3rd International Workshop on Beam Orbit Stabilization - IWBS2004December 6-10, 2004

Hotel Kirchbühl, Grindelwald, SWITZERLAND

Orbit Stabilization

at

Taiwan Light Source

Kuo-Tung Hsu

December 6, 2004

NSRRC Accelerator

Storage Ring(1.51 GeV)

Booster Ring

(1.51 GeV)

LINAC

Transport Line

Bending Magnet

Quadrupole Magnet

Pulsed Injection Magnet

Sextupole Magnet

SRF Cavity

Insertion Device

BM2

BM1

W20 (1.8 T)

U9 (1.25 T)

SWLS (6 T)

EPU5.6 U5 SRF Cavity

(Dec. 2004) SW6 (3.2T)IASW6 (3.5 T) x 3 sets

December 2004

4 Conventional IDs

2 Superconducting IDs

22 Beamlines

56 End-stations

Lattice type Combined function Triple Bend Achromat (TBA)

Operational energy 1.5 GeV

Circumference 120 m

RF frequency 499.654 MHz

Harmonic number 200

Natural beam emittance 25.6 nm-rad

Natural energy spread 0.075%

Momentum compaction factor 0.00678

Damping time

Horizontal 6.959 ms

Vertical 9.372 ms

Longitudinal 5.668 ms

Betatron tunes horizontal/vertical 7.18/4.13

Natural chromaticities

Horizontal -15.292

Vertical -7.868

Synchrotron tune 1.06*10-2 (SRF, 1.52*10-2)

Bunch length 9.2 mm (Doris Cavities, 800 kV)

6.5 mm (SRF, 1.6 MV)

Radiation loss per turn (dipole) 128 keV

Nominal stored current (multibunch) 200 mA (~2004), 300~400 mA (2005~)

Number of stored electrons (multibunch) 5*1011 (SRF cavity, 1012)

Major Parameters of the Storage Ring

Efforts to Improve Beam Stability

Coupled-bunch instability :

RF gap voltage modulation (~ October 2004)

Superconducting RF (December 2004 ~)

(to accompany double the stored beam current).

Coupled bunch feedback system

Orbital stability:

Source Elimination

Ambient temperature, Water temperature, Enhance

data acquisition system, Vibration elimination, Power

quality improvement , Power supply improvement …etc.

Orbit feedback system

Top-up injection

4. Vibration

6. Control System

5. SensorsImplementation

3. Thermal-Mechanical Effects

2. Heat SourceStabilization

1. Utility Capacity Improvement

< 0.1

<1 m rms

<0.3%

< 0.1

(drift<5 m)

~ 0.15<3 m rms

~0.5%

< 0.2

(drift~20 m)

~ 0.25> 1

(drift>50 m)

>1%

Performance, T

orbit

Io/Io

200220012000199919981997

Utility building #2 construction

CHW capacity improvedCTW, CHWStability

Injector energy upgrade

full energy injection

EPU air temp.U5 air temp.

SCADA implementationelectrical sensors

Water control system upgrade

VF controller implementation

Utility data archive system

AHU controller upgradeAHUs reorganized

Air temp. sensorsPosition sensors

Global effect studies

(air/water temp., cable heat)

Girder, thermal insulator , vacuum chamber, SR heat mask

BL-, RF-DIW temp.

AHU, crane

vibration pre-reduced Damping study Floor meas.

Piping improv.

Program initiation

Power supply heating

I monitor

Mechanical Related Source Elimination

(2003)

~ 0.06%

(J. R. Chen, et. al.)

Magnet (Water Temperature)

100 200 300 40022

24

26

28Mag-Water Temp.

Beam Position

Time(min.)

