EUROFEL DS3 -DS4 TASK meeting Proposed synchronization schemes for 4GLS: impact on EUROFEL DS3

EUROFEL DS3-DS4 TASK meeting CEA - Saclay, 16-17 MAY 2005 M. Ferianis

EUROFEL DS3-DS4 TASK meeting

Proposed synchronization schemes for 4GLS: impact on EUROFEL DS3

Mario Ferianis - Sincrotrone Trieste, Italy

CEA, Saclay 16-17 May 2005

• Synchronization issues• Proposed schemes

• Workplan for EUROFEL DS3


Synchronization issues

• In 3GLS, synchronization is used for– Injector: tens of ps s-band / 10 30ps

– Injection and extraction systems: 100s of ps

– Bunch diagnostics and feedback systems: few ps

– Time-resolved experiments (laser to h synch.): few ps

• Being the bunch length tB40psFWHM, typically the required

jitter is at the few psRMS level

• In 4GLS, synchronization is a key feature due to:– sub-ps bunch length, achievable in single pass accelerator

– extensive adoption of fs lasers, both embedded in the radiation production process and foreseen for time resolved experiments

• Minimum required synch. : <10fs level (i.e. seeding laser)


Synchronization issues:some thoughts from my “bulleted list”, 01/2004

• The synchronization (after the greek verb: =to be/to happen at the same time) of a system can be defined as its property of sharing a common time reference and distributed among its different sub-systems. Synchronous sub-systems, therefore, can operate at fixed relative time instants. The term “timing system” is frequently used as synonymous of synchronization system.

• The preservation of the synchronization over long terms is called stability: a synchronous system is also stable when its synchronization is kept over long times, order of magnitude larger than the “event”.

• Femto-second x-ray Light Source test stands have already proven to generate sub-picosecond (10-14s) x-ray pulses, at repetion rates of few Hertz (10-2s), to be delivered during User Shifts of many hours (104s). Therefore, the synchronization of a femto-second x-ray Light Source has to deal with 18 order of magnitude, which calls for state of the art solutions.


Synchronization & Stability

• Do we really need both fs synchronization & stability ?

• During the 1st FERMI User Workshop, there have been some discussions about it…turned out that:“each LINAC shot (FEL pulse) is like one experiment”

• So, what Users are really needing is fs synchronization, but relative to events within one machine cycle: tREP=20ms (50Hz) for FERMI

• What do they really care is that pump pulses are stable (<10fsRMS )

w.r.t. probe pulses, considered as associated pairs!

• Furthermore, both the physical phenomena and the related data taking techniques on fs experiments complete well before milliseconds repetition period


Synchronization issues:sequence of synchronous events

Optical Reference laser pulse

PC laser pulse on cathode

Phase of accelerating voltage

Laser Heater pulse

Seed Laser pulse

Cross-correlator Laser pulse

Probe Laser pulse

Pump Laser pulse

ABSOLUTE STABILITY

JITTER JITTER

ns range

ms range


Proposed schemes:a general layout for a 4GLS synchronization system

0.3 nm 0.1 nm

UV Hall X-ray Hall

Nanometer Hall

SC Linac4 GeV2 GeV1 GeV

1 nm

0.3 nm

100 nm

30 nm

10 nm

10 nm

3 nm

1 nm

Master oscillator

Pump laser

Pump laser

Seed laser

Seed laser

Seed laser

Pump laser

Fiber link synchronization

Injector laser

Undulators

Undulators

Undulators

Future upgrade to 0.1 nm at 8 GeV

SC Linac

MIT Optics & Quantum Electronics Group

04/21/23W.S. Graves,MIT Bates Laboratory


Synchronization issues: adopted optical techniques



Synchronization issues:phase noise representation

f0=carrier

Sidebands

Amplitude vs. carrier

Amp.[dBc/Hz]

“Phase Noise and Jitter”,N. Roberts, ZARLINK semicond.EE Design, July 14 2003


Synchronization issues:phase noise to jitter conversion

1/fm0

White frequency:uncorrelated fluctuations of the osc. period;thermal noise

Flickernoise








10dBc / 3 decades:L(f)=1/SQRT(f3)


SSB Phase noise measurement:effects due to Instrument powering cofiguration:

ac vs. battery source


Proposal for fs synchronization:RLE @ MIT + DESY


Measured fiber laser performanceF.O. Ilday et al. “Laser Technology for X-ray FELs: fs synchronization

…” presented at Erl 2005 workshop, JLAB


Synchronization issues: stabilized FO link Lasers and RF-Timing, Franz X. Kaertner, RLE MIT

CrossCorrelatorFiber

ML - Laser

PZT

Fixed Length L

Far end local end


Proposed schemes:Optical to RF conversion

RF synchronous to optical pulses


Proposed schemes: Optical to RF conversion


Locking of two fiber lasersF.O. Ilday et al. “Laser Technology for X-ray FELs: fs

synchronization …” presented at Erl 2005 workshop, JLAB

Reference laser:LOCAL

Second laser:REMOTE

Long FO link


Stabilization of reference signal distribution25GHz, 1km of fiber, 15 hour test



Stabilization of FO link: LBL proposal R. Wilcox, J. W. Staples LBL, R. Holzwarth, Menlo Systems “Optical Synchronization Systems for fs x-ray sources” PAC ‘05

far end

close end


R. Wilcox, J. W. Staples LBL, R. Holzwarth, Menlo Systems “Optical Synchronization Systems for fs x-ray sources” PAC ‘05

Proof of principle of Mach-Zender interferometer:with 40m of fiber in one arm showing 100s fs of jitter, it has been reduced to 0.85fs!


Stabilization of FO link: LBL proposal R. Wilcox, J. W. Staples LBL, R. Holzwarth, Menlo Systems “Optical Synchronization Systems for fs x-ray sources” PAC ‘05

since heterodyning preserves phases,phase shifts in the 110MHz is reflectedin identical phase shift in the optical signal: 1deg @110MHz==1 deg optical 0.014fs!!!


Multiple frequencies vs. Pulse train

• Multiple frequencies*

– Lower peak optical power, below non-linear effects threshold (Brillouin, Raman)

– With RF beat detection, optical to RF conversion efficiency is maximized, opposed to extraction of a high harmonic from a pulse train (losses in BP filter)

– Phase detection is measuring very high frequency beat between optical waves rather than at lower waves

• Pulse train

– No need to delivery at the far end the 110MHz signal

– Can distribute either 100MHz or 3GHz signal (direct O/E conversion for the RF plants)

– Can be used as seed for other lasers and as ref pulse in diagnostics * R. Wilcox, J. W. Staples LBL, R. Holzwarth, Menlo Systems “Optical Synchronization Systems for fs x-ray sources” PAC ‘05


Workplan for EUROFEL DS3:demonstration of the Optical Clock

REF.osc.

Fiberlaser

O/EConv.

Remote fs laser

Phase

detect

Stabilised Fiber links: star configurationL=300m

Phasenoise

Cross-correlat

e

Electrical signalOptical in fiberOptical in air

RF: 100MHz or-wave: 10GHz

fPULSE

EUROFEL DS3 -DS4 TASK meeting Proposed synchronization schemes for 4GLS: impact on EUROFEL DS3

Documents

stable synchronization

fs synchronization stability

synchronous system

synchronous subsystems

conversionphase noise

required jitter

subps bunch length

s xray pulses