ppt-mall 2 - accelerator · ppt-mall 2 med linje Åke andersson, slri thailand, 15 august, 2016 1...

PPT-mall 2

Med linje

Åke Andersson, SLRI Thailand, 15 August, 2016 1

MAX IV; 3 GeV ring commissioning & RF systems

Åke AnderssonAccelerator development group, MAX IV

Outline

• MAX IV overview

• A little history

• MAX IV – 3 GeV & 1.5 GeV Rings & RF systems

• Cavities – conditioning

• 3 GeV ring commissioning

2Åke Andersson, SLRI Thailand, 15 August, 2016

Conceptual Basis of the MAX IV


• Scientific Case calls for high brightness radiation

over a wide spectral and time range: IR to Hard R-

rays, Short X-Ray Pulses.

• Need for high brightness: low emittance and

optimized insertion devices.

• This is hard to achieve in a single machine:

• higher electron beam energy harder photons

• lower electron beam energy softer photons

One size does not fit all !

The MAX IV Approach

• Different machines for different uses:• A high energy ring with ultra-low emittance for

hard X-ray users.

• A low emittance low energy ring for soft radiation users

• A LINAC based source for generating short pulses and allowing for future development of FEL source.

Åke Andersson, SLRI Thailand, 15 August, 2016

5

Inauguration was June 21, 2016


6

Photo Perry Nordeng

Aerial View of the MAX IV Site


7

Where is MAX IV Laboratory?


8

Inside the Linac building

Photo Annika Nyberg

Klystron gallery

Linear accelerator

Photo Annika Nyberg


Inside the 3 GeV building

9

The experimental hall with one of the beamline’s experimental hutch. Seven are already being built.

Ring tunnel, start ofcommissioning, September 2015.

Photo Annika Nyberg 140828


Inside the 3 GeV building

10

Photo Simon Leemann


Outline

• MAX IV overview






Overview ”MAX-IV Laboratory”

ESLS XIX, Aarhus University, November 23-24, 2011. Åke Andersson

Our new

laboratory

was named,

and the old

MAX-lab

became a

part of it.

Operation Schedule


User operaton 38 weeks/year.

Each week 6 days

5472 hours per year

3 rings 16 416 h per year

However, MAX-I, after 25 years of

user operation, is now retireing from

Synchrotron Radiation operation:

We are only occationally serving the

remaining IR interferometer beam

line and two educational beam lines

since autumn 2011. 9 user beam

lines were in operation over the

years.

MAX-I will continue running as a

pulse strecher for nuclear physics

as scheduled.

The daily operation scheeme


Injection routine


•SC Wigglers ramp

down: 6 min

•Undulator gaps

opening

•Bem dump (RF

off)

•Magnets zero: 0.5

min

•Magnets up to

injection energy

•Injection: 3 min

•Energy ramp: 2

min

•SC Wigglers ramp

up: 6 min

•Undulator gaps

closing

•Landau cavity

towards nominal

tuning, while

SCWs ramping.

No major benefit from

ramping down with

beam (~ 160 mA).

Statistics MAX-II

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829303132333435363738394041424344454647484950

Ava

ila

bil

ity [

%]

Week

Weekly availability MAX-II

2011

2010


Power break

during weekend

Corrector PS

trouble during

weekend

Statistics MAX-III

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829303132333435363738394041424344454647484950

Ava

ila

bil

ity [

%]

Week

Weekly availability MAX-III, 2011

First week after installing the

Landau cavity


Some ”dangerous” activities at the

lab

• MAX-IV Gun test stand

• MAX-IV Landau cavity test in MAX-III

• MAX-IV Cavity conditioning stand

Need some care not to interfere with the user

operation!



