Advanced Accelerator R

1
Advanced Accelerator RD at the A0 Photoinjector

• Ray Fliller III
• FNAL
• February 5, 2008

2
Outline

• Introduction to the A0 Photoinjector
• Transverse to Longitudinal Emittance Exchange
• Coherent Synchrotron Radiation
• Conclusion

3
A0 Photoinjector

• L band 1.5 cell NC RF gun with Cs2Te photocathode
• 35 MV/m maximum cathode gradient
• TESLA technology accelerating cavity
• 12 MV/m accelerating gradient
• Round to Flat beam transformer
• Transverse to Longitudinal Emittance Exchange
  Beamline
• Dipoles recycled from Bunch Compressor
• Quadrupole transport channel
• User experimental area
• Prototype kicker for ILC DR (UIUC, Cornell)

4
Beam Parameters

• Laser energy 16 mJ/pulse _at_ 263nm
• lt5nC/bunch (have had gt12 nC in the past)
• Typically 10 bunches/RF pulse. 1 Hz rep rate
• 4 MeV gun output energy
• 16 MeV total energy
• Dp/p 0.3_at_ 16MeV (1nC)
• Bunch length 2 mm (1nC)
• gez 120 mm-mrad (RMS _at_ 1nC)
• gex,gey4 mm-mrad (RMS _at_ 1nC)

Let those numbers tumble around in the back of
your mind during this talk
5
Experiments

• A0 Photoinjector started about a decade ago as
  part of the TESLA collaboration.
• A0PI has been used for various experiments since
• Laser development (Ph.D. thesis Alan Fry)
• Plasma acceleration
• Channeling radiation
• Electro-Optical Sampling (Ph.D. thesis Mike
  Fitch)
• Flat beam transformation (Ph.D. thesis Yin-E.
  Sun)
• Laser acceleration and bunch Compression (Ph.D.
  thesis Rodion Tikhoplav)
• Polarized RF electron gun development
• Development of NEA GaAs cathode preparation
  chamber for FNAL
• Low Level RF system development
• Beam Diagnostics Development
• Currently Electro-Optical Sampling for bunch
  length measurement
• Martin-Puplett Interferometer for bunch length
  measurement
• Optical Transition Radiation Interferometry for
  energy and transverse emittance mesaurement
• Transverse to Longitudinal Emittance Exchange
• Investigations of Coherent Synchrotron Radiation
• This is only a partial list
• Sorry if I forgot your favorite!

6
Current A0 People

• Helen Edwards The Boss
• Don Edwards
• Me
• Jinhao Ruan Laser, All things optical
• Jamie Santucci Operations
• Tim Koeth Rutgers Ph.D. Student
• Artur Paytan Yerevan U. Ph.D. Student
• Mike Davidsaver UIUC staff, controls
• Grigory Kazakevich Guest Scientist, OTRI
• Manfred Wendt Instrumentation, BPMs
• Randy Thurman-Keup Instrumentation,
  Interferometer
• Vic Scarpine Instrumentation, OTR and cameras
• Alex Lumpkin Instrumentation, Radiation
  Diagnostics
• Ron Rechenmacher CD, controls
• Lucciano Piccoli CD, controls
• Gustavo Cancelo CD, Low Level RF
• Wade Muranyi Mechanical Support, Lead Tech.
• Many Others from AD/RF, AD/MS
• Sorry if I forgot your name!

7
Transverse to Longitudinal Emittance Exchange
8
Longitudinal to Transverse Emittance Exchange

• Cornacchia and Emma (CE) in 2002 proposed using
  a deflecting mode cavity in the center of a
  chicane to exchange a smaller longitudinal
  emittance with a larger transverse emittance for
  an FEL.
• Kim and Sessler in 2005 proposed a deflecting
  mode cavity between 2 doglegs to exchange
  transverse and longitudinal emittances.
• Emma, Huang, Kim, Piot in 2006 proposed using a
  flat beam (exltltey) combined with a deflecting
  mode cavity between 2 doglegs to produce a beam
  with very small transverse emittances and large
  longitudinal emittance to drive an FEL.
• We are going to do a proof of principle emittance
  exchange at A0 using the double dogleg approach
  with a round beam.
• Well be exchanging a larger longitudinal
  emittance with a smaller transverse emittance.
• This will be the Ph.D. thesis of Tim Koeth, a
  Rutgers student.

