Transcript Slides - Agenda INFN

Timing and synchronization at SPARC
M. Bellaveglia
On behalf of the SPARC/X timing,
synchronization and LLRF group
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Summary
• SPARC general overview
• Synchronization and timing systems
– Overview
– Phase detection techniques
– Feedbacks and PLLs in the system
– Two beams synchronization
• SPARC future plans
– Electrons-photons interaction: SPARC towards LI2FE
– LI2FE synchronization
• Conclusion
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC general overview
PHOTOINJECTOR
LASER
SEEDING
LASER
PLASMON-X
THOMSON
HHG
DGL
PHOTOINJECTOR
UNDULATOR
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC general overview
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
LNF Synchronization lab
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC synchronization system overview
• One optical master
oscillator
• Feedbacks with BW
<<1Hz, ≈5kHz and
≈1MHz to synchronize
the subsystems
• Shot-to-shot (10Hz)
analysis for amplitude
and phase calculation
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC timing system overview
Line synchronization
50Hz
PC laser amplification
pump
1KHz
RF reference
79.33MHz
Trigger box
1KHz
PC laser oscillator
79.33MHz
Frequency divider
10Hz
Machine trigger
distribution
10Hz
RF triggers
RF reference
2856MHz
PC and
seeding lasers
Diagnostics
Frequency divider
1KHz
• Input
• Output
• Electrical
• Optical
• Transduction
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Phase detection – standard mixing technique
Sampling
Board
DtRMS ≈ 55 fs
Tested Sampling Boards:
Console application:
- ADLINK 9812: 12-bit, 4-channels, 20 Ms/s
- ADLINK 9820: 14-bit, 2-channels, 65 Ms/s
- NI PXI 5105: 12-bit, 8-channels, 60 Ms/s
- Real time base band analysis: phase noise
and amplitude measurements
- Possibility in choosing waveform analysis
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Phase detection – Resonant method
DtRMS ≈ 250 fs
• 2142MHz cavity design to avoid RF interference
with accelerating structures
• Cavity exited by very short pulses: (i) e.m. field of the beam or (ii) output of a
high voltage PD excited by the UV laser
• Phase detection is possible using the decay time of some us inside the cavity
• 250fsRMS time of arrival jitter observed in both the laser and bunch monitor
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Phase detection – Deflector centroid jitter
SPARC RF deflector
Image of a SPARC
bunch vertically
streaked on a target
Measurement of the
beam centroid
vertical jitter
t ≈ 150 fs
RMS
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
PLLs - Klystron Fast Phase Lock
Phase shifter
To waveguides
PLL on:
Dt≈77fsRMS
Error amp
From reference
• The phase noise introduced at the RF power
PLL off:
generation level can be reduced by phase
Dt≈630fsRMS
locking the klystron output to the RF reference
with an analog loop: same concept as phase
loops in CW machines;
• Short time available to reach steady state (≈1ms): wideband loop transfer
function (≈1 MHz) required.