Tem

per

ature

(¢J)

0.22

0.23

0.24

0.25

0.26

Posi

tion (

mm

)

Caused by the temperature fluctuations of magnet cooling water

Magnet deformed ~ 10 m/

Induced beam orbit drift: 5 ~ 50 m /

Current status

Cooling water temp.: ~ 0.1

Girder Displacement

• Main cause: air temperature

Sensitivity to air temp.: ~10 m /

Induced beam orbit drift: 20-100 m /

• Current status: < 0.1 m per 8 hr shift

Air temp. : < 0.1

(utility control system improved)

Thermal insulator jacket

-200 0 200 400 600 800 1000 1200 14000.92

0.94

0.96

0.98

1.00

-200 0 200 400 600 800 1000 1200 140024.0

24.5

25.0

25.5

26.0

Beam Position

mm

min

Air TemperatureD

egre

e (

C)

0 25 50 75 10024

25

26

27

28

Dis

pla

cem

ent

(£gm

)

Tunnel Air Temp.

Girder Disp. Outer

Girder Disp. Inner

Time (Hours)

Tem

per

ature

(¢J)

-20

-15

-10

-5

0

Expansion of Vacuum Chamber

• Caused by synchrotron light irradiation.Sensitivity to water temp.: ~ 10 m /Move the girder (~ 0.3 m/ ) and BPM (~ 1 m/ )Induced beam orbit drift: ~10-30 m /

• Current statusVacuum cooling water temp.: ~ 0.5

0 6 12 18 2419.5

20.0

20.5

21.0

21.5Girder Displacement

Beam Current

Time (Hours)

Dis

pla

cem

ent

(£gm

)

0

100

200

Bea

m C

urr

ent

(mA

)

0 200 400 600 800 1000 1200 1400

-0.08-0.06-0.04-0.02

0 200 400 600 800 1000 1200 1400

0.51.01.52.0

0 200 400 600 800 1000 1200 1400

2425262728

0 200 400 600 800 1000 1200 1400

0

100

200

Beam Position

mm

min

BPM Displacement

um

Vac-chamber Temp

Te

mp

(C

)

Beam Current

mA

Power Supply Related High Frequency Orbital Noise

(Source Elimination)

@ 136 mA (August 22, 2001)

Beam Spectrum Observation

High Frequency Corrector Power Supply Noise

0 10 0 2 00 3 00 40 0 5 00 6 00 7 00 8 0010

-7

10-6

10-5

10-4

10-3

10-2

F req ue nc y ( H z )

Outp

ut

curr

ent

ripp

le

( am

pere

)

0 0.5 1 1.5 2 2.5

x 104

10-7

10-6

10-5

10-4

10-3

10-2

Frequency ( Hz )

Outp

ut

curr

ent

ripple

(

am

pere

)

Typical Corrector Spectrum

0 100 200 300 400 500 600 700 80010

-7

10-6

10-5

10-4

10-3

10-2

Frequenc y ( Hz )

Ou

tput

curr

ent

ripple

(

am

pere

)

0 0.5 1 1.5 2 2.5

x 104

10-7

10-6

10-5

10-4

10-3

10-2

Frequency ( Hz )

Ou

tput

curr

ent

ripple

(

am

pere

)

Before Improve

After Improve

(K.B. Liu et al.)

Before

Improve

RMS=0.03%

RMS=0.16%

Mar.08,2002

186mA

June 13, 2002

185mA

Observation at Infrared Beamline

After

Improve

Before

Improve

After

Improve

(Y. C. LO, et al.)

Software Environment for BPM Data Access

Closed Orbit Server

Static Database Dynamic Database(Slow Orbit, 10 update/sec)

Device Control

Service ProgramStatic Database Service ProgramData Update

Service Program

Database

Logging

Program

Real-Time

Trend Display

LabVIEW

Interface

Matlab

MEX

Interface

MX-BPM Hardware

Control Ethernet

Software

Components

On

Control

Consoles

BPM

VMEs

Crates(PPC/LynxOS)

Consoles

Computers(Compaq WS/Unix)

(PC/Linux)