MAX-IV Landau cavity

Mechanical design, Elsayed Elafifi, MAX-lab

Conclusion from ”history section”:

Research & Development for future concepts and solutions, sometimes has a negative influence on present performance, but in the end it often pays off:


020406080

100120140160180200220240260

2010 2011 2012 2013 2014 2015

Me

an c

urr

en

t [m

A]

Year

Mean current

Blue: MAX II Red:MAX III

70

75

80

85

90

95

100

2010 2011 2012 2013 2014 2015

Ava

ilab

ility

[%

]

Year

Availability

Outline

• MAX IV overview






MAX IV 1.5 GeV ring

-15 -10 -5 0 5 10 15

X[m]

-30

-25

-20

-15

-10

-5

0

Y[m]

E = 1.5 GeV

C = 96 m

12 double-bend achromats;

length 4.5 m

12 straight sections;

length 3.5 m

Qx = 11.22 ; corr. chroma +2

Qy = 3.14 ; corr. chroma +2

εx = 6.0 nmrad

Radiation losses / turn: 117 keV


MAX IV 1.5 GeV ring


Ring RF 1.5 GeV

-15 -10 -5 0 5 10 15

X[m]

-30

-25

-20

-15

-10

-5

0

Y[m]

One of the twelve SS is

”sacrified” for RF


Ring RF 1.5 GeV ringLund Krakow

100 MHz

60 kW


2 2

MAX IV 3 GeV Storage Ring

-100 -50 0 50 100

X[m]

-160

-140

-120

-100

-80

-60

-40

-20

0

Y[m]

E = 3 GeV

C = 528 m

20 multi-bend achromats; length 21.4 m

20 straight sections; length 5 m

Qx = 42.20 ; corr. chroma +1

Qy = 14.28 ; corr. chroma +1

εx = 0.326 nmrad (bare lattice)

Radiation losses / turn: 360 keV

(bare lattice)

MAX IV 3 GeV ring

One of the 20 achromats in the 3 GeV ring

• Relatively compact magnet structure, except for two ”matching” short straights.


MAX IV 3 GeV ring• The Multi-Bend Achromat gives hor. emittance in the

Intra Beam Scattering regime:

Main radio frequency [MHz] 99.931

Harmonic number 176

Circulating current [mA] 500 Circumference [m] 528

Horizontal emittance (bare lattice) [nm rad] 0.37 [0.326]

Horizontal emittance (with 4 d w and 10 in-vac. Und.) [nm rad] 0.23 [0.201]

Radiation losses per turn (bare lattice) [keV] 360

Radiation losses per turn (with 4 d w and 10 in-vac. Und.) [keV] 854

Natural energy spread (bare lattice) [%] 0.084 [0.077]

Natural energy spread (with 4 d w and 10 in-vac. Und.) [%] 0.094 [0.091]

Momentum compaction factor 3.0 x 10e-4

Required lattice momentum acceptance ± 4.5 %

Rms bunch length with Landau cavities [mm] 50

Vertical emittance [pm rad] 8

[ ] =

without

IBS

Landau cavities are essential in order to reach the design

horizontal emittance!

The difference in

horizontal

emittance

with/without IBS is

kept low by diluting

the electron density

in the bunches.


Ring RF 3 GeV ring

One of the 20 achromats in the 3 GeV ring

• Relatively compact magnet structure, except for two ”matching” short straights.

None of the twenty SS is ”sacrified” for RF. Instead we will

use nine short matching straights upstream the SS.

ID light cannot

pass the cavity


Ring RF overview; 3 GeV Alternative I II

Energy loss with Ids 756keV 1020keV

Circulating current 0.5A 0.5A

Total beam power 378kW 510kW

Total RF voltage 1.5MV 1.8MV

Number of cavities 6 6

Cavity shunt impedance 3.2Mohm 3.2Mohm

Cu losses 117kW 169kW

Total RF power needed 495kW 679kW

Nr of RF stations 6 6

Nr of transmitters 12 12

Transmitter power 41.5kW 56kW

Power to cavity 83kW 113kW

Cu losses/cav 20kW 28kW

Coupling (beta) 4.2 4.0

Cavity voltage 250kV 300kV

Cavity gap 4cm 5cm

Bucket height 4.5 % 4.5 %

Alt I: Represents a solution for a 60% ID equipped ring, with the present MAX II/ MAX III cavities.

Alt II: Represents a solution for a fully ID equipped ring, with slightly modified MAX II/MAX III cavities.

60 kW60 kW

-φφ

120 kW Load

Circ.

Cavity

100 MHz


Chosen!

