Top-Up Injection Schemes John Seeman, SLAC eeFACT, Daresbury October 24-27, 2016

Abstract: Top Injection for PEP-II The schemes for full-energy top-up injection into an e+e- collider will be discussed emphasizing technical issues to be accommodated. The issues include energy stability, energy spread control, transverse emittance reduction, transverse position and angle stability, collimation, background control in the detector, detector blanking of event triggers, and bunch filling strategies. Strategies and observations from PEP-II and KEKB will be included. Top-off injection was developed in PEP-II and KEKB from a linac injector to allow constant luminosity with the BaBar and Belle detectors, repectively, being fully operational during injection. The positron beam top-off in PEP-II was developed initially as its lifetime was the shortest and thus made the luminosity nearly constant. Second, the electron beam top-off was developed making the luminosity fully constant. For PEP-II either electrons or positron could be injection up to 30 Hz if needed, deciding pulse-by-pulse which beam (bunch) was desired. KEKB injector either electron or positrons as determined by the linac and transport line configuration. The implications for top-off into a circular Higgs factory will also be discussed. The injected beam emittances and septum parameters are presented.

Topics for Top-up Injection for Higgs Factory Required injection parameters for a circular e+e- Factory Briefly: PEP-II /BaBar top-up injection Briefly: KEKB/Belle top-up injection Detector background masking Top-up injection commissioning

Acknowledgements Many thanks to: PEP-II: S. Ecklund A. Fisher M. Sullivan J. Turner U. Wienands KEKB: Y. Funakoshi H. Koiso K. Oide S. Uno BaBar: S. Curry W. Kozanecki A. Perazzo C. O’Grady M. Weaver

Definitions (for this talk) Trickle injection = Top-Off = Top-Up

Technical Items Needed for Top-Up injection Measure each bunch charge in real time and determine when it needs refilling. In the injector, initiate a bunch to deliver it to the needed particular bunch (bucket) in the ring. Inject the bunch(es) into the collider with very low losses. Determine the injected beam backgrounds in the particle physics detector and find cures using collimation and launch conditions. Develop methods to monitor relevant backgrounds in real time for accelerator operators to tune on. Develop trigger masking for the detector physics taking by turn and with azimuthal variation.

Future e+e- Collider Parameters 100 km 54 km 27 km

Electron and Positron Production FCCee/CEPC stores about 6.2/1.9 x1013 e- and e+ per beam at the Higgs. FCCee/CEPC with 0.5 hr lifetime needs 4.5/1.4 x1013 e- and e+ per hour or about 1.5/0.5 x1010 e+ and e- per second at full energy (80% capture). Past production rate for positrons: CERN: LEP injection complex delivered ~1011 e+ per second. SLAC: SLC injection complex delivered ~6 x 1012 e+ per second.

Collider Injection Phase Space Septum blade Injection orbit Typical parameters: bx at injection septum (stored) = ~200m bx at injection septum (injection) = ~30m exstored (stored) =9.4 nm exinj (injected) =50 nm sxstored at septum (stored) = 1.4 mm sxinj at septum (injected) = 1.2 mm Xs = Septum blade thickness =~ 5 mm Xc = septum clearance distance = ~6sx Xinj < Ax Xinj = 4Xsinj+Xs+Xc = ~18 mm Ax = machine aperture > ~20 mm Stored Beam Orbit Stored beam Injection bump on Injection bump xc Injected beam xinj Ax Injection bump off (~one turn later) xinj

Top-up injection by full charge exchange Dump Kicker Old bunch Ring orbit New bunch Kicker Injector

PEP-II / BaBar Top-Up (Trickle) Injection Energy = 3.5 x 9 GeV Circumference = 2200 m One collision point (IR) at Lum = 1.2 x 1034 Full energy injection from Linac+Damping Rings Number of bunches = 1732 / ring Beam currents = 2.1 A x 3.2 A Particles = 1.0 or 1.5 x 1014 / beam (HER/LER) Lifetimes: Vacuum = ~10 hours Touschek = ~3 hours (LER) Luminosity = ~1 hour Lost particles per second = 4.2 x 1010 / second Top-up injection = one bunch / pulse, either e+ or e- Injection rate: ~3-15 Hz (30 Hz max) Particles per injection: 3 to 9 x 109 / pulse, selectable Bunch injection controller: pick the lowest bunch Injection efficiency = 50 to 80% Injection kicker pulse length = 0.4 msec Ring path length = 7.3 msec PEP-II BaBar

