C-band Structures at SPARC and Overview of
C-band Technology at other International Projects
D. Alesini
(LNF-INFN, Frascati, Italy)
OUTLINE
1.
WHY C-BAND FOR ACCELERATORS?
2.
C-BAND @ SPARC
3.
C-BAND ACCELERATORS OVERVIEW (PSI, SCSS) AND C-BAND TECHNOLOGY
4.
NEXT STEP:
-STRONG DAMPED C-BAND ACCELERATING STRUCTURES: ELI_NP
-C-BAND GUN DESIGN AND FULL C-BAND LINAC
WHY C-BAND FOR ACCELERATORS?
We refer to high gradient room temperature pulsed electron LINACS with typical parameters:
10-100 Hz rep. rate, 0.5-5 s RF pulse length, single/multi bunch, 20-100 MV/m average accelerating gradients.
S-BAND (2.856 GHz)
X-BAND (12 GHz)
C-BAND (5.712 GHz)
frequency
NLC-CLIC projects:
SLAC LINAC
3 m long sections
3.1 km total accelerator length
960 accelerating structures
up to 50 GeV electron energy
20 MV/m average acc. gradient
Accelerating gradient (f1/2)
Dipole wakefield intensity (f3)
Complication in fabrication technology
Available commercial components
0.5-1 m long sections
up to 100 MV/m acc. gradient
C-BAND ACCELERATING SYSTEM @ SPARC
The energy upgrade of the SPARC photo-injector at LNF-INFN from 150 to more than 240 MeV will be done by replacing a
low gradient S-Band accelerating structure with two C-band structures. The structures are TW and CI, have symmetric axial
input couplers and have been optimized to work with a SLED RF input pulse. In the SPARC photoinjector the choice of the Cband for the energy upgrade was dictated by the opportunity to achieve a higher accelerating gradient, enabled by the
higher frequency, and to explore a C-band acceleration combined with an S-band injector that, at least from beam
dynamics simulations was very promising in terms of achievable beam quality.
2 structures
1.4 m long
>35 MV/m acc. Gradient
Design and built @ LNF
SLED-SKIP RF
compression
system (IHEP,
Beijing)
Low gradient SBand structure
13 MV/m
S-Band SLAC-type structure
22 MV/m
S-Band gun 120 MV/m
DESIGN OF C-BAND TW STRUCTURES FOR SPARC
(D. Alesini, et al, JINST, 8, P05004, 2013)
Structure design criteria:
-CONSTANT IMPEDANCE (all equal irises) to simplify
the fabrication and to reduce the unbalance between
the accelerating field at the entrance and at the end
of the structure, due to the combination of power
dissipation along the structure and SLED pulse profile.
-LARGE IRISES WITH ELLIPTICAL SHAPE to
-reduce the peak surface field obtaining
at the same time an average accelerating
field >35 MV/m with the available power
from the klystron;
-reduce the filling time of the structure
and, consequently, the RF input pulse
length thus reducing the breakdown rate;
-reduce the dipole wake intensity
-increase the pumping speed.
-WAVEGUIDE COUPLER design based on “low pulsed
heating” couplers for high gradient operation of X
Band structures (SLAC).
TEST AT HIGH POWER (@KEK) OF THE PROTOTYPE
The high-power test started on November 5, 2010 and was completed on
December 13, 2010. For almost one month of processing, from
November 5 until December 2, more than 108 RF pulses of 200 ns width
were sent into the structure with a repetition rate of 50 Hz. For a couple
of days the RF pulse length was changed to 300 ns and for one day
(November 12) the repetition rate was decreased to 25 Hz. On November
15, SKIP was switched on.
After the high power test the structure has been cut in slices for
an internal inspection. We have identified the signs of craters and
discharges mainly in the first accelerating cell after the input
coupler, as expected, because the highest field values are excited
at the beginning of CI structures.