RF Rooms

Main Cavities

Harmonic Cavities

RF plants insidering

32

MAXIV Ring RF System

Energy 1.5 GeV 3.0 GeV

RF 99.931

MHz

99.931

MHz

Circumference 96 m 528 m

Harmonic

number

32 176

Current 500 mA 500 mA

No of cavities 2 6

RF station

power

60kW 120kW

Cavity voltage 280kV 300kV

Coupling

(beta)

2.3 4.0

Storage Rings Parameters

1 single 60 kW

transmitters

2 combined

60 kW

transmitters

Lars Malmgren, 19th ESLS RF Workshop Lund, 30 Sept. - 1 Oct., 2015

Ring RF System - 3 GeV Ring RF• The main cavities are

placed in the second short straight section of six consecutive achromats.

• Each RF-room contains two RF power plants.

33Lars Malmgren, 19th ESLS RF Workshop Lund, 30 Sept. - 1 Oct., 2015

Ring RF System - 1.5 GeV Ring RF

• Two Main Cavities and two Harmonic Cavities occupy one straight section

• Two 60 kW Power Plants are placed inside the ring.

34Lars Malmgren, 19th ESLS RF Workshop Lund, 30 Sept. - 1 Oct., 2015

Ring RF System – High Power Plants

• Contracts signed for– High power amplifiers (Electrosys, Italy).

The delivery time was delayed because of severe financial problems in the company. The risk was too large to continue so the contract was canceled (June 2014). A new contract has been signed (September 2014)for delivery of 60 kW liquid cooled solid state power amplifiers ( Rohde & Schwarz, Germany)

– Circulators (AFT, Germany)– Transmission Lines and Integration Work

(Exir Boadcasting AB, Sweden)

• Delivery of high power amplifiers: Two in December 2014 (January 2015) for test of circulators, two in February 2015 , two in March and finally two in June 2015 (1.5 GeV).

Lars Malmgren, 19th ESLS RF Workshop Lund, 30 Sept. - 1 Oct., 2015 35

Ring RF System – High Power Plants• Rohde & Schwartz 60 kW

CW solid state liquid cooled amplifiers based on two 30 kW transmitters/amplifiers with additional power combiner

• >64% overall powerefficiency

• High MTBF

• Compact: 2000 mm × 600 mm × 1100 mm (HxWxD)

• Coolant: glycol/water


Image: Electrosys

One pump unit andheat exchanger per rack

Outline

• MAX IV overview






ESLS-RF Trieste, September 29-30, 2010. Åke Andersson

Main cavity design

• MAX II & MAX III main cavity

Mechanical design:

Leif Thånell, MAX-lab (retired)


Main cavity design

• What we need to do better!• Cu became too soft after soldering

• An ”in air” weld of the shell (Ø 82 cm)

had leaks.

• Water cooling of the shell


Main cavity design

• What we need to do better!• Cu became too soft after soldering

• An ”in air” weld of the shell (Ø 82 cm)

had leaks.

Electron Beam Welding seems to be the solution, but we need to learn:

How stiff OFHC copper can we excpect to get for the end plates, from industry?

Rp0.2 of 180 MPa?

How much does an EBW soften the material around the weld?

Do we really need to stay in the elastic region when we tune the cavity?

Can we safely construct the shell out of two half shells?

For the final weld:

What is the weld shrinkage?

Do we get a decent inner RF contact at the weld stop?


Main cavity design

Cavity profile modification for 250 kV 300 kV

100 MHz Capacity-loaded Cavity for MAX-II and -III, F = 100.08317 MHz

C:\LANL\EXAMPLES\RADIOFREQUENCY\MYONEEXAMPLES\PILL100.AF 9-22-2009 12:40:42

0

5

10

15

20

25

30

35

40

0

5

10

15

20

25

30

35

40

0 5 10 15 20 25 30 35 40 45 50 55

5 cm gap instead of 4 cm

sligthly larger capacitor plate

We want to improve the cooling of

the plate.

New Old

Difficult to avoid water-to-

vacuum joints! OK, or not?

Q = 21000

Rsh = 3.5

Mohm

Q = 21000

Rsh = 3.5

Mohm

Ring RF System - Cavities

42

• All main cavities including two for Solaris Poland• Was Delivered

October -December 2013

July2013

April 2013

Photo: RI

Old spare cavity for MAX-II & MAX-II



43

• Ten Main (two for Solaris) and fiveHarmonic Cavities have beenconditioned at the old MAX-lab.