Actual vertical injection phase space in PEP-II (Decker et al)

PEP-II Interaction Region

PEP-II injected beams

PEP-II Injection Considerations for Top-up

PEP-II Top-up Injection Controls Charge/pulse typically inject "small quanta only“ Maximum trickle (injection) rate about 3/sec average when setup & trickling "Pseudo lifetime" Normal, DCCT-based beam-lifetime for LER un-useable "Pseudo lifetime" calculated from bunch currents avoiding those just injected. Minimum beam current fraction Avoid "trickling from scratch"

PEP-II Bunch Trains

Bunch injection controller Launches triggers Master pattern generator

PEP-II LER Bunch Injection Requests Different bunches have different beam-beam lifetimes and injection rates.

PEP-II top-up injection lengthened the average fill length Both top-up LER top-up No top-up

Reduced injection jitter (by accelerator) Ring orbit

PEP-II Integrated Luminosity Effect from Top-up Injection

Before PEP-II Top-up Injection

After PEP-II Top-up Injection

PEP-II/BaBar Top-Up Injection (Accelerator) Improved peak and average luminosity. Before Top-Up Injection After Top-Up Injection

PEP-II Run Time Improvement with Top-up

KEKB and Belle Detector

KEKB/Linac injection controller and parameters

KEKB-Belle Top-up Injection

Belle Triggers and Masking

KEKB normal injection (<2005)

KEKB top-up injection (>2005)

Comparison of Top-up Injection Belle/BaBar

BaBar Injection Monitoring

BaBar setting of noise sampling gate

BaBar: details of setting noise windows

BaBar Endcap Calorimeter (EMC) signal analysis Starting point before energy fix Injection with energy fix 6.5 kHz = energy error

Top-up injection losses were higher with more stored current

BaBar trigger with LER top-up injection (1)

BaBar trigger with LER top-up injection (2)

PEP-II/BaBar Top-Up Injection (Detector) One turn BaBar trigger masking: Mask all of ring a few tens of turns. Mask injected bunch area for 1250 turns or about 0.9 msec.

BaBar Vetoes versus bunch number and turn number.

BaBar/PEP-II top-up improvements needed during commissioning What improvements did BaBar need (incomplete list) (~late 2004): Smooth out trickle-algorithm in BIC, avoid stoppage (including cleaning up BIC-MPG (bunch injection controller and master pattern generator) communication). Get EPICS bar-chart display showing rate of injection/bunch. Want display of total injection rate Get a hardware real-time injection indicator (pulsed LED or counter). Make sure injection (LESIT) feedbacks don’t stop if too many small quanta. Stabilize setup of quanta (intensity, energy). BaBar needed to update its interlocks (bypassed too many). Speed up refresh of injection-trigger histograms

FCC/CEPC Detector Masking Detectors for FCCee/CEPC need to mask injection bunches. Ramped “Storage Ring” injector: Mask entire ring for ~10 msec every 5 minutes  large injected charge and many bunches (~50-100). Int. luminosity loss ~10%. Ramped “Main Injector” style: Mask entire ring for ~10 msec every 15 seconds  small injected charge and many bunches (~50-100). Int. luminosity loss ~1%. Rapid “synchrotron injector”: Mask ~1/80 of ring for ~10 msec at 0.1 Hz  small injected bunch charge but few bunches (~1-3). Int. luminosity loss << 1%.

Top-up Beam Commissioning Tuning Tools Radiation detectors (diodes, xtals, SVTRAD) useable only when backgrounds are too high! Injection trigger counters count EMC triggers after injection pulse histogram of triggers vs time EPICS variables with integral counts FFT shows effect of beam-energy deviations. normalized to injection rate DCH current quite useful for monitoring of average background not fast, so not useful to assess injection spikes Trigger rate (L3) similar behaviour as DCH current

SPEAR top-up injected beam shape versus turn

Conclusions Top-up injection will work for a circular e+e- factory. Top-up or full charge exchange works. A full energy injector is needed because of the short beam lifetime. The detectors will need to mask out the buckets during damping of injected bunches during data taking but not the whole circumference. Singe bunch injection controller needs to be worked out in detail for both the accelerator and the detector. Commissioning can be complicated as many issues both on the accelerator and detector arise: mainly detector backgrounds and masking.