Download 4 - INFN-LNF

Proton Drivers: Prospects in the US
G. Apollinari, Fermilab
Batavia, IL – PO Box 500, USA
1. Introduction
Recent discoveries of non-zero neutrino mass
and neutrino oscillations have led to a worldwide
resurgence of interest in neutrino physics. The basic
questions, addressed in more detail by other
contributions to these proceedings, can be
summarized in the following way:
 What are the masses of the neutrinos ?
 What is the pattern of mixing among the
different types of neutrinos ?
 Are neutrinos their own antiparticles ?
 Do neutrino violate the CP symmetry ?
 Are there sterile neutrinos ?
 Do neutrinos have unexpected or exotic
properties ?
 What can neutrino tell us about the models
of new physics beyond the Standard Model ?
Neutrino physics studies by APS [1] have
concluded that a multi-megawatt Proton Driver is an
essential element in the study of neutrino oscillations
in any foreseeable scenario. The need for a multimegawatt beam can be easily justified by realizing
that the FNAL Main Injector 0.2 MW beam will
deliver 1013 proton/sec on target, determining a rate
of ~ 10-5 neutrino interaction/kton on a long baseline
experiment located at ~700 km from the source.
The US accelerator community has prepared
two proposals to address the need for a multimegawatt Proton Driver: a proposal for an AGSbased Super Neutrino Beam Facility at Brookhaven
National Laboratory, and a proposal for an 8-GeV
Superconducting Proton Linac used as injector to the
Main Injector at Fermi National Accelerator
Laboratory. This contribution will describe the
accelerator aspects of these two proposals.
2. Super-Neutrino Beam Facility (BNL)
The BNL Super Neutrino Beam [2] Facility
proposal is based on the AGS, a complex that has
achieved world intensity records in number of
accelerated protons (7x1013) in a single pulse. The
requirements for the super neutrino beam are
summarized in Table 1 and a schematic drawing of
the upgraded AGS is shown in Figure 1.
The upgrade is based on increasing the
repetition rate of the AGS and reducing beam losses.
To minimize the injection time to about 1 ms, a 1.2
GeV linac will be used instead of the AGS Booster.
This linac consists of the existing warm linac of 200
MeV and a new superconducting linac of 1 GeV.
The H- injection from a source of 30 mA and 720 s
pulse width is sufficient to accumulate 9 x 10 13
particles per pulse in the AGS. The minimum ramp
time of the AGS will be upgraded from 0.5 s to 0.2
s to reach a repetition rate of 2.5 Hz.
Average Beam Power (MW)
Beam Energy (GeV)
Average Beam Current (A)
Cycle Time (s)
Number of Protons per Fill
Number of Bunches per Fill
Protons per Bunch
Number of Injected Turns
Repetition Rate (Hz)
Linac Energy (MeV)
Linac Av./Peak Current (mA)
Linac Emitt. ( mm mr. nor.)
Pulse Length (ms)
Chopping Rate
Present
0.14
24
6
2
7.0 x 1013
12
5.8 x 1012
190
0.5
200
20/30
2.0
0.5
0.70
Upgrade
1.0
28
36
0.4
8.9x1013
23
3.87 x 1012
240
2.5
1200
21/28
1.0
0.72
0.65
Table 1: Performance of the present and
upgraded AGS.
Figure 1: AGS Proton Driver layout
The extracted proton beam uses an existing
beamline at the AGS, but is directed to a target
station atop a constructed earthen hill. The target is
followed by a downward slopping pion decay
channel. This vertical arrangements keeps the
target and decay pipe well above the water table in
this area. A 3-dimensional view of the beam
transport line, target station and decay tunnel is
shown in Figure 2.
2.1 Superconducting Linac
The 1.2 GeV linac required for injection into
the AGS is placed between the exit of the 200 MeV
linac and the AGS injection point, which is 130
meter long. Only a superconducting linac (SCL)
with sufficient gradient can meet the requirement
of acceleration to 1.2 GeV within this distance, and
the BNL design is therefore based on the SNS [3]
project with an accelerating gradient of about 18
MeV/m. Although an upgrade of the warm linac is
foreseen, the most interesting part of the proposal
rests in the SCL section.