FINAL C-BAND STRUCTURES
First fabricated C-band structure
PARAMETER
prototype final structure
Frequency
5.712 [GHz]
Phase advance per cell
2/3
Number of accel. cells
22
71
Structure length
0.54 [m]
1.4 [m]
Group velocity/c
0.0283
Field attenuation
0.206 [1/m]
Series impedance (Z)
34.1 [M/m2]
Shunt impedance (r)
82.9 [M/m]
Filling time
50 [ns]
150 [ns]
Surf. peak E field/Acc. field
2.17
Surf. Peak H field@ 35 MV/m
87.2 [kA/m]
Pulsed heating @ 35 MV/m
<1 oC
Av. dissipated power @ 10 Hz 7.6 [W]
59.6 [W]
(D. Alesini, et al, JINST 8, P10010, 2013)
MAIN PROBLEMS ENCOUNTERED IN THE STRUCTURE
FABRICATION
The main problems were related to the fact that we do not have a vertical >1.5 m long oven for brazing and we had to
braze the structures in several steps.
Each brazing step is a structure “stress”, requires a full control of the process and introduces unknowns since the success
of the brazing cannot be guarantee at 100% even with a long experience.
The design, machining and brazing of a new (complicated) structure (such as the SPARC C-band structures), require a
strong activity of R&D and prototyping at least in the first phase to investigate all possible criticalities (RF, mechanical).
For the SPARC C-band structures, we have fabricated only one small prototype before starting the construction of the first
complete structure and this is the reason why we had to re-machine and re-cut the first structure a couple of time. In
other words the first structure has been a prototype.
At LNF-INFN this experience has been the first experience of a complete in-house design, realization and test of a such
long multi-cell TW structure. We gain a lot (a lot) of experience thanks to this work.
First realization technique
New design of the junction, re-machining and brazing
OK
C-BAND HIGH POWER STATION @ SPARC
Toshiba ET37202 klystron and solid
state modulator by Scandinova
The new C-band power station will consist mainly of:
• C-band klystron, manufactured by Toshiba Ltd (JP)
• Pulsed HV modulator supplied by ScandiNova (S)
• WR187 waveguide system with double resonator
pulse compressor.
• 500 W solid state klystron driver supplied by
MitecTelecom (CDN)
Test stand for high power test
Frequency
Output RF power
RF pulse length
Pulse rep. rate
Gain
Efficiency
Drive power
5712 MHz
50 MW (max).
2.5 μsec
50 pps max.
44 dB min
40 % min
300 W
C-BAND STRUCTURES: HIGH POWER TEST BENCH @SPARC
Ceramic windows
and power splitter
From KLY
RF pickup
FW-RW
RF Pickup
TRANS
C-Band structure
RF Loads
Ion pumps
C-BAND SLED LOW POWER MEASUREMENTS
The SKIP SLED has been fabricated in IHEP (Beijing). It has been fixed to the
SPARC experimental hall ready for waveguide connection.
HIGH POWER TEST @SPARC: RESULTS (1/2)
1) The RF conditioning has been done in three steps:
test of the Klystron system terminated into a loads
test of the waveguide system up to the SPARC hall terminated
into a load
test of the first accelerating structure
Typical event of discharge
2)The high power test on the first C-band structure started on Nov. 2013.
We operated at 10 Hz with the nominal pulse width of 165 ns (slightly
longer than the filling time of the structure).
3)We progressively increased the power from the klystron (increasing the
HV of the modulator) monitoring:
the current absorption of the 4 ion pumps (3 connected to the
structure and 1 to the waveguide);
the RF signals from pickups.
Control panel
RF reflected
RF transmitted
RF forward
HIGH POWER TEST @SPARC: RESULTS (2/2)
1) The conditioning procedure was semi-automatic and the interlocks on the HV were forced by:
operator
ion pumps current absorption above a certain threshold (50 A corresponding to a vacuum of 10-7 mbar)
including the ion pump absorption directly connected to the KLY output;
KLY interlocks (tube vacuum, modulators interlocks);
2) We normally operate at a vacuum level in the structure between 510-10 mbar and 210-9 mbar.
