Cmw Isis Synch Talk

An Introduction to theISIS Synchrotron

Chris Warsop

Outline

• Introduction

• Acceleration and Trapping

• Transverse Focusing and Injection

• Extraction

• The Challenges of a High Intensity Machine

What does the Synchrotron do for ISIS?

• On Target Require (with 10 Hz Variations for Target 2)

2.5x1013, 800 MeV Protons, in < 1 μs pulse @ 50 Hz

3.7x1013, 800 MeV Protons, in < 1 μs pulse @ 50 Hz

• Linac Provides

2.8x1013, 70 MeV H-, in 250 μs pulse @ 50 Hz

4.0x1013, 70 MeV H-, in 250 μs pulse @ 50 Hz

• Synchrotron

Compresses Pulse Length: 250 → 1 μs

Accelerates Beam: 70 → 800 MeV

52 m

Basic Principles• Acceleration

Circular Machine

Repeated Use of Relatively Low RF Acceleration Voltage

ISIS Ring: Max ~140 kV per turn, (163 m), but over 104 turnsISIS Linac: ~ 70 MV in a single pass (50 m)

Synchrotron Principle

Keep Beam on Constant Bend RadiusVary Confining Field and Particle Momentum in Synchronism

• Compression

Multi-Turn Charge-Exchange Injection, Bunching and Acceleration

Motion in a Dipole Magnet - Uniform Field

BveF

v

BF

Be

P

Bevmv

2

• For particle momentum P, charge e, in uniform field B

Trajectory is Circular Arc, Radius

The Synchrotron

teBP

21

2

01

tBe

cm

c

Ltrev

dt

tdBe

dt

dP dt

tdBce

dt

d ][

• Many Parameters Determined by

Main Dipole Magnet Field Variation with Time: B[t]

dt

tdBeLturn

][

0.005 0 0.005 0.01 0.015Times0.2

0.3

0.4

0.5

0.6

0.7

dleiFT

Dipole Field vs Time

Magnetic Field on ISIS ….

]2[ tfCosBBtB magacdc

fmag = 50 Hz

Binj = 0.176 T

Bext = 0.697 T= 7.002 m

• Main Magnet Field is a Biased 50 Hz Sinusoid

• Allows Energy Recovery in Magnet Power Supply

• Main Dipoles and Quads all in series (White Circuit)

Acceleration

RF Acceleration

• Pass particles through sinusoidally excited RF cavity• Gain energy ε from field across accelerating gap

EeF

]2[][ 0 tfSinVtV RF

]2[0 tfSineV RF ][. tVesEesF

E

dt

tdBeLturn

][

Needs to match energy gain requirements

Acceleration in a Ring

tfSineVdt

tdBeL RFturn 2

][0

• Lock RF frequency to a harmonic of revolution frequency

• Define ideal “Synchronous” particle, which follows

• This particle sees the same phase of RF on every turn

- surfs on wave propagating around machine

• Only two points on the wave where particle may receive correct energy

- Synchronous Phase

0 0.2 0.4 0.6 0.8 1Timeus10

5

0

5

10

dleiFVk

Accelerating Field vs Time

Acceleration Parameters on ISIS

• Two bunches on ISIS REVRF ff 2

• Using relations above with ISIS B[t], get

]2[0 tfSineV RF dt

tdBeLturn

][

21

2

012

tBe

cm

L

cfRF

0 0.002 0.004 0.006 0.008 0.01Times

1.5

1.75

2

2.25

2.5

2.75

3

FR

ycneuqerFzHM

RF Frequency Law

0 0.002 0.004 0.006 0.008 0.01Times0

20

40

60

80

stloVnruT

Vk

Minimum Volts per Turn Required

ISIS Ring RF System Details

• Harmonic Number 2 → 2 bunches

• 6 Tuned Resonant Cavities– Sweep 1.3-3.1 MHz

– 0 - 26 kV per cavity

– Supply ~150 kW of beam power

• Ferrite Loaded Coaxial Resonator– Bias current 200-2000 A

• RF Driver– Two 250 kW tetrodes per cavity

• Multiple Control Loops– Frequency

– Voltage

– Beam position, current etc.

• 4 additional cavities for DHRF

Finite Bunch Length, Oscillations in the Bunch

• What about particles not at correct phase?– i.e. non synchronous particles?

