Beam Feedback System Challenges at SuperKEKB Injector LinacL-band+LAS capture section 50 Hz (e+ or e-) pulse-by-pulse mode switching S-band linac 2.5 GeV e- 0.1nC x 1 PF ... ARC! e+Target!

K.Furukawa, KEK, Oct.2013. ICALEPCS 2013

SuperKEKB Injector Feedback Challenges

Beam Feedback System Challenges at SuperKEKB Injector Linac

Kazuro Furukawa, Ryo Ichimiya, Masako. Iwasaki, Hiroshi Kaji, Fusashi Miyahara, Tatsuro Nakamura,

Masanori Satoh, Tsuyoshi Suwada

KEK, Japan

<[email protected]>



SuperKEKB and electron/positron Injector Linac u  Asymmetric e+/e- Collider for B-physics

v Flavor structure of elementary particles, and possibly new physics

u  40-times higher Luminosity compared with that of KEKB (WR) v Twice larger storage beam à Higher beam current at Linac v 20-times higher collision rate with nano-beam scheme 　 ³ à Low-emittance (even at first turn) à Low-emittance beam from Linac ³ à Shorter storage lifetime (à Higher Linac beam current)

u  Linac challenges v Low emittance e- ³ with high-charge RF-gun

v Low emittance e+ ³ with damping ring

v Higher e+ beam current ³ with new capture section

v Emittance preservation ³ with precise beam control

v 4+1 ring simultaneous injection

2

2x beam current

40x Luminosity

RF-gun

e+ efficiency



Linac Upgrade for SuperKEKB u  Higher Injection Beam Current

v To Meet the larger stored beam current and shorter beam lifetime in the ring v 4~8-times larger bunch current for electron and positron

u  Lower-emittance Injection Beam v To meet nano-beam scheme in the ring v Positron with a damping ring, Electron with a photo-cathode RF gun v Emittance preservation by alignment and beam instrumentation

u  Quasi-simultaneous injections into 4 storage rings (PPM) v SuperKEKB e–/e+ rings, and light sources of PF and PF-AR v  Improvements to beam instrumentation, low-level RF, controls, timing, etc

3

4.0 GeV e+ 4nC x 2

3.5 GeV 10nC x 2 (prim. e-) 5nC x 2 (inj. e-)

1.1 GeV Damping Ring circ. 136 m

ECS

high current gun

Energy-spread Compression System

Bunch Compressor

7.0 GeV e- 5nC x 2

low emittance RF gun

SuperKEKB injector Linac

e+ target & L-band+LAS capture section

50 Hz (e+ or e-) pulse-by-pulse mode switching

S-band linac

2.5 GeV e- 0.1nC x 1

PF

HER

LER PF-AR 6.5 GeV e-

C-band modules

3T RF gun

Ir5Ce

FC



Main features of controls at (previous) KEKB u EPICS as Main control Software Toolkit v Provided a robust basis of equipment controls v Reduced software design efforts much

u Scripting Languages for Operational Software v SADscript/Tk, Python/Tk, Tcl/Tk used much ³ Especially, SADscript as a bridge btw. Accelerator simulation,

Numeric manipulation, Graphic interface and EPICS controls

v Bright new idea in the morning meeting could make the operation much advanced in the evening ³ Great tool to optimize the operation

4

KEKB



SuperKEKB Controls u Inherit Good part of KEKB Controls v EPICS v Scripting languages

u EPICS Channel Access (CA) Everywhere v Embed EPICS control software (IOC) everywhere

possible

v Reduce efforts on protocol design, testing, etc

u Dual Tier: Another layer in addition to EPICS/CA v Event system helps EPICS with another channel v Additional functionality, synchronization and speed

5

SuperKEKB



Dual-tier Controls u IOC controls via Conventional EPICS CA

³ Above 1ms, ordered controls

u Fast FPGA controls via SFP/Fiber (MRF) ³ 10ps ~ 100ms, 114MHz synchronous controls

6

IOC

EVG

IOC

EVR

IOC

EVR

IOC

EVR

OPI

Channel Access

FPGA/SFP

Dual-layer Controls



7

Fast, Global, and Synchronous Controls

u 114.24MHz event rate, 50Hz fiducial

u ~200 dynamic 50Hz-analog/timing param. u Multi/single-mode fiber

u Timing precision is < 10ps. v  < 1ps with external module.

u  MRF’s series-230 (and SINAP's) Event Generator / Receivers u  VME64x and VxWorks v5.5 / v6.8 u  EPICS R3.14.x with mrfioc1/2 u  18 event receivers in KEKB u  More Home-grown receivers

ARC!

e+ Target e+ BT (KEKB: 3.5GeV, 2nC)!

e– BT (KEKB: 8GeV, 2nC, ! PFAR: 3.0GeV, 0.1nC)!

e– BT (PF: 2.5GeV, 0.1nC)!

Event Receivers"

Event Generator"

e- Gun!

