ALICE 350kV setup(s) in 2012.
Status (expected)
Gun 350kV
Eb
6.5MeV / 26.0MeV
Bunch charge : 60pC or 80pC
INJ knobs:
Laser pulse length
Laser beam position on cathode
Bunch charge
SOL-01 ; SOL-02
Buncher power
BC1/2 gradients
BC1/2 phases
Steering
=================================================================
0. Laser pulse length
28ps or 7ps ?
Strong inclination to get rid of pulse stacker ...
 excludes uncertainty re. the pulse profile
 JLab did not use it (increases instability .. ?)
 may increase emittance ... but who cares ! ... and bad shape damages emittance
anyway
 may have “cleaner” longitudinal phase space
(old simulations, small Q, gun
BL only, indicate this !)
To model:
28ps (flat top) and 7ps (Gaussian) in “standard” setup
=================================================================
00. Laser beam position on cathode
Suggest ~ central on cathode
 ALICE is not IR FEL at JLab, we do not expect too much ion back bombardment
1

Ideally – such that the beam is central in SOL-01 (no need to re-steer while wobbling
solenoid)
1. Bunch charge
60pC (Period 12) :
230kV; re-Cs every 7-9 days; noticeable beam loading in booster at 40MHz (THz) .
80pC (feasible in Per. 13)
 350kV; buncher digi LLRF ;
 possibly booster digi LLRF at later stage (?)
 less THz work in 2012 (cells – very little)
 at 16MHz, Q ~ 16pC equivalent  OK !
 But: more frequent re-Cs (NOTE: exponential dependence!)
 But: beam quality could be lower than at 60pC
 But: available buncher power may not be sufficient
Decision:
 60pC as the base option
 80pC as a second option.
=================================================================
2. BC1/2 gradients
after gun
after BC1 (E1)
after BC2 (E2)
Period 12:
0.23MeV
4.0MeV
6.5MeV
Period 13
0.35MeV
4.5MeV
6.5MeV
Could lower BC1 gradient be beneficial ?
This could reduce beam energy drop at the start of BC1 hence better beam longitudinally !?
Both transverse and longitudinal dynamics will change ... Results may surprise ....
To model:
Compare beam quality at E1 = 4.5 & 2.5MeV; E2=6.5MeV in both .
=================================================================
SOL-01/02
Affect strongly the longitudinal phase space (and bunch length)
Model (80pC):
330G / 220G
THZ-04; 230kV; 60pC
Model; 230kV; 80pC
Model; 350kV; 80pC
SOL-01
300
280
330
SOL-02
210
220
220
SOL-01
Range restricted to provide small beam size at buncher ( aperture = 24mm)
2
Practical method to choose SOL-01:
 Find SOL-01 current, I1, at which beam size is minimal at INJ-1.
 Deduct (0.2-0.3)A.
To model:
1) SOL-01 / INJ-1 scan and compare with measurements (at chosen laser pulse)
2) Range: (290-330) G
SOL-02
Very critical !
Two beam can be seen on INJ-2 (booster exit) !
Practical methods (speculative) :
 find I2 such that no two beams seen on INJ-2
 if (2) below holds, measure emittance v SOL-02 and choose current at minimal
emittance
To model:
(1) Range (180-230) G at each I1.
(2) Relation between emittance and longitudinal phase space (LPS) quality on SOL-02: this
may provide a practical means for SOL-02 optimisation.
=================================================================
Buncher power
Does affect transverse dynamics but not too strongly ...
Model (80pC):
2.2MV/m (~1700W)
To model:
1) Range = (1.9 – 2.2) MV/m ; After SOL-01/02 established.
=================================================================

BC1 phase
Do not want to set less than (-10) deg [bad phase space; too short bunch length]
To model:
1) Range = [-10, -20] deg
=================================================================

BC2 phase
If we want the smallest development of the bunch from booster to linac, need to
minimise E :
∆𝐸
𝑆
∆𝑧 =
( 2 − 𝑅56 )
0.511𝛾 𝛾

To model:
1) Range = [+10, +30]
3
=================================================================
Steering
Centre the beam in both SOL-01/02 – this will simplify solenoids scans if necessary.
Use SOL-01 and HVCOR-06 correctors.
Use HVCOR-02 to get “nice” beam at the exit from the booster.
4
Guide: RESETTING INJ FOR 350kV
Initial INJ settings
Laser pulse stacker ON
Burst generator
x5
Bunch charge
60pC
Gun voltage
350kV
SOL-01
3.9A (330G)
SOL-02
2.6A (220G)
Buncher
1600W
BC1/2 Phases
-10 / +20 (BC2 phase such that ~ the minimal E on INJ-5 achieved)
BC1/2 gradients
4.5MeV / 6.5MeV (exits of BC1 and BC2)
Quads etc
~ THZ-04 setup
===================================================
Laser
 Set laser beam at ~ the centre of the cathode
 Set 60pC approximately
crudely : Q~ 100 x QE x LA
[QE]=%
INJ setup : first approximation
 Apply “Initial settings” (see above – apply THZ-04 BURT then modify gun BL
settings)
 Centre beam in SOL-01 (watch INJ-1 while wobbling SOL-01 and move the laser on
cathode)
 Centre beam at INJ-1 position: (i) HVCOR-01/06/02 set to zero; (ii) note position on
INJ-1 and move the beam with HVCOR-01 only towards the centre and overshoot by
~ 1/3 of the initial offset; (iii) centre beam with HVCOR-06 only.
 Buncher, BC1/2 phasing
 Centre beam in SOL-02 by looking at INJ-2/3 (quads off)
 Adjust HVCOR-02 to get “nice” beam post- booster
 Get the beam to INJ-5
 Set 60pC accurately
Gun studies (pulse stacker is still in !)
 Take image of laser beam on cathode (bright enough to determine beam profiles – this
will be used in ASTRA sims)
 Accuracy & repeatability of BZC
 SOL-01 scan on INJ-1
 Set SOL-01 = [current for MIN beam size on INJ-1] – 0.2A
Decision on pulse stacker (3 criteria)
 Phase buncher & BC1/2 as accurately as possible (BC phases = “initial”)
 Set buncher power = 1300 W
 Emittance (Q&D) v SOL-02
 Beam stability on INJ-5
 Measure “quality factor” on INJ-5 (SOL-02 set for MIN emittance)
 Remove pulse stacker
 Re-phase all accurately (phase change expected ~ 50deg)
 Set buncher power = 1600 W (this is to compensate for loss of bunching)
5