July2013

April 2013

Photo: RI



44

• Ten Main (two for Solaris) and five Harmonic Cavities havebeen conditioned at the old MAX-lab.

July2013

April 2013

Photo: RI

100 MHz Tetrode tubetransmitter for conditioning

300 MHz Harmonic Cavities

Main Cavities - Conditioning

45

• The cavities were delivered baked (3 days, 120 degree), with power coupler attached (β = 1).• A 600 l/s ion pump is attached. All cavities in the low 10-10 mbar range.

• So far, 9 main cavities (7 for MAX-IV, 2 for Solaris) have been conditioned to ̴ 25 kW.• Prototype: ̴ 1 year (!)• 2nd Cav # 11: ̴ 3.5 months• 3rd Cav # 08: ̴ 3 months• The following 5 cavities: ̴ 5 * 1 month (now a computer code was used! Robert Lindvall)• 9th cavity # 06: 2 weeks• 10th cavity #09: was only conditioned to ̴ 3 kW (lack of time)

• When all surrounding systems work OK, ̴ 3 weeks of conditioning is sufficient.• ̴1 week up to 50 W (!). Pressure raises up to 5*10-6 mbar!• ̴1 week to pass multipacting regime 3-5 kW. Sometimes a need to attach a turbo!• Finally ̴1 week to reach 25 kW stable operation, without more than 1 ”glitch” per day.• ”Glitch” = Sudden high reflected power, however self extinguishing after ̴ 60 μs.



46

• Multipacting problem origin: Coupler or Cavity body?


Main Cavity – Coupler loop


48

Achromat # 16 17 18 19 20 1

Resonant freq. N2-Vented & Force free [MHz] 100,112 100,019 99,93 100,13 99,973 100,042

Difference compared to FAT [MHz] -0,084 -0,001 0,014 -0,043 0,038

Unloaded Q 20500 20400 20400 20250 20450 19700 Theory cyl-symm: 20923

Degradation due to Ports & Surfaces [%] 2,1 2,5 2,5 3,2 2,3 5,8

Shunt Impedance (linac def.) [MΩ] 3,45 3,43 3,43 3,41 3,44 3,32 Theory cyl-symm: 3,52 MΩ

Required power to reach 300 kV [kW] 26,1 26,2 26,2 26,4 26,2 27,1

After conditioning we vented and turned coupler to β = 2 for installation.We then measured fr and Q0 carefully (by turning coupler to β = 0) :


The two cavities for the 1.5 GeV ring: Unloaded Q were 20100 and 19300.(The last Q-value was surprisingly low, indicating a 7.6 % surface degradation.)



49

A tiny defect in the ceramic windowcaused a leak p 1̴*10-8 mbar



50

Three probe loop ceramics (out of 16) have started leaking.Only those we forgot to 50 Ω terminate! Heating problem?

The last halfyear, moreleaks haveappeared, eventhoughterminatedproperly!!!


Harmonic Cavities - Conditioning

51

• The 7 series cavities (5 MAX-IV, 2 Solaris) were delivered non-baked, only leak tested.• We performed ourselves the bake-out, with an Århus-coupler at β = 1 attached.• Each cavity has two 100 l/s ion pumps. All cavities in the low 10-10 mbar range.

• So far, 5 harmonic cavities have been conditioned to ̴ 4 kW.• Prototype: Is situated in the MAX-III ring since ̴ 4 years. Used only at ̴ 0.5 kW.• The following 5 cavities: ̴ 5 * 2 weeks (manual conditioning from a 300 MHz transm. )

• ̴1 week up to 50 W. Pressure raises up to 5*10-7 mbar!• ̴1 week to pass multipacting regime 0.5-2 kW.• 4 kW without problems, and without ”glitches”.• ”Glitch” = Sudden high reflected power, however self extinguishing.