3) The duration of the RF conditioning (not h24) was about 10-15 full days equivalent! We have finally reached:
 38 MW input power in the structure (44 MW from the klystron), nominal rep. rate and pulse length.
 the corresponding accelerating field was 36 MV/m peak and 32 MV/m average
 BDR <10-5 but even less because a correct measurements of the BDR require a long time and we want to
test the other structure!
 340 KV modulator voltage
C-BAND STRUCTURES FOR SWISS FEL PROJECT
1st construction phase
2013-16
BC1
Injector
2nd construction phase
Athos 0.7-7nm
2018-19
Linac 1
0.35 GeV
2.6-3.4 GeV
BC2
Linac 2
2.0 GeV
Key Parameters:
FEL Wavelength: λ = 1 to 7 Å
Electron Beam Energy: 5.8 GeV max.
Main Linac frequency: 5.712 GHz
Bunch charge: 200 pC or 10 pC
Bunch length: 25 fs or 0.3 fs (ultrashort case)
Core slice emittance (mm·mrad) 0.43
Photon peak brightness* ~1033
Bunch per pulse 2
Bunch spacing 28 ns
Total Length: 715 m
3.0 GeV
user
stations
Linac 3
2.1-5.8 GeV Aramis 0.1-0.7 nm
Main hardware
component: C-band Linac
module x 26
Courtesy A. Citterio and R. Zennaro (PSI)
PSI C-BAND STRUCTURES: REALIZATION
C-band technology:
In house development of ultra-precise machined
accelerating structure without tuning (short
structure program and 2m nominal structure)
In house development of the brazing technique
for the 2m structure
J-type coupler
Specifications:
Phase adv. 2π/3
Filling Time:
vg/c:
329 ns (th.)
113 cells,
constant gradient
3.1% - 1.2% (th.)
Iris radius (20°C): 7.238 mm – 5.447 mm
Length : 2 m
Accelerating gradient 28 MV/m
Courtesy A. Citterio and R. Zennaro (PSI)
SPRING-8 COMPACT SASE SOURCE (SCSS)
A reduction of the machine size for widespread distribution of XFEL sources is the main idea for the SPring-8 compact SASE
source (SCSS). Here, the use of an in-vacuum shorter-period undulator combined with a high gradient C-Band accelerator
allows us to realize a compact XFEL machine with a lower-energy accelerator. Although a reduction of the facility scale gives
a significant advantage, a new technical challenge was requested for generating and accelerating the extremely high-quality
electron beam with a small normalized-slice emittance of less than 1 mm mrad. For this purpose, we proposed to use a
thermionic cathode gun, which has a stable emission property and a long lifetime compared to photocathode rf guns. In
order to make up for its low emission current, a velocity bunching section, which can push the total bunch compression
factor up to a few thousands, was added to the injector.
-Nominal accelerating gradient up to 35 MV/m
-length 1.8 m
-C-Band damped structures in operation
C-BAND TECHNOLOGY
The C-Band technology can be considered “commercial”. Several devices are available on the
market and their cost (in particular for waveguide components) is even lower than for S-Band
Klystrons (TOSHIBA)
structures
RF pulse compression BOC (PSI)
Waveguides standard components
RF pulse compression SKIP
NEXT STEP: DAMPED/HIGH GRADIENT/HIGH REP. RATE
STRUCTURES FOR ELI_NP
In the context of the ELI-NP Research Infrastructure, to be
built at Magurele (Bucharest, Romania), an advanced Source
of Gamma-ray photons is planned, capable to produce beams
of mono-chromatic and high spectral density gamma
photons. The Gamma Beam System is based on a Compton
back-scattering source. Its main specifications are: photon
energy tunable in the range 1-20 MeV, rms bandwidth smaller
than 0.5% and spectral density larger than 104
photons/sec.eV, with source spot sizes smaller than 10-30
microns.