Cont-1"

KL_51/52"

SB_5"SB_4"SB_3"

Cont-5"Cont-4"Cont-3"Cont-2"

SB_2"

KL_B5/B6" SB_A"SB_B"

Central"

Cont-ABC"

SB_1"SB_C"

SH_A1" Injection"

96ns >100ns

>100ns 96ns

Dual-layer Controls

Configuration in 2010

to be expanded



Single Machine, Multiple Virtual Accelerators (VAs) u  Pulse-to-pulse modulation (PPM), one of the VAs is active at a time u  Independent parameter set for each VA, ~200 parameters are switched

every 20ms pulse

8

Dual-layer Controls

ARC e– (2.5GeV, 0.2nC) e− Gun

ARC

e+ Target

e+ (4GeV, 4nC)

e− Gun

Event-based Control System

PF Injection

SuperKEKB-LER Injection e– (3.5GeV, 10nC)

ARC

e– (7GeV, 5nC)

e− Gun

SuperKEKB-HER Injection

ARC e– (6.5GeV, 5nC)

e− Gun

PF-AR Injection

Damping ring Every 20 ms



Two kinds of Virtual Accelerators (VAs) u Virtual Accelerator Models (VAs)

u PPM VAs, that represent independent accelerator views for different purposes v by using dual-tier controls

u Simulation VAs, that represent physics models of accelerators in simulation codes v by using SADscript codes through EPICS CA, normally v even before the machine completion

9

Dual-layer Controls



(Offline) Orbit Correction Test in Sim. VA

Horizontal displacement

Corrector K0

Orbit before correction

Orbit after correction

Vertical displacement

Sugimoto et al.



Beam Feedback Loops u Equipment has inherent instabilities caused by many

sources v At the beginning of the KEKB project, we had to install many

feedback loops for beam energy, orbit, charge, etc. v Simple PID (mostly Proportional and Integral), with limits

v Scripts for prototypes, then ePID on IOCs

u SuperKEKB with demanding beam specification may require further considerations v Emittance preservation v PPM VA handling

u Simulation VA may help organize the loops

11



Energy Stabilization

12

u Energy instability was sometimes found v Closed feedback loops were formed

v Beam positions were measured where dispersion function is large

v RF phases at adjacent stations were changed

v Loop parameters were beam mode dependent

v Energy spread feedback using multi-electrode monitor was also implemented



u If Orbit became unstable v Beam positions were measured where betatron phases are

90degree apart v Corresponding steering magnets are adjusted

v If the orbit fluctuation was large, weighed average of BPMs based on response functions to beam kick or energy change

v The same method was applied to equipment stabilization

Simple Orbit Stabilization

13



Emittance Preservation u If Device is off center of the beam v Focusing magnet (quad) kicks the beam bunch v Accelerating structure (cavity) excites wakefield, to bend the tail

u Distorted bunch in banana shape v Emittance dilution or blow-up v Depending on the beam optics and the beam charge

u Orbit correction is crucial to preserve the emittance

14

Focusing Magnet

Beam

Accelerating Structure

Transverse distribution in time direction

Sugimoto et al.



Emittance Dilution u Offset injection may solve the issue u Orbit have to be maintained precisely

15

100 samples

Mis-alignment leads to Emittance blow-up

Sugimoto et al.

Orbit manipulation compensates it



u Closed loops were installed on each PPM VA independently v Tested at KEKB

Multiple Closed Loop Controls Overlapped

16

ARC!

e+ Target"

e– BT (PF: 2.5GeV, 0.1nC)!e− Gun!

ARC!

e+ Target"e+ BT (KEKB: 3.5GeV, 0.6nC)!

e− Gun!

ARC!

e+ Target"e– BT (KEKB: 8GeV, 1.2nC)!

e− Gun!


PF Injection

KEKB-LER Injection

KEKB-HER Injection

Primary e– (4GeV, 10nC)

Dual-layer Controls



ARC

e– (2.5GeV, 0.2nC)

e− Gun

ARC

e+ Target e+ (4GeV, 4nC)

e− Gun

PF Injection

SuperKEKB-LER Injection e– (3.5GeV, 10nC)

ARC

e– (7GeV, 5nC)

e− Gun

SuperKEKB-HER Injection

ARC

e– (6.5GeV, 5nC)

e− Gun

PF-AR Injection

Damping ring


Every 20 ms

F.B F.B

F.B

F.B

F.B

F.B

F.B

F.B

u Four PPM VAs at least for SuperKEKB (maybe with additional PPM VA for stealth beam)



Feedback loop life cycle u Installation of feedback loops v for beam properties like beam orbit, energy, energy

spread, bunch length, charge, dual-bunch equality, as well as device properties like llrf, timing, magnetic field, etc

v often without knowing the origin of the instability

u Instability hunting, and fixing u Beam study, or machine study using beam

u If fixed, remove corresponding feedback loop

18



Real Accelerator

EPICS Controls

Independent PPM Parameters Common Parameters

Event Synchronous Controls

Simulation Virtual Accelerators

Display and Analysis

PPM Virtual Accelerators

System Construction

Multiple

Multiple Multiple



Conclusion u SuperKEKB requires demanding beam parameters at

injector linac v Even under PPM

u Models of PPM VAs and simulation VAs may help design feedback loops v Energy, orbit, energy spread, charge

v Emittance preservation (via orbit stabilization)

u We should be prepared for full commissioning from 2015

u With Phronesis (Greek: Practical wisdom, Ability to understand the

Universal Truth), we believe we can achieve the goal 20



SuperKEKB dual rings

Mt. Tsukuba

Linac

PF-AR

PF

21

Thank you

Documents

Beam Feedback System Challenges at SuperKEKB Injector LinacL-band+LAS capture section 50 Hz (e+ or e-) pulse-by-pulse mode switching S-band linac 2.5 GeV e- 0.1nC x 1 PF ... ARC! e+Target!