Generally, superconducting linac design
involves choices on frequency, cavity velocity,
numbers of cells per cavity, constant energy gain per
cavity versus constant gradient, number of cavities
per cryomodule and type of focusing lattice.
Figure 2: 3-dimensional view of the neutrino
beamline. The beamline is shown without shielding
on top of the beamline magnets and decay tunnel.
The BNL choice for frequency is 805 MHz in
the low energy (LE) section. This is a multiple of
the linac frequency of 201.25 MHz and the same as
the SNS value. For the medium (ME) and high
energy (HE) sections the chosen frequency is 1610
MHz. During the acceleration process from 200
MeV to 1.2 GeV, the particle  varies from 0.57 to
0.89. The BNL design uses  = 0.615 for the LE
section (same as SNS),  = 0.75 for the ME section
and  = 0.85 for the HE section to make acceleration
more efficient by reducing the variation in the transit
time factor. The maximum accelerating gradient in
the SC cavity is determined by the achievable peak
surface field. The chosen mode of operation is to
operate each section of the SCL with the same
energy increment. This requires the same axial field
from one cryomodule to the next. To achieve this,
and to compensate for the transit time variation from
one cryomodule to next, it may be necessary to
locally adjust the RF phase. Also the coupling power
may have to be adjusted according to the local
transit time factor. Four cavities per cryomodule are
chosen for all three sections to minimize the overall
length and maximize the regularities. A focusing
FODO lattice is chosen for the LE section to reduce
the length of the warm insertion, while a quadrupole
doublets lattice, located in warm sections between
the cryomodules, is chosen for ME and HE sections.
The warm sections consist of ~1-1.4 m long spaces
containing quadrupole magnets, horizontal and
vertical steering dipoles, beam diagnostic, bellow,
pumping ports and gate valves. Table 2 gives the
general parameters of the SCL and figure 3 shows
a field simulation of the SCL cells.
Ave. Beam Power (kW)
Average Beam Current (A)
Initial Kinetic Energy (MeV)
Final Kinetic Energy (MeV)
Frequency (MHz)
No. of Protons/Bunch (x108)
Temperature (0K)
Cells/Cavity
Cavities/Cryo-module
Cavity Separation (cm)
Cold-warm Transition (cm)
Cavity Internal Diameter (cm)
Len. of Warm Insertion (m)
Gradient (MeV/m)
Ave. Gradient (MeV/m)
Cavities/Klystron
No. of RF Couplers/Cavity
RF Phase Angle
Method Trans. Focusing
Phase Advance/FODO Cell
Norm. Emitt. ( mm-mrad)
Bunch Area ( MeV)
LE
7.52
36
200
400
805
8.70
2.1
8
4
32.0
30
10
1.079
10.5
5.29
1
1
300
FODO
900
0.8
0.5
ME
15.0
36
400
800
1610
8.70
2.1
8
4
16.0
30
5
1.379
22.9
9.44
1
1
300
Doublets
900
0.9
0.5
HE
15.0
36
800
1200
1610
8.70
2.1
8
4
16.0
30
5
1.379
22.8
10.01
1
1
300
Doublets
900
1.0
0.
Table 2: SCL General Parameters
Figure 3: Field Simulation results for LE, ME
and HE half cells.
