P. Krejcik UCLA High Power Beams [email protected] November 8-10, 2004 LCLS Diagnostics and Commissioning Injector, Linac and Undulator Diagnostics

P. Krejcik

UCLA High Power Beams Workshop [email protected]

November 8-10, 2004

LCLS Diagnostics and Commissioning

Injector, Linac and Undulator Diagnostics and Beam Position Monitors

P. Krejcik

P. Krejcik


November 8-10, 2004

Context of Diagnostics in Commissioning

Review scope of proposed diagnostics

Emphasize that diagnostics themselves need commissioning

Consider if full features (resolution, automation) are needed at beginning of commissioning

Implicit sequence of commissioning: e.g. feedbacks after BPMs commissioned; slice parameters need prof. monitors and TCAVs

P. Krejcik


November 8-10, 2004

Readiness of diagnostic systems

Which SLC diagnostics should be preserved?Technology choices still being made on some new systems

BPM modules – trying to attain desired resolutionProf monitor cameras – resolution, controls integration, data rate

Some diagnostics are turnkey systems, others are R&D projectsR&D still required for ultrafast diagnostics

CSR THz power bunch length monitorsEO bunch profiling

P. Krejcik


November 8-10, 2004

Dynamic Aspects of Commissioning

Initially diagnose a wildly mis-tuned and unstable machine

Yet the same diagnostics should ultimately have finesse to optimize SASE operation

Deal with imperfect and uncalibrated settings

Detective work for finding hardware faults

Quantify magnitude and sources of jitter

P. Krejcik


November 8-10, 2004

Diagnostics Roadmap for electrons

Trajectoryresolution

•Position•Angle•Energy

Beam sizeresolution

•Emittance•Energy spread

Bunch length & Tdevelopment

•Longit. profile•Single shot rms

Slice parametersresolution

• Emittance• Energy spread

Bunch charge FEEDBACK

Stabilizationresponse•Jitter characterization

Noninvasive Invasive

• Setup•Tuning

120 Hz

P. Krejcik


November 8-10, 2004

Accelerator System Diagnostics*• 180 BPMs at quadrupoles and in each bend system180 BPMs at quadrupoles and in each bend system

upstream linac

L0

RFgun

L3L1 X L2

• 8 Energy (BPM) 8 Energy (BPM) E, energy spread (Prof) , energy spread (Prof) E measurements : measurements :

• 2 Transverse RF deflecting Cavities for slice measurements2 Transverse RF deflecting Cavities for slice measurements

BC1 BC2 DL2DL1undulator

LTU

• 5 Bunch length monitors5 Bunch length monitors

DumpE, E

• 5 Emittance 5 Emittance x,y measurements (Profs, Wire Scanners) : measurements (Profs, Wire Scanners) :

3 prof. mon.’s3 prof. mon.’s((xx,,yy = 60 = 60°°))

* See also P. Emma talk how optics is optimized for diagnostics

P. Krejcik


November 8-10, 2004

Beam Position Monitoring requirements

P. Krejcik


November 8-10, 2004

Beam Position MonitorsStripline BPMs in the injector and linac (existing) and in the LTU

Differencing large numbersMechanical precision

Fabrication by printing electrodes on ceramic tubes

Drift in electronicsDigital signal processing

Cavity BPMs in the undulator, LTU launchSignal inherently zero at geometric centerC-band (inexpensive) signal needs to be mixed down in the tunnel

P. Krejcik


November 8-10, 2004

Stripline versus Cavity BPM SignalsP

f700 MHz

500 MHzBP filter

ADCx4

119 MHzClock

24th harmonic

DigitalprocessingRF in

Controlsystem

/4

Stripline

Mixer

LO sync’ed to RF

IF

• noise (resolution) minimized by removing analog devices in front of ADC that cause attenuation• drift minimized by removing active devices in front of ADC

• noise (resolution) minimized by removing analog devices in front of ADC that cause attenuation• drift minimized by removing active devices in front of ADC

C-bandcavity

Dipole mode

coupler

~5 GHz

P. Krejcik


November 8-10, 2004

Simplistic View of Digital BPMs

Is the purely digital approach the best way to go?Must always maximize signal to noise for best resolution

So eliminate any cause of attenuation: couplers, hybrids, active devices etc.

