FNAL for FPGA
  software development
• Plan to equip SMTF with next revisions of SimCon
• As a universal digital control system SimCon is
  applicable to many accelerators (in principle it
  is not restricted to LLRF control only)

13
(No Transcript)