9
Deflecting Mode Cavity (DMC)
Derived from Figure 1 of CE. Electric field at
synchronous phase. Magnetic field a quarter
period later.

• No longitudinal electric field on axis.
• Electric field imparts an energy kick
  proportional to distance off axis.
• Plan to use this to change the momentum deviation
  in presence of dispersion!
• Electro-magnetic field provides deflection as a
  function of arrival time.
• This type of cavity can be used as a crab cavity
  or for bunch length measurement.

k is the integrated longitudinal energy gain at a
reference radius a normalized to the beam energy
E.
10
Beam Matrix Transport

• The 4x4 emittance matrix at two points in an
  accelerator are related by
• R is the 4x4 transport matrix between these
  points
• B and C typically have zero determinant and
  couple transverse and longitudinal emittances
  through dispersion.
• The emittances after the transport line are given
  by

Derivation follows CE
11
Requirements for an Exchange

• These equations show that for perfect exchange we
  need
• How to get l20? Its not obvious, but only
  happens if
• We have no coupling at all
• Or
• If l2?0 there is still exchange, but the
  emittances are coupled.

Follows from the symplectic condition
Derivation follows CE
12
How to Make an EEX Beamline

• Assume that the beamline consists of a before
  cavity section, a DMC, and an after cavity
  section.
• Assume that the before cavity section produces
  some dispersion, h, with a slope h.
• Assume that the cavity is a zero length element
• What does the cavity strength need to be?
• What are the needed properties for the after
  cavity section?
• These equations come out of nothing more than the
  symplectic condition and the condition that the A
  and D blocks of the R matrix are all zeros.

FNAL Beams Doc 2553
13
Types of Solutions

• Some solutions are
• A dogleg after the cavity (K.J. Kims idea)
• A beamline segment that gives additional
  dispersion with no slope in the absence of the
  DMC
• A quad followed by a dipole (Helen Edwards idea)
• A beamline segment that gives no dispersion (at
  the end of the beamline) with a slope in the
  absence of the DMC
• A chicane is unable to exchange the emittances
  without residual coupling. In fact, any solution
  that would be dispersionless without the cavity
  cannot produce a complete uncoupled emittance
  exchange (l2?0) independent of the incoming beam
  parameters.
• CE stated the coupling for a chicane in their
  paper.
• Although the input beam parameters can be
  adjusted to minimize it

14
Thick Cavities

• The previous analysis assumes a zero length
  cavity
• Don Edwards has derived the matrix for a single
  cell, pillbox cavity
• An n cell pillbox cavity of total length L, and
  strength k, can be constructed

15
Thick Cavities and Emittance Coupling

• What does this do to the exchange?
• The finite length cavity, particularly the 4,3
  term leads to a coupling of the exchange that
  cannot be eliminated by beamline design alone.
• The A and D blocks cannot be made all zeros.
• How can you minimize the coupling??
• The optimal cavity strength remains unchanged.
• The after cavity section obeys the same
  equations.
• The dispersion going into the cavity should have
  no slope.
• This will make the A and D determinants zero
• The residual coupling for our beamline in terms
  of the beam parameters at the cavity
• So at the cavity, you want to
  eliminate the coupling.