2.2 RF Source and Cryogenic System
The SCL RF source system will consists of
klystrons, power supplies, transmitters, circulators
and waveguides, with a basic architecture identical
to all three sections. Each cavity will be driven by a
single klystron through the circulator. The total of
23 cryomodules will have a power coupler for each
cavity, which will be immersed in a helium vessel
at 2.1 K and 0.04 bar. The heat shield will be
cooled between 30 and 50 K and all 23 modules
will be cooled in parallel to allow replacement of a
cryomodule without warming up of the other
modules. The Helium refrigerator is located outside
Table 3: Refrigeration parameters
2.3 AGS Upgrades
In its current operation, the AGS receives four
batches of 1.5 GeV proton beam from the Booster
synchrotron in about 0.5 second. In the proposed
upgrade, the linac can provide ~1012 protons with
injection time of less than one ms. To provide 1 MW
beam power for neutrino production, the AGS has to
be cycled at 2.5 Hz. For this improved capability,
the following major upgrade of the AGS have to be
implemented: a new direct injection from the SCL
with H- stripping foil system, a new main magnet
power supply system, a new RF accelerating cavity
and its associated power switching system to double
the accelerating voltage operated at 2.5 Hz, a new
single turn fast extraction system for beam delivery
to the target and a new collimation and radiation
shielding system to keep the beam losses at an
acceptable level.
Proton Driver Linac - Technology Flow
Other Labs &
Universities
JHF
(KEK)
325 MHz
RFQ and
Klystron
RIA (ANL)
APT (LANL)
SCRF
Spoke
Cavities
Linac
Accel.
Physics
The Proton Driver Linac [4] merges design
concepts and technology from the ILC [5], the
Spallation Neutron Source (SNS) [3], the Rare
Isotope Accelerator (RIA) [6], JPARC, and other
SCRF projects as shown in Figure 4. The design of
the Proton Driver linac has evolved substantially in
response to the ILC technology selection. ILCcompatible operating frequencies (1300 MHz and
325 MHz) have been selected. This means that the
Proton Driver main linac (from ~1.3 - 8 GeV) is an
exact copy of the TESLA design, with identical
cavities, cryomodules, Klystrons, assembly tooling
fixtures, and so on.
The Proton Driver proposal is staged in two
phases, an initial phase with a 0.5 MW linac beam
power at 8 GeV and a second phase with 2.0 MW
linac beam power.
The overall layout and component count of the
Proton Driver Linac is given in Figure 5 with its
three major sections. The Main “TESLA” Linac
from 1.3 - 8 GeV uses 1300 MHz ILC cavities,
cryomodules, and klystrons. The β < 1 section uses
1300 MHz ILC cavities modified for operation with
SNS
Production
Experience
Fast
Ferrite
Shifters
<1
Cavity
Design
TESLA
COLLABORATION
FNAL
ANL / SNS
“SNS / RIA”
H R “PULSED RIA” Beta < 1
_ F SCRF Spoke
Elliptical
Q Cavity Linac Cavity Linac
Pulsed
Modulators
“TESLA”
Elliptical Cavity SCRF Linac
Beta = 1
1300 MHz 8 GeV
1.3 GeV
New FNAL Proton Source Linear Collider Test Facility
PROTON DRIVER
NUMI Beamline &
Infrastructure
Main
Injector
@2 MW
FNAL
Proton Plan
Upgrades
Neutrino
Super-beams
Beam Transport
and Collimation
Design
8 GeV beams:
P, n, , , e…
Technological
& HEP Applications
BNL / SNS
Figure 4: The Technology base for the FNAL
Proton Driver is derived from many SCRF
projects.
0.5 MW Initial
8 GeV Linac
“PULSED RIA”
11 Klystrons (2 types)
449 Cavities
51 Cryomodules
325 MHz
0-110 MeV
Front End Linac
ß<1 TESLA LINAC
1300 MHz
0.1-1.2 GeV
2 Klystrons
96 Elliptical Cavities
12 Cryomodules
10 MW
TESLA
Klystrons
Modulator
Single
3 MW
JPARC
Klystron
Multi-Cavity Fanout at 10 - 50 kW/cavity
Phase and Amplitude Control w/ Ferrite Tuners
H- RFQ MEBT RTSR SSR
Modulator
DSR
Modulator
Elliptical Option
48 Cavites / Klystron
DSR
10 MW
TESLA
Multi-Beam
Klystrons
ß=.47 ß=.47 ß=.61 ß=.61 ß=.61 ß=.61 ß=.81 ß=.81 ß=.81 ß=.81 ß=.81 ß=.81
or… 325 MHz Spoke Resonators
TESLA LINAC 1300 MHz
Modulator
3. Proton Driver (FNAL)
SNS (JLAB)
RIA (MSU)
Klystrons
Shield
35-55 K
~4 bar
4600 W
0W
4600 W
6200 W
~35 %
RF
Distribution
Secondary
4.5 K
3.0 bar
4.6 g/s
2.3 g/s
6.9 g/s
15 g/s
~100%
Cavities
Primary
2.1 K
0.04 bar
645 W
21 W`
666 W
1300 W
~100%
Cryogenics
23 Cryomodules
Temperature
Pressure
Static Load
Dynamic Load
Total Load
Refrigerator Capacity
Margin
non-relativistic protons.