This also eliminates drift which causes offsetsOther approaches also try to do this: e.g. AM to PM conversion with a hybrid and then digitizeMight as well digitize first, eliminate the middle men, and do the conversions digitallyUltimately left with calibrating the drift in the BPM cables, because ADCs are now very stable.

P. Krejcik


November 8-10, 2004

Linac stripline BPMsNeed to replace old BPM electronicsNeed to replace old BPM electronicsCommercially available processing units look promisingCommercially available processing units look promisingBeam testing of module as soon as funding availableBeam testing of module as soon as funding availableTest new BPM fabrication techniquesTest new BPM fabrication techniques

http://www.i-tech.si

P. Krejcik


November 8-10, 2004

Analysis of Test Signals in the “Libera” module – S. Smith

Measured signal to noise ratio implies resolution of 7 m in a 10 mm radius BPM

Identified fixable artifacts in data processing

P. Krejcik


November 8-10, 2004

Cavity beam position monitors for the undulator and LTU

Coordinate measuring machine verification of cavity interior

• X-band cavity shown

• Dipole-mode couplers



R&D at SLAC – S. Smith





NLC studies of cavity BPMs, S. Smith et al

NLC studies of cavity BPMs, S. Smith et al

P. Krejcik


November 8-10, 2004

C-band beam tests of the cavity BPM – S. Smith

25 m

200 nm

• Raw digitizer records from beam measurements at ATF

• Raw digitizer records from beam measurements at ATF

cavity BPM signal versus predicted position at bunch charge 1.6 nC

cavity BPM signal versus predicted position at bunch charge 1.6 nC

• plot of residual deviation from linear response• << 1 m LCLS resolution requirement

• plot of residual deviation from linear response• << 1 m LCLS resolution requirement

• C-band chosen for compatibility with wireless communications technology

• C-band chosen for compatibility with wireless communications technology

P. Krejcik


November 8-10, 2004

LCLS BPM Testing

Testing is planned at the “Controls Test Stand” to be located at the FFTB, 2005.

Evaluation of processor electronicsResolution determined by comparing several adjacet BPMs

Possibility to test new striplines

Copy the design of NLC C-band cavity BPMS

P. Krejcik


November 8-10, 2004

Beam Size MeasurementWire scanners, based on existing SLAC systemsWire scanners, based on existing SLAC systems

Measures average projected emittanceMeasures average projected emittanceBut is minimally invasive and can be automated for regular monitoringBut is minimally invasive and can be automated for regular monitoring

Profile monitorsProfile monitorsSingle shot, full transverse profileSingle shot, full transverse profileYAG screen in the injector for greater intensityYAG screen in the injector for greater intensityOTR screensOTR screens in the linac and LTU for high resolution in the linac and LTU for high resolution1 1 m foilsm foils successfully tested in the SPPS: successfully tested in the SPPS:

Small emittance increase disrupts FEL, Small emittance increase disrupts FEL, but no beam lossbut no beam loss

-1:1 imaging optics => ~ 9 -1:1 imaging optics => ~ 9 m resolutionm resolutionUsed in combination with TCAV Used in combination with TCAV

for slice energy spread and emittancefor slice energy spread and emittanceCTR for bunch length measurementCTR for bunch length measurement

OTR image taken in the SPPSCourtesy M. Hogan, P. Muggli et al

OTR image taken in the SPPSCourtesy M. Hogan, P. Muggli et al

P. Krejcik


November 8-10, 2004

Profile Monitor Camera Specification• Digital camera technology

• Not TV camera that subsequently needs a frame grabber• External trigger supplied to the camera by control system

• 30 fps at 1280x960 pixels, 10 bit resolution• Digital image read out over ethernet or firewire• Inexpensive, commercially available ~$1k – Z. Salata.