16
Thick Cavity at A0PI

• For the A0 experiment, this ratio is limited to
  2h
• The bunch length is limited by the input beam and
  the R56 of the beamline
• Quads upstream of the beamline already focus the
  b function at the cavity
• However, it is a small effect anyway
• The transverse emittance after EEX is unaffected
• This is the larger of the emittances.
• For the finicky it is increased 0.04....
• The longitudinal emittance after EEX increases 6
• There are bigger fish to fry when it comes to
  diluting the emittances
• More on this in the next section of the talk

17
Emittance Exchange - CDR

• Incoming beam has energy chirp that partially
  compresses the bunch at the DMC.
• Quads prior to the first dipole focus the beam at
  the cavity.
• First dogleg provides dispersion at DMC.
• The cavity reduces the momentum spread of the
  beam and gives a shear deflection to the beam.
• The second dogleg finishes the exchange.

Diagram by Tim Koeth.
18
Watching the Exchange
Input to the EEX line
Before Dipole 2
Before DMC
After DMC
Before Dipole 4
Exchange Complete
19
Simulation of the Exchange
Dogleg Start
DMC
Dogleg Finish
20
Beamline Layout with the Cavity Off
Deflecting Mode Cavity
21
Beamline Installed
22
Deflecting Mode Cavity Design

• Our dogleg is designed to produce h0.33m.
• The beam energy is 14.3 MeV.
• This means the cavity needs to produce a kick of
  45keV/mm of transverse offset.
• We will use a 3.9GHz, 5 cell copper cavity.
• Based on the CKM SRF deflecting cavity
• Cooled with liquid N2
• This choice is driven by
• The need to produce a cavity quickly
• An available 80 kW klystron
• The mode spacing of the cavity vs. the number of
  cells

23
Cavity Construction Timeline

• Half cells arrived Aug. 9, 06
• Dumbbells brazed Aug. 29
• Dumbbell trim spec Sept. 20
• Dumbbell leak check Sept. 25
• OOOPS - Deformed the dumbbells
• Expanded the dumbbells - Oct. 13
• Did not restore the profile
• Replacement dumbbells arrive Jan. 2 07
• Cavity brazed Feb. 6
• Bead pulls finished Mar. 1
• First low power cold test Apr. 5
• Conditioning started May 25
• High Power cold test May 29
• Accidentally overpressurized vessel June 7
• Retuning cavity completed July 23
• High Power test Aug. 8
• New exhaust installed on vessel!!!
• Cavity installed in beamline Aug. 25
• Cryo clearance Sept. 6

Pictures by Reidar Hahn and Tim Koeth
24
Cavity Polarization and Field Flatness
Vertical

• Longitudinal electric field vs angle in cells 2-4
  determined by bead pull.
• Cavity polarization is set by input coupler.

• Bead pull results of cavity field flatness tuning.

25
First Deflected Beam by a CKM type Cavity
Operating phase for exchange

• The induced kick is about ½ of what was expected
  for the input power
• Tim and Leo were still quite excited!
• There were a few reasons for that, which were
  rectified
• But even now, the maximum cavity strength is only
  70 of what we need
• We think this is due to a change in field
  flatness

26
Diagnostics for EEX

• Prior to the exchange we have the following beam
  diagnostics
• OTR screens for beam spot size measurement
• Slits for uncorrelated beam divergence
  measurement
• Combined with above gives a transverse emittance
  measurement
• Horizontally bending spectrometer for energy and
  energy spread measurement
• Streak Camera for bunch length
• After the exchange
• Same diagnostics for measuring transverse
  emittance
• Vertically bending spectrometer for momentum
  spread
• Allows us to decouple the momentum spread
  measurement from any residual horizontal
  dispersion from the doglegs
• Martin-Puplett interferometer for bunch length
  measurement
• The bunch length is too short for the streak
  camera

27
WARNING Some of the data that you are about
to see are VERY preliminary results. Some of
these data are hot off the press and the
analysis needs some refinement....
28
Emittance Data

• We measure emittance in a 2 step process
• Measure the spot size at a location
• Insert slits at same location
• Measure the size of the slits at a downstream
  location
• This gives us
• Spot size
• Uncorrelated beam divergance
• The product is emittance
• Spacing between the slits
• sxx /sx2 -a/b

29
Input Emittance Data and Courant-Snyder Parameters

• The measured emittance is larger than socially
  acceptable
• It does not agree with previous measurements
• Tim is investigating the problem
• The disagreement between the a and b is only
  because of the emittance
• Otherwise we have good agreement

30
Incoming Momentum Spread

• The incoming beam momentum spread agrees well
  with simulations.
• Below is a beam image after our spectrometer in
  the straight ahead line.