Self-Configuration

Table

Data Acquisition

Reflective

Memory

(RM)Fast Orbit, 1

K update/sec

Machine Parameters

Service Routines

Machine Modeling

(MPAP,APAP)Control Applications

Data Access Routines

Data Access Routines

April 2003

Matlab

Scripts

AP

MATLAB

LabVIEW

VI

LabVIEW

System Status

Display

Alarm Checking

ProgramParameter Page

ProgramData Logging

& ArchivingHardware Diagnostics

Program

AP

Turn-by-Turn Server Program

Closed Orbit Data

Reflective

Memory

(RM)Fast Orbit, 1 K

update/sec

DBPM Hardware, LR-BPM Hardware

Mode Control BPM Data

Control Program

To

other

DBPM

nodes

Closed

Orbit

Server

+

MX-BPM

Data

Acquisition

DBPM

LR-BPM

Data

Acquisition

Node

Turn-by-Turn

Data Access

Environment of BPM System

Orbit Feedback System Architecture

Control

Network

Control

Consoles

PC/Linux

V

M

E

H

O

S

T

16

Bit

D

A

C

RM/

DMA

Private Ethernet Switch

Reflective Memory

Hub

Keithley 2701

PBPM Front-end

Keithley 2701

PBPM Front-end

Feedback

VME

Crate

Corrector Control

VME CrateOrbit Acquisition

VME Crates

RM/

DMA

16

Bit

D

A

C

16

Bit

D

A

C

Electron BPM

(MX-BPM & DBPM) Correction Magnet Power Supply

Photon BPM

V

M

E

H

O

S

T

RM/

DMA

16

Bit

A

D

C

16

Bit

A

D

C

16

Bit

A

D

C

Tim

ing G

enerato

r

VM

E H

ost

Po

werP

C 7

410

RM/

DMA

V

M

E

H

O

S

T

Feedback

Data Capturer

VME Crate

VM

E H

ost

Po

werP

C 7

410

PC/WindowsWS/UnixWS/Unix

Local RAM

Disk

V

M

E

H

O

S

T

RM/

DMA

16

Bit

A

D

C

16

Bit

A

D

C

Tim

ing G

enerato

r

16

Bit

A

D

C

Fib

er In

terfa

ce

Orbit Feedback System Summary

* 22 correctors + 30 BPMs for both plane

* Using measured response matrix

* Singular Value Decomposition (SVD) approach to invert the response

matrix

* Proportional, Integral and Derivative (PID) control algorithm

* Orbit acquisition rate 1 kHz

* Loop bandwidth is about 5 Hz now (30 Hz before December 2003)

* Unified system for global feedback and local feedback system

* Remote enable/disable of the feedback loop

* Orbit excursion reduced to less than ~ um level for ID operation

- U5, U9 and EPU5.6

* Suppress orbit leakage due to non-ideal dynamic local bump

- EPBM

* Prototype local feedback loop by using e-BPMs

Orbit Feedback System Summary (cont.)

* DSP feedback engine was replaced by general purpose PPC module

User friendly development environment

Lift maintenance difficults

Slightly increase jitter < 50 µsec (DSP < 10 µsec)

* Robustness enhancement of the feedback loop

BPM Check

RMS, data change rate, …etc

Limited correct setting range

Problems and Plan of the Orbit Feedback System

Major Problems :

* Loop bandwidth is too small right now

=> cannot eliminate mechanical related oscillation

=> Limited gap/phase changing speed of IDs

Short-term Plans:

* Increase sampling rate o 2 kHz to 4 kHz

* Modify corrector power supply regulation loop

* Increase loop bandwidth > 100 Hz

Closed Orbit : ~10 µµµµm rms with DC correction schemes.

(Peace Chang et al.)

Orbit distortions: < 10 µµµµm rms during insertion gap scan

can be compensated for using look-up correction tables.

(Peace Chang et al.)

Beam orbit stability: a few µµµµm level (peak-to-peak) with a

global feedback system. (C. H. Kuo et al.)

Closed Orbit and Orbit StabilityClosed Orbit and Orbit Stability

e-

Power

Supply

#1

Power

Supply

#2

Power

Supply

#4

Power

Supply

#3

EPBM

Model

Enable

Bump

Amplitude

Control

(0-5 V)

Bump Coefficient

(± 10 V)