SOL-01 scan on INJ-1
Set SOL-01 = [current for MIN beam size on INJ-1] – 0.2A
Emittance (Q&D) v SOL-02
Beam stability on INJ-5
Measure “quality factor” on INJ-5 (SOL-02 set for MIN emittance)
Decide on whether to use laser pulse stacker ...
Settings for optimisation scans
Can fix some parameters from start
 SOL-01 (set as described above; “fixed” by the buncher aperture)
 Buncher power ~ 1600W (~ hard limit anyway)
The remaining essential parameters
 SOL-02 (controls both transverse & longitudinal)
 BC1 phase (controls both transverse & longitudinal)
 BC2 phase (controls mostly energy chirp; small transverse effect; small effect on
bunch length)
 BC1 gradient (lower = better ?) ; this should be decided mostly on the basis of
simulations
To model:
- reduce BC1 exit energy to 2.5MeV; keep 6.5MeV after BC2
- parameter scans (BC1 phase, SOL-02, Pb)
SOL-02 scans
At a given BC1 gradient (decided earlier).
Vary SOL-02 at BC1 phase = -5, -10, -15, -20.
Parameters to observe:
 INJ-2 (two beam structure)
 emittance
 at each phase and SOL-02 with lowest emittance, - “quality factor” on INJ-5
6
Effect of R51 and how to deal with it ...
NOTE: not even thinking about R52 ... !!!
R51 in INJ and its effect on overall bunch compression
If R51 is not killed at the end of the INJ line, the bunch is injected into the linac having an
elongated bunch with (z-x) correlation.
If this (z-x) correlation is NOT preserved within the linac, this initial elongation of the bunch
may become “uncorrelated” and will determine the minimal bunch length after the
compression chicane.
Question : does the linac preserve (z-x) correlation ? Difficult to believe in it ... because in
the linac : the (E-z) correlation is gradually established and this should distort (x-z)
correlation potentially existing at the entrance to the linac; not to mention RF focussing !
In any case, if the bunch longitudinal phase space has a chirp and is “smeared” by R51 and
this bunch is injected into linac, this “smear” will remain after final bunch compression thus
limiting the minimal possible bunch length (the extent of “smear” appears to reduce however
depending on the energy chirp the linac introduces).
If the bunch from the booster does NOT have a chirp, the bunch lengthening due to R51
does not seem to affect the minimal bunch length achievable after the compression
chicane.
R51 and compression chicane.
R51 is quite strong in compression chicane (~ 1.0 at the entrance to the last dipole and =0 at
its exit).
THz setup needs R51=0 in the middle of the last dipole
FEL setup needs R51=0 at the exit from the last dipole.
Hence two setups cannot be fully compatible wrt R51 effect.
MAX compression for FEL required here
THz generation point
R51 ~ -1.0 (entrance to last chicane dipole)
That could be the reason for strong effect of ST2-Q-03/04 on THz levels.
7
What if the beam size is very small in compression chicane ?
 will help if x is small everywhere in the chicane
 but, intuitively, will not help if the beam is small only in the last dipole but not
everywhere ... perhaps some interplay with R52 ...
How to take care of R51 ? ...
It appears we do need to take care of R51 at the end of INJ beam line (entrance to linac) .
How ?
From ALICE optical functions: R51 is killed when Dx and Dpx are killed. Are they related
in some mysterious way ?
At the exit of a dipole with a bend angle of 0 :
𝐷𝑥′ ≅ 𝛼0
𝑅51 ≅ 𝛼0
(NOTE: not quite good match with MAD from which R51 ~ 0.1 while the formula estimates
it as ~ 0.5)
Quads are acting in the same “direction” for both Dpx and R51 (although physics is
obviously different).
As a side note ... to kill Dx and Dpx at the exit of INJ-DIP-02 , the Q-05 should have
position/strength relation as follows:
(𝑘𝑙)𝑠 ~ 2.0
where s is the distance between the quad and the dipoles.
8
 R51 behaviour
Practical solution for INJ #1 (non-zero chirp):
For the bunch from the booster with an energy chirp.
Assuming that killing Dx and Dpx will automatically kill R51:
 Kill Dx at the exit from DIP-02 (watch BPM-04)
 Kill Dx at the exit from the dog-leg (watch ST1-1 screen and/or ST1-BPM-01 )
 Hope that R51 is also killed along the way ...
Practical solution for INJ #2 (zero chirp):
Looks like there is nothing to worry about ...
9
10