For this LINAC high gradient/high rep rate and damped
structures (for multi-bunch operation) are required to allow
an high gamma flux and a compact source.
Bunch charge
250 pC
Number of
bunches
32
Bunch distance
16 ns
C-band average
accelerating
gradient
33 MV/m
Norm.
emittance
0.2-0.6
mmmrad
Bunch length
<300 m
RF rep Rate
100 Hz
FIRST REQUIREMENT: HOM DAMPED STRUCTURES
ELI-NP operates in multi-bunch mode. The passage of electron bunches through accelerating structures
excites electromagnetic wakefield. This field can have longitudinal and transverse components and,
interacting with subsequent bunches, can affect the longitudinal and the transverse beam dynamics.
TRANSVERSE EFFECTS  Cumulative beam break-up (BBU)
LONGITUDINAL EFFECTS  beam loading
Eacc=average accelerating field
Ideal axis
Injector S-band
Einj=beam energy
after injector
Booster LINAC (C-band)
Normalized Courant Snyder Invariant at the
exit of the linac for an initial displacement of
all bunches of 500 m
DAMPING OF DIPOLE MODES CHOICE
Dipoles modes
propagate in the
waveguide and dissipate
into a load
CLIC structures X-band, high gradient
C-Band structures Spring-8
Advantages
Advantages
1. Strong damping of all modes above
waveguide cut-off
2. Possibility of tuning the cells
3. Good cooling (high rep. rate)
1. Easy machining of cells (turning)
2. 2D geometry: no multipole field
components
Disadvantages
Disadvantages
1. Machining: need a 3D milling machine
2. Multipole field components (octupole)
but not critical at least for CLIC
1. Critical e.m. design: notch filter can reflect
also other modes.
2. Not possible to tune the structure
4. Cooling at 100 Hz, long pulse length (?)
DAMPING: HFSS AND GDFIDL SIMULATIONS
w=13 mm
w=7 mm
w
First dipole mode
passband
ACCELERATING STRUCTURES DESIGN
The power released by the beam on the dipole modes is dissipated into
SiC absorbers. Several different solutions are possible to design the
absorber. The final geometry has been optimized to:
- simplify the realization procedure and the overall cost of the
structures.
- Reduce the transverse size dimensions to allow solenoids positioning
absorber
PARAMETER
VALUE
Type
TW- quasi CG
Frequency (fRF)
5.712 [GHz]
Phase advance per cell
2/3
Structure Length
1.8 m (102 cells)
Iris aperture (a)
6.8-5.8 mm
group velocity (vg /c):
0.034-0.013
Quality factor (Q)
8800
Shunt imp. (r)
RF input power
EACC_average@ PIN=40 MW
Rep. Rate (frep)
Average dissipated power
67-73 [M/m]
40 MW
33 MV/m
100 Hz
2.3 [kW]
first solution (CLIC-type)
absorber
PROTOTYPES REALIZATION
First prototypes have been realized to verify the feasibility of the machining of
the cells and of the brazing process. We are now focalizing in the realization of
two prototypes previous the realization of the first complete structure.
-The first prototype is a full scale device without precise internal dimensions
that we would like to build in order to test the full brazing process, verifying
eventual structure deformations and vacuum.
-The second prototype is a device with a reduced number of cells that we would
like to realize to test the RF properties of the structure at low and high power.
2nd STEP: A FULL C-BAND LINAC
All existing (under realization) C-Band LINACS have an S-band injector. A full C-band linac with a
C-Band RF gun is the next and definitive step in C-Band photoijnectors.