Block Diagram of Control Electronics for

Elliptical Polarization from Bending Magnet (EPBM) ProjectDRAFT

Ratio

Amplifier

x 4

#1 #4#3#2

BM2 Bending MagnetEPBM VC #1

EPBM VC #2

EPBM VC #3

EPBM VC #4

EPBM Control Electronics

SYNC to

Experimental

Station

Clock

Generator

12 12

Corrector Strength

ILC12ILC03

#1 #2 #3 #4#1 #2 #3 #4

Line Driver x 2

EPROM

12 Bits

Counter

t1

t2

t2 : user specify

Ratio

Amplifier Instrumentation

Amplifier x 812 Bits DAC

++ + +

t1 : 40 msec

Polarization

Control

EPBM

Mode

Request

EPBM

Model

Request

EPBM

Mode

Enable

3.4o

1220 mm460 mm 375 mm 540 mm 345 mm

325 mm 135 mm 170 mm 175 mm

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0 5 10 15 20 25 30 35 40 45 50

Vert

ical P

ositio

n (

mm

)

BPM Index

Difference Orbit - EPBM 'on', DGFB 'off'

"ep0925f.dat"

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0 5 10 15 20 25 30 35 40 45 50

Vert

ical P

ositio

n (

mm

)

BPM Index

Difference Orbit - EPBM 'on', DGFB 'on'

"ep0925g.dat"

* EPBM dynamic bump with feedback

EPBM flip bump without feedback

Eliminated Orbit Leakage for EPBM Operation

(C.H. Kuo et. al.)

Insertion Devices in the TLS Storage RingInsertion Devices in the TLS Storage Ring

2005Apr.

1999

Dec.

1994

Dec.

2003

Mar.

1997

Sep.

1999

Apr.

2002

Installation Date

0.050.050.0330.0330.0360.0360.0360.0360.0080.0080.0110.011

((--0.012)0.012)

0.05040.0504

((--0.014)0.014)

Vertical Tune Shift Vertical Tune Shift

(Horizontal Tune Shift)(Horizontal Tune Shift)

19.610.463317.92.993.52

(2.37)

190.5Defection Parameter Ky

(Kx)

1400010015000140001500140038000Max.

50005800500060804000Min.Photon Energy

(eV)

[Used Range]

3.51.251.83.20.640.67

(0.45)

6Maximum By (Bx) Field

(Tesla)

7.548131676661.5Number of Periods

18.5182218181855(Min.) Gap (mm)

6920655.625Period Length (cm)

0.854.53.01.4043.93.90.835Magnet Length (M)

SCHyb.Hyb.SCHyb.PureSCType

Arc

2,4,6

S6S5S4/RFS3S2S1Location Section

IASWU9W20SW6U5EPU5.6SWLSInsertion Device

The photon flux vs. photon

energy of bending magnets and

IDs at TLS. Two Taiwan

beamlines, SP8-BM and SP8-

U3.2, at Spring 8 are also

depicted

with

Feed-forward

Tune

Correction

without

Feed-forward

Tune

Correction

Orbit stability of the NSRRC Storage RingOrbit stability of the NSRRC Storage Ring

With the follow-gap look-up table and global orbit feedback system, the orbit drift

of the field scan of U5, U9 and EPU5.6 can be reduced from a few hundred

microns to a few microns. When the gaps of U5, U9 and EPU5.6 are scanned from

19, 21, and 21 mm to 39, 41, and 41 mm, the closed orbit distortions indicated by

the R4BPM5X (blue) and R3BPM3Y (green), those are used in the faster orbit

feedback system, are within 2 microns in both horizontal and vertical planes.

(Peace Chang, et. al.)

Photon Beam Stability Monitor (∆∆∆∆Io / Io)

BM3

BM3

VFM

VFM

50 um pinhole

Sensitive to

Beam size instability

Orbit instability

Energy instability

Signpost of overall beam stability form users’ viewpoints

Demagnification factor : 3

Photo

DiodeThermal Deformation

Backlash

Driver Mechanism

(Optimization)

Electron Beam Stability

(Transverse,

Longitudinal,

Beamsize, …)

History of the Photon Beam Stability (∆∆∆∆I0/I0)

May Jan Jan Jan2001 2002 2003 2004

0

20

40

60

80

100

Ach

ieved

Perc

en

tag

e i

n U

ser

Tim

e

long shutdown long shutdown

Improve cavity

temperature controlat 30% user time

U5, U9, EPU dynamic tuning

at 100% user timeRelax U5, U9, EPU

dynamic tuning speed

(K.K. Lin et. al.)

