EMMA Extraction / Diagnostic line Bruno Muratori STFC, Daresbury Laboratory 01/09/08.

EMMA Extraction / Diagnostic line

Bruno Muratori

STFC, Daresbury Laboratory

01/09/08

• Septum is - 70 degrees• Dipole to close dispersion is

only 10 degrees• Requirements are:

– Small beam– Dispersion = 0 on exit of

dipole• Difference between the two

bends very large:– Introduce more

quadrupoles ?– Introduce different

dipoles ?– Extract at different

location ?

EPAC08: EMMA diagnostic line

Option 2:10 degree bend→ 2 × 5 degrees& more quads

Option 3: Different extraction→ -70 degree septum &43 degree dipole & 4 quads

Option 1: More quadrupoles

Possible solutions

Option1: More quadrupoles

• Impossible to have small beam together with a matched dispersion

• Number of quadrupoles appears to be insignificant in this

• Resulting quadrupole strengths huge ! (3 × injection)

• Difference in bend angles far too great

• Also does not work with genetic algorithm due to James Jones

• Can this be shown analytically ?

Option2: Splitting the dipole & extra quads

• Problem actually worsens– Dispersion not closed– Beta very large– Dispersion very big (m)– Quad strength very large

• Again number of quadrupoles appears to have no effect in this

• Therefore third option is the only option available

• Yet another change for the engineers to endure …

Option3: Changing extraction point

• Dispersion matched easily• Betas very low• Quadrupole specifications

same as injection line !• Line very flexible

New diagnostic line layout

• Additional length → could re-use six SRS quadrupoles !• Rest of line unchanged• Still have

– Tomography– Space for possible deflecting cavity– EO section

Diagnostic line (1)

tomography EO

spectrometer

Diagnostic line (2)

• Ease of matching & tunability of line should be reflected in control room … (I hope !)

• Extra room for all the diagnostics which may be required for any eventual upgrades

• Betas remain very low• Deflecting cavity may be

located in optimum position (if we have it)

• Quadrupole strengths remain low for entire line

Tomography (1)

• Beam quite small on screen– ≥ 230 µm– need good OTRs or YAGs (?)– need good cameras

• Due to space restrictions– only 3 screens– 60° phase advance / screen

Tomography (2)

• Example taken from FLASH (45° phase advance / screen)

courtesy Ch. Gerth

Measurements

• Energy– First dipole & spectrometer at end with OTRs

• Projected transverse emittance– Quadrupole scans & tomography 60° phase advance / screen– Equivalent set-up in injection line for comparisons

• Bunch length– EO monitor downstream of the tomography section– No profile information

• Possibility of introducing a transverse deflecting cavity (TDC) to measure additional bunch properties

σz

L

0x

x

deflecting voltage

deflector bunch

screen

z

TDC Resolution (1)

• In absence of quadrupoles resolution increases with distance (L) from TDC to screen

σz

x

deflecting voltage

deflector bunch

screen

z

TDC Resolution (2)

• In the presence of interspersed quadrupoles this is not so and we must take into account of the entire transfer matrix from TDC to screen – there can be as many quadrupoles as desired

01 11 12''01 21 22

xx R R

xx R R

• Transverse displacement on screen is

• Beam size on the screen

• Transfer Matrix to screen gives

βd – deflector, βs – screen

• Want R12 big → sinΔψ = 1, βs fixed → make βd large

Transverse deflecting cavity (1)

deflecting cavity tomography EO

spectrometer

Transverse deflecting cavity (2)

Transverse deflecting cavity (3)

deflecting cavity tomography EO

spectrometer

0.95

1.35

1.6

Δµx = 90°

Δµy = 65°

1.13

Transverse deflecting cavity (4)

• Reverse of formula gives requirement of cavity voltage

• Take Δµ = 65° and φ = 0• For streaked bunch to be comparable to un-streaked bunch

• βx,y = 9 m at the deflecting cavity therefore we need, assuming an emmitance degradation to 10 µm and a bunch length of 4 ps

eV0 ≥ 0.23 MV @ 1.3 GHz

• Equality gives a streaked beam which is √2 times un-streaked beam– only rough idea of requirements– not enough for ≥ 10 slices (what we would like) → ~ 1 MV ?– longer bunch lengths / better emittance → lower voltage

20 0| sin cos |

N

z d

eV pcm c

Measurements with TDC

• Slice emittance & transverse profiles given by

– knowledge of R12 from TDC to screen

– one dimension on screen gives slice emittance– other dimension gives bunch length

• Slice energy spread given by– streaked beam and spectrometer

12 sind sR 01 11 12''01 21 22

xx R R

xx R R

• R matrix from TDC to screen known• Only need width at screen → slice emittance• Bunch length given by knowledge of kick

Longitudinalslices ~250 μm

Slice emittance measurements @ FLASH

courtesy Ch. Gerth

Energy

Time

Head

Tail

Unstreaked bunch

Energy spread measurements

courtesy Ch. Gerth

• FLASH operation at 650 MeV

Energy spread measurements (FLASH)

courtesy Ch. Gerth

Conclusions / Discussion

• Nice solution at long last … (Thursday of last week )– Beam very small throughout line

• Lots of tests still to be done but looks very flexible & workable• Steerers / correctors still to be included• Matching much easier than used to be (plus it works ! )

– Indication that this should be better during operation as well• Good characterisation of the beam at injection & extraction even

without TDC• Have good location for TDC should it be used in the future

– Realistic voltage parameters– Extra beam properties not available with EO– Currently looking at requirements for TDC with RF engineers