P. Krejcik


November 8-10, 2004

Profile Monitor Camera Dynamic Range• How many bits are necessary to see the tails?

saturation

3 needs 10 bits

4 needs 12 bits

P. Krejcik


November 8-10, 2004

Profile monitor commissioning

Can be tested off the beamline at the Controls Test Stand

Evaluate data acquisition and integration into the control system

test a complete optical setup and measure optical resolution and wavelength response

P. Krejcik


November 8-10, 2004

Bunch length diagnostic comparisonDevice Type Invasive

measurementSingle shot measurement

Abs. or rel. measurement

Timing measurement

Detect bunchin

g

RF Transverse Deflecting Cavity

Yes: Steal 3 pulses

No: 3 pulses Absolute No No

Coherent Radiation Spectral power

No for CSR Yes for CTR

Yes Relative No Yes

Coherent RadiationAutocorrelation

No for CSR Yes for CTR

No Absolute(2nd moment

only)

No No

Electro Optic Sampling

No Yes Absolute Yes No

Energy Wake-loss

Yes No Relative No No

P. Krejcik


November 8-10, 2004

Bunch Length Measurements with the RF Transverse Deflecting CavityBunch Length Measurements with the RF Transverse Deflecting Cavity

yy

Asymmetric parabola indicates incoming tilt to beam

A = 1.6696E-02 STD DEV = 1.3536E-03B = 28.23 STD DEV = 3.084C = 1328. STD DEV = 8.235RMS FIT ERROR = 23.63

-80 -40 0 40 80

SBST LI29 1 PDES (S-29-1)

1.7

1.6

1.5

1.4

1.3

X103

****

***

**

**

*

****

********

0 40 80

SBST LI29 1 PDES (S-29-1)

1.7

1.6

1.5

1.4

1.3

X103

E

0 40 80

1.7

1.6

1.5

1.4

1.3

X103

0 40 80

1.7

1.6

1.5

1.4

1.3

X103

0 40 80

1.7

1.6

1.5

1.4

1.3

X103

0 40 80

1.7

1.6

1.5

1.4

1.3

X103

E

1-APR-03 20:21:16

Cavity on

Cavity off

Cavity on- 180°

Bunch length reconstructionMeasure streak at 3 different phases

z = 90 m

(Str

eak

size

)2

0° 180°

2.4 m 30 MW

P. Krejcik


November 8-10, 2004

Commissioning of the Transverse Cavities

Calibration of the deflection strength in units of pixels on the profile monitor

Also requires beam trajectory feedback to stabilize the RF phase of the deflecting cavity

Prof monitor image acquisition fully integrated into the control system

P. Krejcik


November 8-10, 2004

Calibration scan for RF transverse deflecting cavity

Beam centroid[pixels]

Cavity phase [deg. S-Band]

• Bunch length calibrated in units of the wavelength of the S-band RF

Further requirements for LCLS:

•High resolution OTR screen•Wide angle, linear view optics

P. Krejcik


November 8-10, 2004

OTR Profile Monitor in combination withRF Transverse Deflecting Cavity- detailed applications in P. Emma talk

Simulated digitized video image

Injector DL1 beam line is shown

Best resolution for slice energy spread measurement would be in adjacent spectrometer beam line.

P. Krejcik


November 8-10, 2004

Coherent radiation from the electron bunchCoherent radiation from the electron bunch

Frequency domainFrequency domainSpectral power in a fixed bandwidthSpectral power in a fixed bandwidth

SpectrometrySpectrometry

Autocorrelation Autocorrelation

Time domainTime domainElectro optic samplingElectro optic sampling

Measured directly near the bunchMeasured directly near the bunch

Or transported out of the beam lineOr transported out of the beam line

P. Krejcik


November 8-10, 2004

m 190≈z

Diagnosing Coherent Radiation1. spectral power

Smooth Gaussian bunch spectrum from BC1

With 5% microbunching

• Measure bunch length

• Detect microbunching

• Measure bunch length

• Detect microbunching

Fixed BW detectorSignal prop. 1/z

Bunch length signal for RF feedback

- J. Wu

P. Krejcik


November 8-10, 2004

4 THz main peak

BC2 Bunch length monitor spectrum - based on coherent spectral power detection

BC2 bunch length feedback requires THz CSR detector

Demonstrated with CTR at SPPS

Spikes in the distribution now have same spectral signature as microbunching

P. Krejcik


November 8-10, 2004

Transition radiation is coherent at wavelengths longer than the bunch length,>(2)1/2 z