The front end linac
operating at 325 MHz uses a mixture of warm
copper structures and superconducting spoke
resonators modeled on those of the Rare Isotope
Accelerator (RIA) project.
SNS & DESY
the tunnel and the expected heat loads and pressure
requirements for the SCL are given in Table 3.
Modulator
ß=1
8 Cavites / Cryomodule
8 Klystrons
288 Cavities in 36 Cryomodules
Modulator
Modulator
36 Cavites / Klystron
ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1
Modulator
Modulator
Modulator
Modulator
ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1 ß=1
Figure 5: Layout and component counts for the
baseline “Initial” configuration of the Proton
Driver Linac. The 1300 MHz elliptical-cell cavity
option for 110 MeV < E < 400 MeV is shown.
A common feature of proton linacs is that most
of the accelerator physics complexity and
performance risk are in the front end of the linac,
whereas most of the cost is in the back end. Thus
the design emphasis in the Proton Driver Front End
is on conservative beam dynamics, accurate
alignment, and instrumentation; while the emphasis
in the Main Linac is on having a lean and
economical design. This emphasis is shared with
the main linac of the ILC. Table 4 lists the segment
lengths, output energies and number of modules in
each section of the linac.
An outstanding feature of the design is the
small number of Klystrons. Only eleven Klystrons
are required for the baseline linac. This is a
reflection of the efficiency of superconducting RF,
and the use of a TESLA-style RF fan out in which
one large Klystron drives many cavities. Only two
types of Klystron are used, and both types are
already in production for other projects. A single,
standard modulator design is used throughout.
LINAC SEGMENT LENGTHS
Ion Source (H- and P)
Low-Energy Beam Transport (LEBT)
Radio-Frequency Quad (RFQ)
Medium-Energy Beam Transport (MEBT)
Room Temperature Front End
SCRF Single-Spoke Resonator
SCRF Double-Spoke Resonator
SCRF Triple-Spoke Resonator(OPTION)
Low Beta=0.47 elliptical cavity (OPTION)
Medium Beta=0.61 elliptical (OPTION)
High Beta=0.81 SCRF
Beta=1 SCRF
LINAC ACTIVE LENGTH *
Transfer Line to Ring
Tunnel to Front End Equipment Drop
TUNNEL TOTAL LENGTH *
frequency-scaled copy of the 402.5 MHz RFQ for
the LBL/SNS front end.
8 GeV Linac
Length
~0.1 m
~0.1 m
~4.0 m
3.6 m
10.4 m
12.5 m
17.2 m
64.0 m
18.8 m
38.5 m
70.1 m
438.3 m
613.6 m
972.5 m
20.0 m
1606.0 m
Eout
0.065 MeV
0.065 MeV
3.0 MeV
3.0 MeV
15.8 MeV
33 MeV
110 MeV
400 MeV
175 MeV
400 MeV
1203 MeV
8000 MeV
8000 MeV
8000 MeV
# Modules
TBD
2
21
1
2
6
2
4
6
36
47
RFQ modules
Rebuncher Cavities
Room Temp 3-Spoke Resonators
Cryomodules
Cryomodules
Cryomodules
Either 3-Spoke
or Elliptical for
Cryomodules
110-400 MeV
Cryomodules
Cryomodules
Cryomodules
half-cells (quads)
TBD
Table 4: Segment length and output energies of
the Proton Driver linac.