31
Incoming Momentum Spread Mystery

• For some unknown reason, currently, we are seeing
  multiple peaks in the energy spread of the beam.
• Even in our straight ahead line
• It is even intermittant here one day, gone the
  next
• Possible reasons are
• A double guassian laser pulse
• Initial diagnostics show that is not the case
• We have been having RF issues with our 9 cell
  lately
• Is this related to the emittance issue??
• Below is an image of the double peaks beam is
  26 off crest

32
Double Peak Investigation

• We did a simple experiment
• Measure the energy difference between the peaks
  at different RF phases.
• Ask Is the energy difference consistent with
  two pulses separated by some time interval?
• We measured 26 and 36 degrees off crest.
• In both cases double peak spacing corresponds to
  2.00.1 deg, otherwise known as
• 4.3ps, or 1.3 mm, or 232 GHz.
• Conventional wisdom is that the laser is the
  cause.
• Jinhao pleads not guilty!
• We have checked laser with
• Streak camera - UV only
• CW Autocorrelator -IR only
• Single Shot Autocorrelator -IR only
• All evidence supports Jinhaos claim!
• Streak Camera measurements on the electrons show
  a single gaussian pulse.
• Are there distant fellow travellers?
• Looked at the laser nothing.
• Looked at BPMs nothing.
• Need to redo streak camera on electron and laser.

33
Incoming Bunch Length Measurements

• Bunch length measurements for the incoming beam
• Reasonable agreement with simulations
• Chirping the beam does not shrink the bunch
  length
• After the exchange line bunch length is near the
  resolution limit of the streak camera
• The expected bunch length is 1.6 ps after the EEX
• The streak camera resolution 1 ps
• The bunch length without the cavity is 8.1 ps
• But it will have been over-compressed in the
  process

34
Horizontal Dispersion through EEX Line

• Cavity is off during these measurements

35
Beam Spot Image after EEX line

• This is a typical cavity off beam image after the
  EEX line
• The beam is too big to fit on the screen
• Horizontal beam size here is due to 56 cm of
  dispersion

36
Horizontal Spot Size Preliminary!!!!
Screen edge

• Above is a typical cavity off horizontal profile
  after the second dogleg
• The screen edges are clearly visible
• As the cavity strength is increased, the beam is
  clearly manipulated
• Further analysis needs to be done
• The vertical spot sizes dont look this bad

k44kideal
k59kideal
k66kideal
k79kideal
37
Vertical Spot Size

• Above is a typical cavity off vertical profile
  after the second dogleg .
• Same screen
• Same conditions as before

38
Horizontal Spot Size Along Beamline vs. Cavity
Strength
Dogleg Ends
Cavity
Dogleg Begins

• Beam images prior to the first dipole agree with
  simulations.
• Obviously the images after the dipole are
  truncated so the fit quality suffers.
• So we really dont believe the data in this plot.

39
Vertical Spot Size Along Beamline vs. Cavity
Strength
Dogleg Ends
Dogleg Begins
Cavity

• Agreement is much better here because of the
  better profiles
• We dont expect the cavity to change anything in
  the vertical plane, so we have confidence that
  the cavity polarization is good.