SLAC SPPS measurement: P. Muggli, M. Hogan

Limited by long wavelength cutoff and absorption resonances

0

0.4

0.8

1.2

1.6

-100 -50 0 50 100CombinedCTRInterferogramsSm0

0.4

0.8

1.2

1.6

-100 -50 0 50 100CombinedCTRInterferogramsSm

zz 9 9 mm

Diagnosing Coherent Radiation2. autocorrelation

P. Krejcik


November 8-10, 2004

10-17

10-15

10-13

10-11

10-9

10-7

10-5

0.001

0.1

10

10-4

10-3

10-2

10-1

100

10 100 1000CTRFSpecSigmaz20Mylar12.5_3

Wavelength (µm)

Mylarresonances

Simple model: Gaussian, z=20 µm, d=12.7 µm, n=3 Mylar window+splitter

Transport issues for THz radiationTransport issues for THz radiation

• Smaller measured width:

Autocorrelation < bunch !

• Modulation/dips in the interferogram• Fabry-Perot resonance: =2d/m, m=1,2,…

• Signal attenuated by Mylar: (RT)2 per sheet

• Fabry-Perot resonance: =2d/m, m=1,2,…

• Signal attenuated by Mylar: (RT)2 per sheet

P. Muggli, M. Hogan

P. Krejcik


November 8-10, 2004

Developments in autocorrelation techniques

Investigate other detector types for wavelength dependance

Golay cell

Beam splitters without wavelength dependance

Single shot autocorrelatorCamera records fringes on single shot

Use CSR from chicane bed

P. Krejcik


November 8-10, 2004

Bunch length scan performed while observing spectral Bunch length scan performed while observing spectral power with THz detectorpower with THz detector

Coherent transition radiation wavelength comparable

to bunch length

LINAC

FFTB

Comparison of bunch length minimized according to

wakefield loss and THz power

Pyroelectric detector

foil

GADC-26 -24 -22 -20 -18 -16 -14 -120

100

200

300

400

500

600

Pyrometer signal [arb. units]

linac phase offset from crest [deg. S-Band]

FFTB Pyrometer Signal

-26 -24 -22 -20 -18 -16 -14 -12200

250

300

350

400

450

500

energy loss [MeV]

Linac Wake Loss

Linac phase

Wake energy loss

THz power

P. Krejcik


November 8-10, 2004

Dither feedback control of bunch length minimization at SPPS - L. Hendrickson

Dither time steps of 10 seconds

Bunch length monitor response Feedback correction

signal

Linac phase

“ping”

optimum

Jitter in bunch length signal over 10 seconds ~10% rms

Jitter in bunch length signal over 10 seconds ~10% rms

P. Krejcik


November 8-10, 2004

Diagnosing Longitudinal phase space:Energy spectrum versus Bunch length signal

- Muggli, Hogan et al

Jitter in the compressor phase:

Resuting energy profile Corresponding bunch length signal

Single shot measurements

jitter

signal

P. Krejcik


November 8-10, 2004

LB=1.80 mB=1.60 T

LB=1.80 mB=1.60 T

SPPS Four Dipole Chicane

s

LT=14.3 m

9 GeV9 GeV

BPM - energyProf. Monitor - E

Momentum compactionR56= –75 mm

Momentum compactionR56= –75 mm

z

50 m

1.6%

z0

1.2 mm

Correlated energy spread

Linac chirp

Measured energy spread

SR background

P. Krejcik


November 8-10, 2004

Measured and predicted energy spread from wakefield chirp in SPPS

Measured and predicted energy spread from wakefield chirp in SPPS

Special setup to give 100 m bunch length with more charge at the head of the bunch

Measured at end of linacMeasured at end of linac

P. Krejcik


November 8-10, 2004

Wakefields change not only the energy spread in the bunch

But also the centroid energy of the bunch

Fast means of determining relative bunch length

Wakefields change not only the energy spread in the bunch

But also the centroid energy of the bunch

Fast means of determining relative bunch length

P. Krejcik


November 8-10, 2004

Relative bunch length measurementbased on wakefield energy loss scan

Relative bunch length measurementbased on wakefield energy loss scan

Energy change measured at the end of the linac

as a function of the linac phase (chirp) upstream of the compressor chicane

Shortest bunch has greatest energy loss

Predicted wakeloss___

For bunch length z

__

Predicted shape due to wakeloss plus RF curvatureP. Emma, K. Bane

P. Krejcik


November 8-10, 2004

Coherent radiation from the electron bunchCoherent radiation from the electron bunch