The RF distribution is identical to TESLA for
beam energies above ~2 GeV. Above this energy,
the protons are sufficiently relativistic that the
TESLA passive RF power split with “vector sum
regulation” is effective.
Below this energy,
extending the TESLA RF split technique to proton
linacs required the development of fast high power
“Ferrite Vector Modulators” (FVM’s).
These
provide independent phase and amplitude control on
individual cavities driven from a common Klystron
and they have been the subject of an intense and
successful R&D program in the Proton Driver
project.
3.1 325 MHz Front-End Linac Overview
The front end linac consists of an H- ion source,
Radio-Frequency Quadrupole (RFQ) with a 3 MeV
output energy, a Medium-Energy Beam Transport
(MEBT) section, followed by room-temperature and
superconducting spoke resonators. The front end
runs at 325 MHz (one quarter of the ILC’s 1300
MHz frequency) so that while every RF bucket is
occupied at 325 MHz, only every 4th RF bucket is
occupied in the 1300 MHz main linac. The entire
front end linac (up to 110 MeV) is driven by a single
Klystron. A schematic view of the front-end is
shown in Figure 6.
The H- Ion source will be copied from the
LBL/SNS, JPARC, or DESY designs.
The
specifications for the ultimate configuration (30 mA
x 1 msec x 10 Hz) are already exceeded by the
LBL/SNS and we also collaborated with the SNS to
verify that their source operates well at the
10 mA x 3 msec pulse lengths needed for the Initial
scenario. The 325 MHz RFQ will be a either a direct
copy of the 324 MHz JPARC design, or a
Figure 6: Layout of the 325 MHz Front End
Linac, including the H- Ion Source (IS),
Radiofrequency Quadrupole (RFQ), MediumEnergy Beam Transport (MEBT), RoomTemperature Triple-Spoke Resonators (RT-TSR),
Superconducting Single-Spoke Resonators (SSR),
Double-Spoke Resonators (DSR), Triple-Spoke
Resonators (TSR).
The Medium-Energy Beam Transport
(MEBT) performs matching and analysis of the 3
MeV output beam of the RFQ. Two bunching
cavities (325 MHz spoke resonators) are used to
maintain the longitudinal beam structure. The
MEBT also contains a beam chopper to generate
the Main Injector beam extraction gap, and
optionally also chops the beam to fill the 53 MHz
RF buckets in the Main Injector. Transverse
focusing in the MEBT and following sections is
provided by superconducting solenoids. It is
expected that these will have significantly better
properties of emittance growth and beam halo
formation than the quadrupole focusing normally
used. The RFQ output section has been designed
to produce an axisymmetric (round) beam at its
output to match into this solenoidal focusing.
Additionally, a flexible alignment rail is planned to
allow module-by-module commissioning and
convenient reconfiguration and replacement of
beamline components and diagnostics.
Room-temperature triple-spoke resonators
(RT-TSR) are used to accelerate the beam from 3
to 15 MeV. Each room temperature tank (Figure
7) is individually tailored to the proton velocity as
the beam is accelerated. The room temperature
cavities are interleaved with superconducting
warm-bore solenoids.
Superconducting spoke resonators are used
starting at 15 MeV. The resonators are similar to
those developed for the RIA and APT projects
(Figure 8), but are operated in pulsed mode at
325 MHz. A single cryomodule containing 16
single-spoke resonators and 16 cold-bore
superconducting solenoids accelerates the beam to
33 MeV.
Two more RIA-style cryomodules
containing 14 double-spoke resonators and
solenoids accelerate the beam to 110 MeV. A pool
boiling 4.5 K cryogenic distribution system is
included in each cryomodule. A cold-to-warm
transition and warm gate valve is included at the end
of each cryomodule.