SPARC
S-Band injector
C-Band Booster
PSI SWISSFEL
C-BAND LINAC
SCSS- SPRING 8
C-BAND RF GUN
The designed gun integrate a waveguide coupler that allows:
-high efficiency cooling of the accelerating cells
-low pulsed heating of the coupler surfaces
-arbitrary solenoids position around the accelerating cells and on the
beam pipe
-100 Hz operation in multi bunch
-fabrication of the gun without brazing processes (hard copper->higher
gradients)
PARAMETERS
f [GHz]
5712
Q0
10900
Ecathode @10 MW Pin
200 [MV/m]

2
Filling time 
200 [ns]
NEXT STEP: $/€/time/people for the first prototype realization
A. Bacci
CC
PARAMETER
Charge
Laser pulse length
laser spot size
Output energy
Output emittance
Output bunch length
Output energy spread
42 MW, 100 Hz
TOSHIBA
E37210
(Nominal output
power 50 MW)
0
5
10
15
20
25
30
35
40
C:\DOC_LAVORO\SPARC\20_SOLENOIDS\01_GUN\02_CBAND_GUN\03_NEW_SOL_CGUN_8CM_FROM_CATHODE_R2.4\SGUNCBAND_Z8R24_NO0.AM 2-22-2012 15:20:20
-10
-8
-6
35 MW
-5
0
2 -24 -2 012
25
14: 25 :44
-4
-2
0
2
4
6
8
10
7 MW
5
10
15
0
5
10
15
20
LA VO RO\ SP ARC \2 0_ SOL EN OID S\ 01_ GU N\0 2_ CBA ND _GU N\ 04 _NE W_ SOL _C GUN _O NCE LS _R6 .5 _A ND_ R2 .4\ NE W_S OL ENO ID \AL L_ SOL EN OI DS_ 1. AM
20
-8
adjustable, Gun C-band solenoid, N=34, ++++
-6
-4
-2
0
2
4
6
8
10
adjustable, Gun C-band solenoid, N=34, ++++
-10
-5
0
5
10
15
20
POSSIBLE CONFIGURATION OF A MULTI-BUNCH C-BAND INJECTOR
C-BAND INJECTOR
250 pC
8.5ps
250 m
95 MeV
0.25 mm mrad
800 fs
0.38%
1.1 s
Flat top
32 bunches
250 pC
16 ns bunch spacing
L=1.8 m
Eacc=31 MV/m
E_cathode=170 MV/m
60 MeV
beam
Beam loading
compensation
CONCLUSIONS
1. C-Band adventure started @ LNF for the SPARC energy upgrade-single bunch operation.
2. The first prototype has been tested at gradients >50 MV/m
3. The two final C-band structures have been designed and fabricated. The first one has been
already tested at high power up to 32 MV/m average accelerating gradient. The second one
is now under installation. The high power tests will start next December.
4. C-Band accelerators developed all over the world (PSI, SPRING8) have made the C-Band
technology “accessible and commercial” in term of waveguide components, RF
compressors, RF sources.
5. Damped C-band structures for multi-bunch acceleration with >100 Hz rep. rate have been
developed for ELI-NP and are now under construction. They are the next step in C-Band
accelerating structure technology and their interest goes above the ELI_NP proposal .
6. The full C-band injector (>100 Hz, Multi-bunch) is now the next and definitive step in this
technology. It includes a C-Band RF GUN at gradients >180 MV/m that has been designed
(RF+mechanical) but has to be fabricated and tested if we want to go in this direction.
THANKS TO…
V. Lollo, R. Di Raddo, P. Chimenti, R. Boni, M. Ferrario, R. Clementi, M. Bellaveglia, A. Gallo,
M.E.Biagini, G. Di Pirro (LNF-INFN)
Toshiyasu Higo, Shuji Matsumoto, K. Kakihara (KEK)
M. Migliorati, A. Mostacci, V. Spizzo, S. Tocci, L. Palumbo, S. Persichelli, G. Campogiani (La
Sapienza)
A. Grudiev, G. De Michele, G. Riddone, Silvia Verdu Andres (CERN)
R. Zennaro, A. Citterio (PSI)
…AND THANK YOU FOR YOUR ATTENTION