Low Emittance Tuning (LET).Through Dispersion Free Steering.

Low emittance:requirments for SuperB

HER e+εx = 2 nmrad εy = 5 pmrad

LER e−εx = 2.5 nmradεy = 6.2 pmrad

SuperB e+e- collider luminosity is:

This requires very low emittances for both rings:

Low Emittance Tuning (LET) Algorithm

Within the reach of 3rd generation light sources

S.M.Liuzzo PhD at LNF-INFN [Slide 2]

L.E.T. CORRECTORS USED• V & H steerers • Skew quad gradients• Bpm Tilts-Gains (in progress)

Only few measurements:1 to ~10 RM columns , orbit (P) and dispersion (η) TOTAL TIME: 5 min

dispersion

1 Off-diagonal block ORM column

1 diagonal block ORM column

P is the orbit H or V

SVD inversion of RESPONSE MATRIX M determines the Correction

DFS

Coupling

β-beating

Time

• Matrix M simulated from Model• SVD inversion for simultaneous

minimization of dispersion coupling and β-beating

LET compared to ORM and DFS

Vertical Emittance and vertical correctors rms as a function of the eigenvalues

cut used in SVD pseudo-inversion.

Vertical alignment errors corrected with:

Orbit Response MatrixDispersion Free Steering

Low Emittance Tuning

Vertical Correctors rms grows while vertical

emittance is reduced for all correction algorithms

εy

rms V cor.

ORM

DFS

LET

S.M.Liuzzo PhD at LNF-INFN [Slide 4]

Simulations of correction

S.M.Liuzzo PhD at LNF-INFN [Slide 5]

SuperB HER V12 Arcs rms misalignments:

Quadrupoles [0 300]μm ∆H, ∆V

[0 300]μrad ∆φSextupoles

[0 300]μm ∆H, ∆VBPM

Offsets [0 300]μm168 H-V correctors

168 H-V BPM

Produced table of tolerances for SuperB

R.Ba

rtol

ini

Comparison between LET and LOCO

S.M.Liuzzo PhD at LNF-INFN [Slide 6]

Simulations for Diamond

Simulations: 50 machine with random misalignments and bpm offsets, corrected using DFS and LET

εy =5.28 10-12

εy =3.27 10-13

DFS

LET

Correction applied in two iterations using H and V steerers only. No BPM Tilts

Comparison between LET and LOCOMeasurements at Diamond

S.M.Liuzzo PhD at LNF-INFN [Slide 7]

Coupling [%] H emittance [nm] Lifetime [h] Current [mA]

Decay mode900 bounches

150 mA

Initial conditions 21.1 h

After 2 LOCO iterations with skew quad and quad

After 4 LET iterations with skew quad

LETLOCO LOCO

5.9 h

5.5 h

Final Lifetime measurements

preformed after injection

BPM Tilt estimate by LOCO

Coupling estimated from lifetime:

εy =1.6 10-12 m rad (LET)

Same measurements. Parameters

vs iteration number

350 μm (LOCO:700 μm)

Vertical dispersion [m]Lifetime [h]

Coupling from Lifetime [h]Residual Vertical orbit

Residual Vertical orbit + H corrCurrent*Lifetime [mAh]

LET LET REDLOCO BLUE

818 mAh

881mAh

32 μm (LOCO: 1 μm)

5.5 h

5.9 h

0.06 %

0.07 %

Measurements at SLS

S.M.Liuzzo PhD at LNF-INFN [Slide 9]

Measurements aimed to achieve low vertical emittance

Same Tool used for Diamond, modified for direct access to Control System

SWISS LIGHT SOURCE2.4 GeV, 288m, 12 beamlines, 400 mA, 5.4 nm Hor. Emit.

Vertical beam size measurements performed using vertically polarized Syhnchrotron Light Monitor

TUPB27 Proceedings of DIPAC 2007, Venice, Italy

Measurements at SLS

S.M.Liuzzo PhD at LNF-INFN [Slide 10]

Top-up mode400 mA

Normal user operations

Beam size measurement for 3 subsequent correction performed using only

VERTICAL STEERERS

400 μm residual vertical orbit.

Best observedσy =6.8 μm

SLS best σy = 4.9 μm is obtianed using skew quadrupolesTested also

Horizontal correction and Skew quadrupoles correction, but still work is in progress.

Conclusion

S.M.Liuzzo PhD at LNF-INFN [Slide 11]

• LET Algorithm performing well but still not reaching simulation expected results.

• Need to further test and debug to control H corrections and Tilt detection

• Future application in DAΦNE

• Developed and applied Low Emittance Tuning Tool for SuperB.

• Tested and compared algorithm to LOCO at Diamond (skew quarupoles)

• Tested algorithm at SLS with promising starting results. (same tool as dimond)

Summary