Figure 7: Room-Temperature Triple-Spoke
Resonator (RT-TSR)
the beams from 1.2 GeV to 8 GeV. These 9-cell
cavities are capable of accelerating both protons
and electrons with an accelerating gradient of 26
MV/m. Between 0.4 and 1.2 GeV the linac uses a
single Klystron which drives “squeezed TESLA”
SCRF cavities optimized for protons with
relativistic  = 0.81. Finally, an alternative to the
triple spoke SC cavities described previously in the
0.1-0.4 GeV region are =0.47 and =0.61
elliptical cavities. Both options use six
cryomodules approximately 10 m long driven by a
single modulator and 1-2 klystrons. At our present
level of understanding the beam dynamics of either
choice appear workable, and the final choice will
be based on economic considerations.
As in the TESLA design, transverse focusing
is provided by quadrupoles located between the
cavities.
1300 MHz ELLIPTICAL CAVITY CRYOMODULES: 2-4 TYPES
Beta = 0.47
2 Cryomodules
16 Cavities
Beta = 0.61
4 Cryomodules
32 Cavities
OPTION OF
ELLIPTICAL
MEDIUM-BETA
CAVITES
110 - 400 MeV
Beta = 0.81
6 Cryomodules
48 Cavities
Beta = 1.00
36 Cryomods
288 Cavities
Figure 8: Single Spoke SCRF Resonator,
designed by an ANL-FNAL Collaboration.
In the energy range 110 - 400 MeV the baseline
design uses 325 MHz superconducting triple-spoke
resonators similar to the single and double spoke
resonators of the front-end SCRF linac. A single
modulator feeding two 325 MHz klystrons drives six
10 m long cryomodules containing a total of 42
triple-spoke resonators. Focusing in this section is
provided by superconducting quadrupoles spaced
between every two cavities. An alternative design
for this energy range based on 1300 MHz TESLAstyle elliptical cavities can be considered.
3.2 1300 MHz SCRF Linac Overview
The 1300 MHz part of the Linac is divided into
two sections with different cavity and cryomodule
designs. The =1 Main Linac section (~85% of the
Proton Driver linac) is an exact copy of the TESLA
(TTF3) design for the International Linear Collider.
It contains 8 copies of the TESLA RF unit (36
cavities driven by a single Klystron) and accelerates
Figure 9: 1300 MHz Elliptical Cavity
Cryomodules for the Proton Driver. The Main
Linac uses 36 standard TESLA (TTF3)
cryomodules containing 288 standard β=1 9-cell
TESLA cavities.
The “high beta” section
contains six cryomodules with 8-cell cavities
optimized for protons with β=0.81. The two lower
 cryomodule types are optional and they perform
the same function of the triple-spoke cavities.
3.3 RF System and Fast Ferrite Tuners
R&D
The Proton Driver Linac has two RF systems.
A 325 MHz front-end RF system drives the RFQ
and the Front End Linac. In the initial
configuration, a single modulator and a single 325
MHz, 2.5 MW klystron is used to power the entire
front end linac up to an energy of 110 MeV. A
second RF station (one modulator with 2
Klystrons) feeds the spoke resonator option in the
energy range 110-400 MeV. The use of
superconducting RF reduces the number of
Klystrons needed to cover this energy range by
about a factor of 3-5 compared to a warm-copper
linac. At 1300 MHz, the Main Linac is driven by an
RF system with 9 TESLA Multi-Beam klystrons
(Thales TH-1801 or equivalent). The =0.81 and the
first =1 RF stations use “fast ferrite tuners” to
independently control each cavity. The final 7 RF
stations (above 1.8 GeV) use a passive RF power
split and “vector sum regulation” to control the
beam energy. A standard TESLA modulator
(FNAL/TTF, 17 MW peak, 300kW avg.) is used
throughout. This standardization is a major strength
of the 8 GeV Linac concept: there is a single type of
pulsed power source for the entire linac. The
modulators are built to be reconfigurable for either
the initial (3 msec x 2.5 Hz) or ultimate (1msec x 10
Hz) beam pulse.
E-H TUNER CONCEPT
MICROWAVE INPUT POWER
from Klystron and Circulator
E-H
TUNER
MICROWAVE POWER
IS SPLIT INTO TWO
INSIDE MAGIC TEE,
REFLECTED FROM
TWO ADJUSTABLE
SHORTING STUBS,
AND RECOMBINED
AT OUTPUT PORT.