• Depends on Revolution Time– variation of velocity with P– variation of path length with P

• On ISIS obeys ‘common sense’– Higher energy particles take less time – Lower energy particles take more time

2 4 6

-0.02

-0.01

0.01

0.02

Graphics

2 4 6

-1

-0.5

0.5

1

RF Phase →

RF

Vol

ts →

P E

rror

→

The ISIS Machine Cycle

0.005 0 0.005 0.01 0.015Times

0.2

0.3

0.4

0.5

0.6

0.7dlei

FT

Dipole Field vs Time

Extraction

Acceleration

Trapping

Injection

Bunch Size : Stable Regions

• Injected beam is effectively unbunched – Fills the machine circumference

• Phase Stable region shrinks as dB[t]/dt increases• Most particles trapped in stable regions

– Not all Trapping Loss!

• Have simplified processes– Complicated Trajectories– Space Charge 2 4 6

-0.02

-0.01

0.01

0.02

2 4 6

-0.02

-0.01

0.01

0.02

2 4 6

-0.02

-0.01

0.01

0.02

2 4 6

-0.02

-0.01

0.01

0.02

2 4 6

-0.02

-0.01

0.01

0.02

2 4 6

-0.02

-0.01

0.01

0.02

RF Phase →

P E

rror

→

Dual Harmonic RF Upgrade

• Add More RF Cavities– Run at 2 x frequency– Half the volts– Carefully Phased

• Enlarge Stable Region– Optimise Trapping– Reduce Loss -3 -2 -1 0 1 2 3

-2

-1

0

1

2

-3 -2 -1 1 2 3

-60000

-40000

-20000

20000

40000

60000

-3 -2 -1 0 1 2 3

-2

-1

0

1

2

-3 -2 -1 1 2 3

-40000

-20000

20000

40000

P E

rror

→

RF Phase →

Single Harmonic

RF Phase →

Dual Harmonic

New DHRF Cavities

6 h=2 cavities

2 h=4 cavities

2 h=4 cavities

Some Measurements of Longitudinal Motion

Injected Chopped Beam (100 ns, 1/15 of circumference)

RF Off RF On

Transverse Motion• Beam consists of many particles, not all aligned with reference orbit

– small angular spread about beam direction

• Need to focus beam to prevent it diverging– hitting vacuum vessel

• Quadrupole Magnet provides the focusing required

Beam with angular spread

Focusing Elements

Focused Beam

Unfocused Beam

Motion in a Quadrupole – Linear Focusing Field

BveF

v

BF

yxgveF

xygB

,

,

• Linear force with transverse displacement– Focusing in one plane– Defocusing in the other

• Arrange Alternating Channel– Overall Focusing in both planes

Form a Stable Focusing Channel

0 2.5 5 7.5 10 12.5 15

0

0.02

0.04

0.06

0.08

• Design repeating pattern of magnets for optimal stability (lattice)• "Modified" Simple Harmonic Motion

– distorted sinusoidal oscillation about equilibrium orbit

• Beam larger in focusing elements → overall focusing• ISIS Synchrotron Lattice

– Horizontal → [QD QF QD BF] x 10 (plus trim quads)– Vertical → [QF QD QF QD] x 10