1130 1140 1150 1160 1170 1180 1190 1200197

198

199

200

201

202

203

Time (min.)

Sto

red B

eam

Curren

t (m

A)

-15

-10

-5

0

5

Orb

it C

han

ge

(mic

ron)

Date: Dec. 31, 2002 Top-up mode Test

∆R4BPM5X (blue line)

*Orbit change referred to the data at 1130 minute is taken with 10 Hz sampling rate.

∆R3BPM3Y (red line)

Stored beam current *Injection threshold current is 198 mA

Top-up injection scheme planned.

Feasibility demonstrated.

(G.H. Luo et. al.)

TopTop--up Injection Projectup Injection Project

Experiment of Top-up injection

Filling pattern at the beginning of

top-up injection

Filling pattern at the end of top-up

injection

User’s shift decay mode Test of top-up mode

Zoom in, injection interval 2 min.

(G.H. Luo et. al.)

Cooling Water and Chamber Temperature

Outlet cooling

water temperature

Chamber’s

temperature

Top-up

test run

(G.H. Luo et. al.)

Compound Displacement of the of Beam Position

Monitor Reading

Stored beam current

in mAR1 BPM3 X-displacement

in µm

R1 BPM3 Y-

displacement in µm

δx ~15 µm

δy ~5 µm

-Intensity dependency of BPM electronic and BPM support fixtures (~ µm)

-Filling pattern dependency of BPM electronics (~ µm)

-RF gap voltage condition variation

-BPM support structure (~ 10 µm)

-Thermal related (~ 10 µm)

Top-up test run

(G.H. Luo et. al.)

R1BPM1Y

R2BPM7Y

R6BPM4Y

Vertical Beam Position Variation

20 µm

20 µm

20 µm

Top-up test run

R1BPM1X

R4BPM3X

R3BPM8X

Horizontal Beam Position Variation

20 µm

9 µm

5 µm

Top-up test run

Renew 1.5 µµµµsec half-since injection kicker to reduced jitter

and to improve waveform matching .

Gate of the orbit feedback loop is a provincial solution to

remedy mismatching problem of the injection kickers.

User experiment with injection gate and without is performed,

acceptable results get up to now.

Improve gun pulser and linac performance to improve filling

pattern control.

Studies of the injector reliability, injection efficiency, and

minimization of orbit perturbation during injection, etc., are

ongoing.

Top-up operation is scheduled in late 2005.

TopTop--up Injection Planup Injection Plan

223 m OD (700 m circumference)

139 m ID (437 m circumference)

Building

NSRRC in Hsinchu Science Park , Taiwan Site

720 kW (4 SRF cavities) RF Max. Power

4.8 MV (4 SRF cavities) RF Max. Voltage

500 MHz RF Frequency

1 % Coupling

< 2 nm·rad at 3 GeV (Distributed dispersion) Emittance

x 12 Bending-section

10.5 m x 6 ( v = 10.5 m, h = 160 m) 6 m x 18 ( v = 8 m, h = 110 m) 3 m x 12 ( v = 4.5 m, h = 250 m; In-achromat)

Straight-section

24-cell DBA Lattice

499.2 m (h = 832 = 26·13, dia.= 158.9 m) BR Circumference

518.4 m (h = 864 = 25·33 , dia.= 165.0 m) SR Circumference

400 mA at 3 GeV or 300 mA at 3.2 GeV (Top-up injection) Current

3 ~ 3.3 GeVElectron Energy

Parameters of the Proposed Taiwan Photon Source

Sub-µm orbital performance is one of a design goal.

Summary

* To achieve µm (sub-µm) orbital performance is a short term goal at

TLS.

* Further develop of fast orbit feedback system is needed in

following area:

• Corrector PS improvement

• High sampling rate (~ 2 KHz/4 KHz)

• PS control interface

• BPM system improvement in engineering as well as software

functionality.

* Top-up operation mode is scheduled, beam orbit stability will be

improve further.

* Sub-µm orbital performance is one of a challenge for the newly

proposed 3 ~ 3.3 GeV Taiwan Photon Source