40
EEX Spectrometer Spot vs. k

• Vertical image size 61 mm 930 keV
• Horizontal image size 80 mm

Cavity OFF
Cavity 10
Cavity 20
Cavity 30
Cavity 40
Cavity 50
Cavity 60
Cavity 70
Cavity 80
Cavity 100
41
Current Status

• We are beginning to get results from the EEX
  experiment.
• We are starting to measure the EEX matrix
  elements.
• We have a bunch of data that we need to analyze
• Spot sizes, slit images, momentum spread,
  incoming bunch length.
• Certainly need to take more data to determine
  that the emittance is swapped.
• And a few operational questions to answer
• Is the cavity strength high enough?
• What causes double peak in the energy spread?
• This may be a recent RF problem in the 9 cell
• Possibly the laser
• Otherwise a head scratcher
• Some other smaller bug-a-boos to deal with.
• Interferometer installation is finishing up.
• Required for bunch length measurement after EEX.
• Tim may be seeing the first light at the end of
  the tunnel

42
Coherent Synchrotron Radiation
43
Coherent Synchrotron Radiation

• Coherent Synchrotron Radiation occurs when a
  short bunch passes through a bending magnet.
  Synchrotron radiation from the rear of the bunch
  can interact with the front of the bunch. This
  leads to

• Beam Energy loss
• Energy spread changes within the bunch
• Large decrease in energy at the tail of the bunch
• Smaller increase in energy at the head of the
  bunch
• Transverse emittance growth
• This is from the particles with changed energies
  not following the orbit given by the original
  energy deviation and dispersion. When the
  dispersion is eliminated, these particles have a
  betatron offset.

44
CSR Induced Emittance Growth

• CSR causes emittance growth because of the
  redistribution of the energy inside of the bunch.
  This causes particles to follow paths dictated
  by their energy loss. After a dipole this leads
  to position and angular offsets.

Figure Taken from A Bunch Compressor forSmall
Emittances and High Peak Currents at athe VUV
Free-Electron Laser Frank Stulle, Ph.D. thesis
University of Hamburg
45
CSR Induced Energy Spread Changes

• Saldin et.al. worked out the energy loss vs.
  longitudinal position in 1996. Their model
  assumes that the transverse sizes of the bunch
  are negligible.

Bunch Profile
15 cm in dipole
Energy increase
50 cm in dipole
60 cm in dipole
Steady state

• As the beam is in the dipole,
• the bunch head gains energy
• A longer dipole pushes the energy gain region
  forward
• Most of the beam looses energy

Bunch Head
Figure Taken from A Bunch Compressor forSmall
Emittances and High Peak Currents at athe VUV
Free-Electron Laser Frank Stulle, Ph.D. thesis
University of Hamburg
46
CSR Radiation Spectrum
wmax17 mm
Coherent part
Log P (W)
wc200 mm
Incoherent part
Log wc (rad/s)

• Calculated Energy spectrum for a 1nC bunch
  passing through dipole 3 of the A0 bunch
  compressor.
• Total radiated energy 10 mJ
• Incoherent loss scales as , Coherent
  losses scale as
• This spectrum does not reflect the shielding
  effect of beam pipe

47
Important CSR Parameters

• Overtaking length is the distance needed in a
  dipole for radiation generated at the rear of the
  bunch to catch up to the front of the bunch.
• Suppression Parameter calculates the maximum beam
  pipe radius one can use if one wishes to shield
  against CSR.
• Derbenev Criterion is a relation between the
  transverse size, bunch length, and bending
  radius. A small number (compared to 1) means
  Saldins model is valid. If this is comparable
  or larger than one, the model overestimates the
  effects on the beam.

48
CSR at A0

• CSR might be seen in two places at the
  Photoinjector
• The old bunch compressor
• The emittance exchange experiment
• Prior to decommissioning the bunch compressor we
  took data to see the effect of CSR on the
  electron bunch.
• The vacuum chamber did not allow measurement of
  the CSR itself.
• Preliminary CSR measurements were done with the
  EEX line.

