Transcript Electron Beam measurements @ SPARC Daniele

Electron Beam measurements @
SPARC
Daniele Filippetto
On behalf of SPARC team
LIFE, 20-02-09
[email protected]
SPARC LAYOUT
PC Laser
LIFE, 20-02-09
[email protected]
e- beam detectors:
3 different types of screens:
•
Optical Transition Radiation
Best resolution, no saturation,
Low Photon Yield in the visible region, need precise alignement
•
Single Crystal Ce:Yag
Good resolution, high photon yield centered on 500nm
Saturation, multiple scattering can limit the resolution (thickness)
•
Crom-ox powder
High saturation level, only visible light emitted
Poor resolution (depending on the grain size),
8-12 bit Firewire/ethernet digital camera used as detectors;
Commercial Macro obj. used to image the screen on the CCD
LIFE, 20-02-09
[email protected]
beam imaging: Screens
Chosen screens:
Yag:Ce single crystal with aluminum coating to assure
Conduction ( the material itself is an insulator).
The system resolution is a combination of all these causes:
1.
2.
3.
4.
Yag:Ce
Electron multiple scattering inside the screen;
Bremsstrahlung radiation creation, that generates x-rays causing scintillation
The lens diffraction;
The pixel
size divide
the optical
system
magnification.
single
crystal
has by
best
resolution
(that
is dominated by the optical
system).Linearity assured on the SPARC density range, saturation excluded.
A saturation limit approx. 0.01 pC/um^2 for YAG:CE crystal has been found,
while the maximum for the SPARC case is about 0.18 fC/um^2.
We need to choose the screen thickness to assure the best resolution,
but also the bigger signal possible (bulk effect)
LIFE, 20-02-09
[email protected]
e- beam Detectors:
Resolution 100um
Non intercepting;
BPMs
sensitivity;
beam position dependency error <10-4
3E6 particles resolution (about 500fC)
Non intercepting;
Both low and high energy;
ICT
High sensitivity;
beam dump;
Used just for low energy;
Faraday cup
LIFE, 20-02-09
[email protected]
RF Deflector:
5 cells standing wave, TM110 mode,
2.856 GHz
Input power = 1MW
Shunt Imp.=2.5 MΩ
 z _ RES 
cE e   y _ screen y
V RF cos  RF L
 ˆ

2
    Vy cost   y2   E2   defl
c

2
DEFLECTING
VOLTAGE
y_BEAM
y
Beam
axis
z
z
σz_res = 65fs = 20um
 defl
BUNCH
DEFLECTOR
2
E
SPARC case:
y
y’

' 2
E
L
SCREEN
LIFE, 20-02-09
E
max
 8 10 5 (hyp σy_defl100um)
[email protected]
Gun diagnostics:
2 screens
Faraday
ICT
cup
BPMs
Slit mask
•
•
•
•
•
Feedback on the RF phase and amplitude;
ICT and Faraday cup; (Phase scan)
Yag screen (beam centering, solenoid scan,
th. emittance).
2 orthogonal slit systems (emittance)
2 bpm to control the positon of the beam
injected in the LINAC;
LIFE, 20-02-09
[email protected]
LINAC:
Orbit control and correction:
•BPMs each setion entrance;
•H/V Corrector magnets each section entrance/exit;
Envelope control for emittance minimization:
•“Cromox” at each section exit, to image the beam. Rough
estimate of RF kick, beam envelope all along the LINAC.
•Solenoids to compensate for the emittance growth (usefull expecially
in the velocity bunching setup).
LIFE, 20-02-09
[email protected]
TL:
time
3 points for beam imaging (2 different screens,OTR and Ce:Yag crystal)
~7ps
time
Quadruoles for beam matching in the undulator used to measured the
projected emittance
Dispersive section (dipole) used to measure energy and energy spread (14
degrees,1.07 m trj. radius)
RF deflector: slice emittance, slice energy spread, longitudinal Trace Space.
LIFE, 20-02-09
[email protected]
...NEXT FUTURE...
LIFE, 20-02-09
[email protected]
Cavity BPM
Dipole modes are used for position detection, since their amplitude
depends linearly on the beam position and is zero for a centered beam.
The signals excited in the cavity are coupled into an external circuit, and the
amplitude of this particular mode can be separated in the frequency domain.
In principle, no additional subtraction is needed, the information about the
position is given directly. High resolution because of the large signal per
micron displacement.
N =√Z0kTkΔf
k is the Boltzmann’s constant,
Tk the temperature in Kelvin
Δf = f110/Qext
Resolution down to nm
LIFE, 20-02-09
[email protected]
Electro-Optical Technique:
EO
crystal
pol.
/4
Wollaston
Prism +
Balanced
Detector
fs laser
-
b
e-beam
Ebeam _ THz (r , t )
wd
d 3
  ( )( n1  n2 ) 
n0 r41Ea 1  3 cos 2 ( )
c