Frequency domainSpectral power

Spectrometry

Autocorrelation

Time domainTime domainElectro optic samplingElectro optic sampling

Measured directly near the bunchMeasured directly near the bunch

Or transported out of the beam lineOr transported out of the beam line

P. Krejcik


November 8-10, 2004

SPPS Electro Optic Bunch Length Measurement with

in-vacuum crystal

Probe laser

Defining aperture

Beam axis

M1 M2EO xtal

Geometry chosen to measure direct

electric field from bunch, not wakefieldModelled by H. Schlarb

electrons

P. Krejcik


November 8-10, 2004

Features of the SPPS Electro Optic SetupCompressed pulse from the users pump-probe Ti:Sa laser Compressed pulse from the users pump-probe Ti:Sa laser oscillatoroscillator

Transported low power pulse over ~150 m fiber to the Transported low power pulse over ~150 m fiber to the electron beam lineelectron beam line

OTR provides coarse timingOTR provides coarse timing

Ti:Saoscillator

Stretcher Shaper Fiber

launch

e-

EO xtl

polarizing beamsplitter

s

pimaging optics

~150 m fiber

OTR

P. Krejcik


November 8-10, 2004

Features of the SPPS Electro Optic Setup

Fiber incorporated in pulse compression setup Fiber incorporated in pulse compression setup including compensating fiber dispersion with a including compensating fiber dispersion with a spatial light modulator spatial light modulator

Cavalieri et al, FOCUS Group U. MichiganCavalieri et al, FOCUS Group U. Michigan

Fiber incorporated in pulse compression setup Fiber incorporated in pulse compression setup including compensating fiber dispersion with a including compensating fiber dispersion with a spatial light modulator spatial light modulator

Cavalieri et al, FOCUS Group U. MichiganCavalieri et al, FOCUS Group U. Michigan

SLM 640-pixel

f f f f

To fiber

Grating pair

From stretcher

P. Krejcik


November 8-10, 2004

Features of the SPPS Electro Optic Setup

Crystal mounted close to electron beamAvoid wakefields from smaller apertures

ZnTe crystal: 200 um thick

EO coefficient,

phase match,

phonon resonances

Crystal mounted close to electron beamAvoid wakefields from smaller apertures

ZnTe crystal: 200 um thick

EO coefficient,

phase match,

phonon resonances

P. Krejcik


November 8-10, 2004

Electro-Optical Sampling at SPPS Electro-Optical Sampling at SPPS – A. Cavalieri et al.– A. Cavalieri et al.Single-ShotSingle-Shot

<300 fs<300 fs

170 fs rms170 fs rmsTiming JitterTiming Jitter

Er

Line image camera

polarizer

analyzer

Pol. Laser pulse

Electron bunch

EO crystal

Bunch length scan

P. Krejcik


November 8-10, 2004

Electro optic resolution limitsElectro optic resolution limits

Spatial imaging resolution limits time resolution

Crossing angle determines width of time window and temporal resolution

Resolution limit then set by crystal thickness and the phase velocity mismatch

Crystal material chosen to minimize phase mismatch

Spatial imaging resolution limits time resolution

Crossing angle determines width of time window and temporal resolution

Resolution limit then set by crystal thickness and the phase velocity mismatch

Crystal material chosen to minimize phase mismatch

P. Krejcik


November 8-10, 2004

Electro optic resolution limits

Future experimentsSmaller crossing angle

Smaller angle magnifies time coordinate on spatial axis

But reduces the time window to accommodate beam jitter

EO polymer filmsStrong EO coefficient

May not last long

Higher laser power cross correlation techniques (Jamison et al)

Laser amplifier located near beamline

Future experimentsSmaller crossing angle

Smaller angle magnifies time coordinate on spatial axis

But reduces the time window to accommodate beam jitter

EO polymer filmsStrong EO coefficient

May not last long

Higher laser power cross correlation techniques (Jamison et al)