49
The A0 Bunch Compressor

• A0 Bunch Compressor was a 4 dipole C type
  compressor.
• Vital Statistics
• Beam Energy 12.9 MeV
• 9 cell phase -39 off crest
• Path Length 1.89 m
• Magnet Arc Length 0.2904 m

• Magnet Bend Angle 22.5
• Maximum Dispersion 14 cm
• R56 8 cm
• Compression Ratio 71

50
CSR in the Bunch Compressor
Dipole 3 and the Spectrometer may produce some
CSR related effects. But the Derbenev
criterion is not satisfied so the model is not
so accurate.
Dipoles 1, 2, and 4 are not long enough for the
radiation at the rear to overtake the front of
the bunch

• This table shows some calculated important
  parameters for CSR in the A0 bunch compressor.
• CAVEAT EMPTOR The energy spread and emittance
  growth assume the 1D model is accurate.
• CAVEAT EMPTOR The magnets are close together so
  SR generated in one dipole can effect the beam in
  the nextnot included in calculation.
• WALK AWAY WITH CSR could be a factor in the
  compressor and the spectrometer. 3D simulations
  are needed.

51
CSR in the Bunch Compressor

• Beam Images after spectrometer for 3 situations
• 16MeV, uncompressed, small energy spread
• 12MeV chirped but uncompressed
• 12MeV chirped and compressed
• For 3 different charges
• 1, 2, 2.7 nC/bunch
• The plan was to compare the energy spread in each
  case to see if CSR is necessary to explain the
  data in the compressed case.
• ASTRA was used for all straight sections
• Including the gun and 9 cell
• Including 2D space charge
• 3D field maps.
• CSRTrack simulates compressor and spectrometer
• CSRTrack does not differentiate between velocity
  (space charge) and acceleration (synchrotron
  radiation) fields
• Simulations done with fields off and on

52
Some Beam Images

• Below are typical background subtracted images of
  1nC bunches after the spectrometer.
• Beam Energy and chirp are identical in each case.
• Higher Energy is to the left.
• Image is 6.1cm across. This is equivalent to
  2.3MeV.

12 MeV Uncompressed
12 MeV Compressed
53
Some Beam Profiles

• 47.8 keV/mm _at_ 16 MeV
• 37.2 keV/mm _at_ 12.2 MeV

• Profiles (actually projections) are extracted
  from images
• Each profile above is for 1nC/bunch. No attempt
  has been made to normalize the profiles to one
  another.
• Camera gains where adjusted to get the best images

54
16MeV Comparison 2.7nC
WHAT!!!

• 47.8 keV/mm

• Simulation scaled vertically and shifted
  horizontally to align peak.
• Comparison is for 3nC, as the higher charge
  should enhance any CSR effects.
• Agreement is reasonable for the CSR on case.
• Reminder This is an uncompressed beam.

55
Why Do We See CSR Effects in the Spectrometer??

• The arc length is 30 cm.
• The overtaking length in the spectrometer is 25
  cm.
• The Derbenev criterion is 0.35
• Uh Ohh.
• Saldins model predicts
• An energy increase of 4keV
• The horizontal emittance should grow 12
• The simulations predict
• An energy increase of 8keV
• An emittance increase of 9
• Most importantly The simulation and calculation
  more closely replicate the data!
• Did we see CSR??
• We did not have a port to look in when this data
  was taken.
• We do now but havent looked yet, we will soon.
• This will give us confidence that we know what we
  are talking about!

56
12 MeV Uncompressed Comparison 2.7nC
Screen Edge

• 37.2 keV/mm

• Comparison of the chirped, uncompressed beam with
  simulations.
• Simulated energy spread is larger than the data.
• Shape is not well matched either

57
Compressed Beam Comparison 2.7nC
Screen Edge

• 37.2 keV/mm

• Simulations with and without CSR are shown here.
• Qualitative agreement is reasonable.
• Including CSR does not enhance the agreement.