Γ is the phase shift between the two polarization
α is the angle between Ethz and the crystal axis
Ea is the Thz field coming with the e-beam
ZnTe or GaP Crystals are used (110 cut).
LIFE, 20-02-09
[email protected]
EOS Single shot schemes
SPECTRAL DECODING
• Simplest implementation
• Limited resolution>600fs.
TEMPORAL DECODING
• Highest resolution demonstrated ~ 50 fs
• Needs complex amplified laser
and laser transport
SPATIAL ENCODING
•Resolution expected to be same as TD
• Low power laser sufficient: e.g.
femtosecond fiber lasers
LIFE, 20-02-09
[email protected]
EO scheme @ SPARC:
/2
Seed
laser
Pol
Diode75
COMPRESSOR
Delay line
BBO
Fiber receiver box
(413x178x90)
AUTO
CORR
EO chamber
concept by
D.Fritz LCLS
nsec
gated
ICCD
ICCD
Menlo
Laser
TC-780
E-beam
SPARC-FERMI collaboration
Cube
beamsplit
/4
/2
Stretcher
O2E conv.box
lens
(300x80x100)
lens
EO
Longitudinal electron bunch shape
Laser-electron
beam jitter,
etalon with fs resolution
mot mirror,
CCD
Fiber link
Periscope
of raw data
for
single shot jitter)
(can be example
used to measure
the
FLAME-SPARC
measurement of short bunchFAST
@DIODE
FLASH
Cylindircal
(temporal
decoding)
Mirror
Thanks to
lens
mirror
Optical
Enclosure
box
M.Veronese
LIFE, 20-02-09
14
[email protected]
Absolute bunch length measurements by incoherent
radiation fluctuation analysis
A new method for bunch length measurement
Photon radiation by charged particles is a stochastic process.
(synchrotron radiation, transition radiation, Cherenkov radiation,
…)
By fixing a bandwidth  , we define a
mode with coherence length:
t 
C
1
2 
The photons radiated by the electrons within a single mode adds
coherently.
LIFE, 20-02-09
[email protected]
Fluctuations as function of number of modes
In a bunch of length σt and zero
transverse size, there are
M=σt /σtc independent modes
radiating simultaneously.
M ~ 10
In this situation, the fluctuation of the energy radiated per
pulse becomes (M combined Poisson processes):
 V

 V



  1  tC  1

M 
2   

2
 
1
2   V V
 

LIFE, 20-02-09

2
N

2
Possibility of
absolute
bunch length
measurements !
[email protected]
A step more toward shorter bunches measurements
In order to minimize the
experimental error is good to
have no more than ~ 100 modes
M
And because   
2 
Shorter bunches relax the
requirements on the bandwidth…
No intrinsic limit in measuring shorter and shorter bunches!
LIFE, 20-02-09
[email protected]
First results:
20
mV/div
Measured
amplitude
2 ns/div
Measurement
range
V
V
 0.111
In our case :   
4
 
ALS - BL 7.2 - Sept. 22, 2006
• Filter: 632.8 nm, 1 nm FWHM
• APD: Perkin Elmer C30902S,
Vbias = -238 V (G~100)
• AMP. Ortec VT 120 G ~ 200
LIFE, 20-02-09

1
V

V

2
ps  20.4 ps
2
 0.4%
N
1.9 GeV - ICS ~ 3.7 mA
Nat. bunch length = 20.3 ps
[email protected]
Conclusions:
• the actual Photo-injector diagnostic has been tested and optimized.
•All the parameters needed for the present machine goal can be measured with
good accuracy.
The accuracy decreases going toward shorter and bunches and lower currents;
Going toward:
More sensitive, precise and accurate tools;
Non interceptive systems;
Using noise may help.
LIFE, 20-02-09
[email protected]