Bea

m W

idth

→

Distance along beam axis →

Horizontal

Vertical

Lattice and Main Magnets

• 10 super-periods– 16.3 m long

• Main Dipole– 4.4 m long– 0.16 – 0.69 T

• Main Quads– Doublet & Singlet– 0.7 m long– 0.8 – 3.0 T/m

• Half Apertures– ~ 60 mm x 80 mm

• IAC = 400 A, IDC = 660 A• Power 1.8 MW• Also 20 Trim Quads

– Programmable

Particle Motion and Formation of a Beam

0 25 50 75 100 125 150Distance Around Machinem-0.15

-0.1

-0.05

0

0.05

0.1

0.15

esrevsnarT

tnemecalpsiD

mParticleTrajectories

0 25 50 75 100 125 150Distance Around Machinem-0.15

-0.1

-0.05

0

0.05

0.1

0.15

esrevsnarT

tnemecalpsiD

mParticle Trajectories

0 25 50 75 100 125 150Distance Around Machinem-0.15

-0.1

-0.05

0

0.05

0.1

0.15

esrevsnarT

tnemecalpsiD

mParticle Trajectories

0 25 50 75 100 125 150Distance Around Machinem-0.15

-0.1

-0.05

0

0.05

0.1

0.15

esrevsnarT

tnemecalpsiD

mParticle Trajectories

0 25 50 75 100 125 150Distance Around Machinem-0.15

-0.1

-0.05

0

0.05

0.1

0.15

esrevsnarT

tnemecalpsiD

mParticle Trajectories

-0.01 -0.005 0 0.005 0.01Transverse Displacement

-0.002

-0.001

0

0.001

0.002

elgnA

Oscillationin Phase Space

• Single Particle TrajectoryOscillation about equilibrium orbit

• Number of Oscillations per turn– Q Value– On ISIS QH=4.31, QV=3.83

• Avoid Integer Q– Stability

• Beam formed by– many particles (~1013)– incoherent oscillations

• Maximum Extent of beam– Beam Envelope

• Oscillation in (x,x') space

Tra

nsve

rse

Ang

le

Transverse Position

Beam Self-Field for Uniform Charge Distribution

0 0.05 0.1 0.15 0.2rm0

2000

4000

6000

8000

10000

rEVm

0 0.05 0.1 0.15 0.2rm0

210-6

410-6

610-6

810-6

0.00001

0.000012

BT

BvEeF

ca

reB

202

22

0

2

12

a

reBeveEF zrr

• High Intensity Beam – must allow for beam's own field

• Simplest case– long uniform cylinder of charge– moving at velocity βc

c

2

0

1

2 a

reEr

ar

• Gauss/Ampere Law• Defocusing Radial Force

– Strong energy dependence

• At low energy affects stability– Charge distribution is important– Take great care to control it

Relationship BetweenTransverse Particle Motion and Charge Distribution

12

34

56

-1-0.75

-0.5

-0.25

0.250.5

0.75

1

12

34

56

-1

-0.50.51

• Transverse Motion– Modified SHM

• Average Charge Density– Depends on amplitude distribution– Avoid small amplitudes

• Optimise Amplitude Distribution– Approach Uniform

• Injection Process Crucial– Determines initial distribution

particles with large amplitudes

-0.5 0 0.5 1

1

2

3

4

5

-0.5 0 0.5 1

1

2

3

4

5

particles with small amplitudes

Transverse Profile

Par

ticl

e D

ensi

ty →

Transverse Axis Transverse Axis

Transverse Axis Transverse Axis

Bea

m A

xis

→

Corresponding Trajectory

Injection – Simplest Model

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

Injection Septum Magnet

Injected Particles

Machine Circumference

Tra

nsve

rse

Dis

plac

emen

t

Simple Injection with a Bump ~ a few turns

Injection Septum Magnet

Injection Bump Magnets

• Use Magnet Bump– Equilibrium Orbit near Septum

during Injection (reasonable amplitudes)

– Remove bump after Injection so particles miss Septum

• OK for injecting a few turns• Not for the 100's required on ISIS

H- Charge Exchange Injection ~ many turns

• Inject an H- Beam

• Bring together H- with circulating beam in dipole

• Strip to H+ with foil (~2% loss H-, H0)

• Circulating beam Passes through foil (20 times)

• Inject on top of circulating beam over 100's of turns

• Allows HI Beam! Injection Bump Magnets

Stripping Foil

2 e- + H

H

H

Re-circulating H

Injection Painting

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

5 10 15 20 25

-1

-0.75

-0.5

-0.25

0.25

0.5

0.75

1

• To introduce a range of oscillation amplitudes during injection:

1. Vary Injection Point 2. Vary Equilibrium Orbit

• On ISIS do both1. Vertical Plane2. Horizontal Plane

• Inject current at constant rate– Vary amplitude non linearly

with time

1. Vary Injection Point

2. Vary Equilibrium Orbit

Injection on the Falling Magnet Field

• Inject before field minimum

• Field is 'too high'– Beam circulates on smaller orbit

• As field drops, orbit moves out– At 0.0 ms is on design orbit

• Inject at one point on inside radius– Movement of orbit gives painting

• Best configuration:– Painting process exploits B[t]