OUTPUT
TO CAVITY
( STANDARD
SETTING * )
Magic Tee
PHASE CHANGE OF OUTPUT IS PRODUCED BY
SYMMETRIC MOTION OF TWO TUNING ARMS
* “Standard Setting” actually requires g/4 offset between legs
AMPLITUDE ADJUSTMENT
MICRO W AVE INPUT POW ER
from Klystron and Circulator
Reflected Power
(absorbed by circulator)
E-H
TUNER
ATTENUATED
OUTPUT
TO CAVITY
(ANTI-SYMMETRIC
OFFSETS )
ASYMMETRIC OFFSETS OF
SHORTING STUBS CAUSES
FRACTION OF MICROWAVE
POWER TO RECOMBINE INTO
OUTPUT PORT,
PLUS SOME REFLECTED
POWER TO INPUT PORT.
(RELECTED POWER
IS
EATEN BY CIRCULATOR).
Magic Tee
AMPLITUDE CHANGE IS PRODUCED BY
ANTI-SYMMETRIC OFFSETS OF TWO TUNING ARMS
MICROWAVE INPUT POWER
from Klystron and Circulator
E-H
TUNER
ATTENUATED
OUTPUT
TO CAVITY
FERRITE LOADED
SHORTED STUBS
CHANGE ELECTRICAL
LENGTH DEPENDING ON
DC MAGNETIC BIAS.
Ferrite
Loaded
Stub
Bias Coil
ELECTRONIC CONTROL
Reflected Power
(absorbed by circulator)
ELECTRONIC TU NING
W ITH BIASED FERRITE
phase shift of the two branches, the hybrid junction
will cause a fraction of the recombined power to be
sent forward to the cavity, with the remainder of
the power being rejected to RF loads. Figure 10
shows the principle of operation and figure 11
shows some initial R&D results.
Magic Tee
TWO COILS PROVIDE INDEPENDENT
PHASE AND AMPLITUDE CONTROL OF CAVITIES
Figure 10: Principle of operation of waveguidestyle single-hybrid FVM (sometimes called an E-H
tuner or I-Q modulator).
The Proton Driver linac design uses fast, high
power Ferrite Vector Modulators (FVMs) to provide
individual RF phase and amplitude control for each
cavity up to 1.5 GeV. The FVMs are high-powered
versions of devices which are commonly used in
microwave world. The basic concept of the FVM is
to split incoming RF power into two equal branches,
independently phase shift each branch under
electronic control, then recombine each branch in a
4-port hybrid junction. Depending on the relative
Figure 11: Measured phase shift of a 5” long
Coaxial Phase Shifter at 325 MHz. The usable
range of the ferrite stub tuner, corresponding to the
region where RF losses are below 0.2 dB, is about
120 degrees.
4.
Conclusions
We presented the technical proposals for two
high intensity Proton Drivers in the US: the FNAL
Proton Driver and the BNL Super-Beam Neutrino
Facility. The designs are based on proven
principles and existing facilities and therefore carry
a limited technical risk. A future implementation of
one of these two proposal will undoubtedly provide
a superb neutrino beam facility to address all the
questions raised by the most recent discoveries in
the neutrino sector.
References
[1] APS Multi-Divisional Study on the physics of
Neutrinos: http://www.aps.org/neutrino/
[2] “The AGS-based Super Neutrino Beam Facility
– TDR” –BNL-73210-2004-IR
[3] SCRF Linac for the Spallation Neutron Source,
SNS-110020300TR0001R000, Nov. 22, 1999.
[4] Proton Driver Design Study, FNAL-TM 2369
http://www-bd.fnal.gov/pdriver/8GEV/
[5] TESLA Technical design Report –
http://tesla.desy.de/new_pages/TDR_CD/start.html
[6] “Medium  Cavity and Cryomodule
prototyping for RIA” J.D. Fuerst
http:// www.orau.org/ria/r&dworkshop/fuerst.pdf