Laser amplifier located near beamline

Ti:Sapphire Ti:Sapphire laserlaserTi:Sapphire Ti:Sapphire laserlaser

200 200 m thick ZnTe crystalm thick ZnTe crystal

ee

New chamber T. Montagne

P. Krejcik


November 8-10, 2004

Synchronization of the Laser timing

Jitter in the laser timing effects

Electro optic bunch timing measurement

Pump-probe timing for the users

Enhancement schemes using short pulse lasers

Jitter in the laser timing effects

Electro optic bunch timing measurement

Pump-probe timing for the users

Enhancement schemes using short pulse lasers

P. Krejcik


November 8-10, 2004

SPPS Laser Phase Noise Measurements – R. Akre 476 MHz

M.O.

x62856 MHz

to linac

MDL3 km

fiber~1 km

VCO

Ti:Salaser osc

diode

EO

scope

Phase detector

2856 MHz

P. Krejcik


November 8-10, 2004

Energy and Bunch Length Feedback Loops

L0 L1

DL1

DL1Spectr. BC1 BC2

L2 L3

BSY 50B1

DL2

Vrf(L0)

Φrf(L2)Vrf(L1) Φrf(L3)E E E

Φrf(L2) zΦrf(L1) zE

4 energy feedback loops2 bunch length feedback loops120 Hz nominal operation, <1 pulse delay

Progressive commissioning schedule

P. Krejcik


November 8-10, 2004

Undulator trajectory launch loop to operate at 120 Hz, <1 pulse delay

Damps jitter below 10 Hz

i.e. need stability above 10 Hz!

At lower rep. rates, less damping

Linac orbit loops to operate at 10 Hz because of corrector response time

Antidamp

Damp

Gain bandwidth shown for different loop delays- L. Hendrickson

Closed Loop Response of Orbit Feedback

P. Krejcik


November 8-10, 2004

Remaining intra-undulator diagnostics – from Bingxin Yang, Lehman Review August ‘04

Location: every long break (905 mm)Diagnostics chamber length: 425 mm

Functional components

RF BPM, Cherenkov detector, OTR profiler, wire scanner, x-ray (intensity) diagnostics

P. Krejcik


November 8-10, 2004

FY04 accomplishments– from Bingxin Yang, Lehman Review August ‘04

Layout of diagnostics chamber

OTR profilerCamera module designed

Wire scannerScanner design in progress

Wire card adapt SLAC design

X-ray diagnostics designBeam intensity: double crystal

Beam profile: imaging detector

P. Krejcik


November 8-10, 2004

Major issues at UCLA workshop– from Bingxin Yang, Lehman Review August ‘04

Beam damage of optical componentsExample from Marc Ross’ coupon test, LINAC 2000

Saturated FEL beam deposit higher energy density

Desirable informationTrajectory accuracy (x~1m)

Effective K (K/K ~ 1.5×10-4)

Relative phase (~10º)

Intensity gain (E/E~0.1%, z-)

Undulator field quality

P. Krejcik


November 8-10, 2004

Rethink x-ray diagnostics (Galayda) – from Bingxin Yang, Lehman Review August ‘04

Intra-undulator diagnosticsElectron beam position monitor (BPM)

Electron beam profiler (OTR & wire scanner)

Low power x-ray Intensity measurements (R&D)

Beam loss Monitor

Far-field low-power x-ray diagnostics (R&D)Clean signature from spontaneous radiation

Space for larger optics / detectors

Single set advantage (consistency, lower cost)

Goal = obtain “desirable information”

P. Krejcik


November 8-10, 2004

Final Beam Dump

Sensitive measurement of beam energy

Optimized for energy spread resolution of 4*10-5 (P.Emma)

Bends smear out microbunching

Dispersion hides emittance measurementMight be possible in the vertical plane

P. Krejcik


November 8-10, 2004

Summary

Diagnostics integrated into the LCLS design

All systems require commissioning time to achieve LCLS resolution requirements

New diagnostics still require R&D for bunch length and timing

Developmental work at SPPS is critical

Diagnostics being developed hand-in-hand with controls and feedbacks

P. Krejcik


November 8-10, 2004

Appendix

Documents

P. Krejcik UCLA High Power Beams [email protected] November 8-10, 2004 LCLS Diagnostics and Commissioning Injector, Linac and Undulator Diagnostics