58
CSR in the Bunch Compressor Roundup

• We took beam images after spectrometer for 3
  situations
• 16MeV, uncompressed, small energy spread
• 12MeV chirped but uncompressed
• 12MeV chirped and compressed
• For 3 different charges
• 1, 2, 2.7 nC/bunch
• CSR was not a major effect in the bunch
  compressor.
• Surprisingly, the spectrometer is showing hints
  with an uncompressed beam.

59
Emittance Exchange Quick Reminder

• Incoming beam has energy chirp that partially
  compresses the bunch at the DMC.
• Quads prior to the first dipole focus the beam at
  the cavity.
• First dogleg provides dispersion at DMC.
• The cavity reduces the momentum spread of the
  beam and crabs the beam.
• The second dogleg finishes the exchange.

Diagram by Tim Koeth.
60
CSR in the Emittance Exchange Experiment
Same two trouble spots

• This is the table of CSR parameters for the EEX
  experiment.
• It appears that CSR is only an issue in Dipole 3
  and the Spectrometer.
• However, the 1D model is suspicious in those
  places so 3D simulations are needed.
• BUT, looking does not require simulations
• So that is what we did!

61
CSR in the EEX Experiment

• When the EEX experiment was installed, ports were
  included to measure any radiation from dipoles
  2,3,4, spectrometer.
• On December 20 we tried to measure any Coherent
  Radiation at dipole 3. We measured
• Coherent Transition Radiation prior to dipole to
  determine if the bunch was compressed.
• Radiation through straight-through port of Dipole
  3.
• Varied the charge to investigate charge
  dependence.
• The goal was not a precise measurement of
  anything. We wanted to know if we would see
  anything at all!

62
Experiment Setup
Fused Silica Window
Removable OTR Screen
D3
D2
2
1
Pryoelectric Detectors
Crystalline Quartz Window

• Pyroelectric detectors measure the THz radiation
  expected from the Coherent Transition and
  Coherent Synchrotron radiation.
• Transition radiation used to ensure that bunch is
  compressed.
• Not a measure of length per se, more of an
  optimization knob.
• Straight ahead detector measures edge radiation
  from D2 and synchrotron radiation from D3.
• Optimally a crystalline quartz window should be
  placed on both ports as fused silica reduces the
  transmission of certain wavelengths.

63
Results Detector 2 Signal

• Detector 2 shows increased signal with a
  compressed beam.
• Doubling the charge increases the compressed beam
  signal a factor of 3.
• The coherent power scales as q2.
• The incoherent power only scales as q.
• So we have some confirmation that some sort of
  coherent radiation is emitted!
• Be it edge radiation from D2 or synchrotron
  radiation from D3.

64
CSR in EEX Experiment

• We have evidence that some sort of radiation is
  occurring when a compressed bunch passes through
  D3.
• If we look back a few slides we see that
• The energy spread after the exchange is predicted
  to be 0.06 (without CSR) vs. the 1.4 prior to
  the exchange.
• Saldins model predicts the energy spread to
  increase to 0.3 after the spectrometer.
• If this is the only effect on the longitudinal
  emittance, the measured post exchange emittance
  grows from 5 mm-mrad to 40 mm-mrad.
• This is still less than the 120 mm-mrad at the
  start.
• The transverse emittance increase after the
  exchange line is predicted to be 8.
• 3D simulations are needed to understand what the
  effect on the beam will really be.
• Nonetheless we will still see an exchange.
• Weve seen that the 1D model can overestimate the
  effect on the beam.
• If we go to lower charges, we can reduce the
  effect and get a cleaner signature as well.

65
Conclusion

• A0 has an exciting AARD program.
• Ive highlighted a only two of the projects.
• A successful emittance exchange experiment at A0
  would be an exciting development.
• Especially for Tim!
• Coherent Synchrotron Radiation studies are
  underway.
• We have seen something.
• Weve detected radiation where we expected to.
• We see the effects on the beam where we didnt!
• The effects on the EEX experiment are still being
  understood.
• More studies are being planned.
• Stay tuned for more intriguing results!!!

66
Thanks for your attention