– Gives more time for trapping

Inject

Eqm Orbit Moves

Out

0.005 0 0.005 0.01 0.015Times

0.2

0.3

0.4

0.5

0.6

0.7

dleiFT

Dipole Field vs Time

Extraction

Acceleration

Trapping

Injection

Injection Details

• Accumulate 2.8x1013 protons over 150 turns

• Stripped with alumina foil – 0.3 μm thick

– 120 mm x 40 mm

– 98% efficient (2% H0, H-)

– Protons pass through foil ~20 times

• Injection Bump Magnets

– Single turn, 14 000 A

– Pulsed (45 mr)

• Septum– 6 turn, 4000 A

– DC (285 mr)

Extraction of Beam

• At Extraction– Protons Circulating at 800 MeV (=0.84)– 3.2 kJ per pulse– Two bunches with 200 ns gap

• Extraction System – 3 fast kicker magnets

deflect the beam into …

– a septum magnet

which lifts it into the EPB

• Kickers need to be fast to avoid beam loss– Go from zero to full field between passage

of bunches

Extraction Details

• Kickers– 3 units give 15 mr kick

– 0 to 0.04 T in < 210 ns

– Single turn, 5000 A

– 0.5 m long

• Septum – ~ 8 m downstream

– 8 Turn, 8900 A

– DC

– 1.8 m long (21 degrees)

– Lifts beam out of machine

Measurements of Transverse Motion

Transverse oscillation observed at one point in ring over 30 turns

Measurements of Transverse Motion

-150 -100 -50 0 50 100 150-1

0

1

2

3

4

5

6

7

8

9High Intensity Horizontal Scan: R5HPM1, 2.5E13 ppp

Position (mm)

Det

ecto

r S

igna

l (V

)

-100 -50 0 50 100

-400

-200

0

200

400

600

800

1000

1200

1400

High Intensity Horizontal Scan: R5HPM1, 2.5E13 ppp

Position (mm)

Tim

e (

us

)

-150 -100 -50 0 50 100 150-1

0

1

2

3

4

5

6

7

8

9High Intensity Vertical Scan: R6VPM1, 2.5E13 ppp

Position (mm)

Det

ecto

r S

igna

l (V

)

-100 -50 0 50 100

-400

-200

0

200

400

600

800

1000

1200

1400

High Intensity Vertical Scan: R6VPM1, 2.5E13 ppp

Position (mm)

Tim

e (

us

)

Development of Horizontal and Vertical Profiles -0.5 – 1.5 ms

Why is a High Intensity Machine Difficult to Run?

• Must Have Tight Loss Control on a High Power Beam– Mean Power: Injection 16 kW → Extraction 160 kW

– Fractional losses must be low

– Minimal Activation for Maintenance

– Prevent damage

• Losses on a High Intensity Beam are not easy to control!– Complicated loss mechanisms which are very sensitive to many parameters

– E.G. Beam Distributions, Correction Dipoles, Quadrupoles, RF, Linac, etc.

• Some important aspects of beam physics are not yet fully understood

Ring Tuning and Loss Control

• Loss control– Absolute levels

– Location

– Time ~ i.e. Particle Energy

• Monitoring– 40 Beam Loss Monitors

– Beam Toroids

– 5 Profile Monitors

– 30 Position Monitors

– Wire scanners, scintillators …

• Optimising Handles– Correction Dipoles (14)

– Correction Quadrupoles (20)

– Injection System

– RF System

– Extraction System

– Linac

• Many 100's of parameters

Diagnostics ~ Some Machine Details

So What’s New on a 20 year old Machine?

• Plenty of R&D Underway– Important for ISIS ~ to achieve optimal running with major upgrade – Important for New Machines

• Main Topics– Longitudinal Trapping i.e. DHRF upgrade and optimisation– Transverse Space Charge and Related Losses– Instabilities– Loss Control

• Involves– New Diagnostics & Measurements– Computer Simulation/Theory– Improved Beam Control and Manipulation (Software/Hardware)

• Paves the way for MW ISIS Upgrades

Closing Comment …

• First injected beam into the ring in 1984

– 20 years on ISIS is still a world leader

• Enormous credit to those who designed, built, commissioned machine

– Few things stay ahead for so long …

• Still upgrading, improving … to get to 0.24 MW

• Next steps 1 MW, 4 MW …