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
PLLs – Long term drifts correction
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Two beams synchronization – IR-UV on cathode
IR beam
UV beam
RF GUN
electrons
DEFLECTOR
SCREEN
• Experiment to try to modulate the electron beam during the photo emission
• Delay line in the IR optical transfer line to synchronize the beams
• Coarse synch (same bucket, <360ps): HV photodiode illuminated by the two
beams and 5GHz oscilloscope to see the pulses overlap in time
• Fine synch (“zero” delay): electron extracted by both the beams and
accelerated until the RF deflector
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Two beams synchronization – IR-UV on cathode
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Two beams synchronization – FEL seeding
Undulator
SR
LASER
SEED
SCREEN
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Two beams synchronization – FEL seeding
• Same approach than the IR+UV experiment
• Both the beams are optical and more precisely: (i) the spontaneous radiation of
the electron beam and (ii) the seeding laser pulse
• Coarse synchronization (about 1ns): photodiode and oscilloscope (500MHz BW
is enough)
• Tuning the undulator section for spontaneous we were able to see both the
signals on the oscilloscope and to overlap the electrical pulses
• Also we observed the spectrum of the two beams tuning the spontaneous in
the same wavelength range of the seed (400nm)
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Two beams synchronization – FEL seeding
• Fine synchronization: when the coarse synchronization was
finished, we tuned the undulator sections to let the electron
beam interact with the seed and we slightly moved (5mm) the
optical delay line of the seeding laser
• We immediately observed the seed amplification at the
spectrometer screen @400nm
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC future plans - SPARC towards LI2FE
SPARC nominal parameters
Electron Beam Energy
155
MeV
RMS normalized transverse emittance
<2
mm-mrad
Bunch Charge
1.1
nC
RMS slice norm. transverse emittance
<1
mm-mrad
Repetition rate
10
Hz
RMS total corralated energy spread
0.2
%
Photocathode spot size
1.1
mm
RMS incorrelated energy spread
0.06
%
Laser pulse duration (flat top)
10
ps
RMS bunch spot size @linac exit
0.4
mm
Laser pulse rise time (10÷90%)
1
ps
RMS bunch length @linac exit
1
mm
Bunch peak current (50% beam)
100
A
PLASMON-X
SEEDING
LASER
PHOTOINJECTOR
LASER
FLAME parameters
THOMSON
HHG
DGL
PHOTOINJECTOR
Wavelength
800
nm
Compressed pulse energy
5
J
Pulse duration (bandwidth)
30 (80)
fs (nm)
Repetition rate
10
Hz
Energy stability
10%
Pointing stability
<2
UNDULATOR
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
urad
SPARC future plans - LI2FE Experiments
• Thomson scattering
• Plasma acceleration
• Requires physical overlapping of
SPARC and FLAME beams within the
depth of focus of the laser focusing
optics.
• SPARC and Flame pulses injected in a
gas jet, requires synchronization at
the level of the period of the plasma
wave.
• Experiments: PLASMONX, MAMBO
• Experiments: PLASMONX
• Request: Δt<1psRMS
• Request: Δt<100fsRMS
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC future plans - LI2FE synchronization
FLAME AREA
FLAME oscillator
In 2856MHz
Out 79.33MHz
• Current layout
• PC laser oscillator is the OMO
• Electrical reference
distribution
• FLAME not yet considered
SPARC HALL
PC laser
Oscillator
79.33MHz
RF reference
2856MHz
RF reference
79.33MHz
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC future plans - LI2FE synchronization
FLAME AREA
FLAME oscillator
In 2856MHz
Out 79.33MHz
RF reference
2856MHz
• Easiest and quickest
• Low cost
• Coaxial cable distribution
SPARC HALL
PC laser
Oscillator
79.33MHz
• 1st solution – electrical
distribution
• Possible temperature stabilized
cable bundle
• Hundreds of femto-seconds
performance
RF reference
2856MHz
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
SPARC future plans - LI2FE synchronization
FLAME AREA
FLAME oscillator
79.33MHz
• 2nd solution – optical
distribution
• Fiber laser OMO (Optical Master
Oscillator)
• Major system modification
needed
• Higher cost
PC laser
Oscillator
79.33MHz
OMO
79.33MHz
RF reference
2856MHz
SPARC HALL
• Fiber links to distribute the signal
(active length stabilization)
• Optical mixing (cross correlation)
for laser clients
• Sub-100fs performance
June 24th 2010, 40th Meeting of the LNF Scientific Commitee
Conclusion
• Synchronization system performance
– System is inside project specifications
– New diagnostic methods developed (BAM, LAM)
– e-beam to RF jitter: ≈200 fsRMS
– Seeding experiment successful
• Future plans: electrons-photons interaction
– Possible solutions and relative jitter:
• Electrical reference distribution: 100÷200 fsRMS
• Optical reference distribution: <100 fsRMS
June 24th 2010, 40th Meeting of the LNF Scientific Commitee