Harmonic Cavities - Conditioning

52

Achromat # 13 14 15

Resonant freq. @ FAT [MHz] 299,89 299,749 299,575

Resonant freq. Pumped & Force free [MHz] 299,766 299,561 299,44

Unloaded Q 20800 20800 21000 Theory cyl-symm: 21656

Degradation due to Ports & Surfaces [%] 3,95 3,95 3,03

Shunt Impedance (linac def.) [MΩ] 5,32 5,32 5,37 Theory cyl-symm: 5,54 MΩ

After bake-out, conditioning, removal of coupler, and installationwe measured fr and Q0. A Δfr = -140 kHz is expected.


Main Cav. – Transport to site


Installed and baked in 3 GeV ring

Photo courtesy S. C. Leemann

Outline

• MAX IV overview






Outline

• MAX IV overview




• 3 GeV ring commissioningFirst a few glances at magnet and vacuum techniqal solutions:


Slide by Martin Johansson

MAX IV 3 GeV ring DC magnets• Each cell is realized as one mechanical unit containing all magnet elements. •Each unit consists of a bottom and a top yoke half, machined out of one solid iron block, 2.3-3.4 m long.

MAX IV 3 GeV ring magnets

Strong, 25 mm bore, sextupoles & achromatic octupoles for non-linear optics.All those carry auxiliary windingsthat can be used as:• Additional H/V correctors• Auxiliary sextupoles• Skew quadrupoles (coupling &

vertical dispersion correction)• Upright quadrupoles (calibrate

BPMs to the center of adjacentsextupole/octupole)


MAX IV 3 GeV ring vacuum

• First presentation


MAX IV 3 GeV ring vacuumBPM

Ion pump locationAbsorber location

Sector valve location

VC10VC1

VC2VC3

VC4

VC5

VC6

VC7

VC8

VC9


3 GeV ring, commissioning timeline

Commissioning 3 GeV Ring - RF

62

• During 2015 we had different difficulties to maintain all five cavities at nominal fields. ( The sixth cavity was removed from the ring because an early RF incident, that lead to a broken power coupler window). Mainlyvacuum trips even at low power was common. We would have neededeven longer conditioning time to remove multipacting regimes. Due to lack of time it was then easier to leave the power station off.

• However, for commissioning it was always enough with two or threecavities in operation. Injection was even more efficient at low total cavityvoltage, and 120 mA could easily be stored with three cavities in operation.

• The amplitude and phase loops were always regulating on the forward fields, from a directional coupler just before the power coupler. We found it not neccessary yet to regulate on the cavity fields.

• The frequency loops worked well. Still problems with the mechanical part, until it was found out that the bearings were not sufficiently greased.

• The frequency loop needs a delay of some hundred ms, to avoid regulatingon a mechanical vibration when the motor is moving. Alternatively an averaging of the signal over longer time can be used (low-pass filter).

3 GeV ring, BPM offsets

-1.0

-0.5

0.0

0.5

1.0

Offset [m

m]

500400300200100

S[m]

Horizontal Vertical


3 GeV ring, integer tunes

-1.0

-0.5

0.0

0.5

1.0

DY

[mm

]

500400300200100

S[m]

Model Experiment

M1-COAY-01 changed by 0.1 mrad

-2

-1

0

1

2

DX

[m

m]

500400300200100

S[m]

Model Experiment

M1-COAX-01 changed by 0.1 mrad


3 GeV ring, bare vertical orbit

Plot by M. Sjöström

A random seed from our misalignmentmodel. Internal note 201211071, by S.C. Leemann

3 GeV ring, vertical aperture

2011, simulations by S.C. Leemann2016-02-03, measurements by J. Sundberg

Measured vertical aperture scaled to center of LS: 2.3 mm


3 GeV ring, vacuum conditioning

5

6

7

8

910

-11

2

3

4

5

6

7

8

910

-10

Ave

rag

e N

orm

aliz

ed

Pre

ssu

re [

mb

ar/

mA

]

12 3 4 5 6 7 8 9

102 3

Dose[A.h]

P=P0*D^(power)power= -0.88 ± 0.004P0 = 1.78e-010 ± 1.28e-012

All vacuum gauges in S2except RF cavitiesI=50 to 55 mA

Commissioning – March 2016

68

• During 2016 we had a longer shutdown in February for installation of the two first IDs.

• From March to June the commissioning time was split between machine studies and BL commissioning at lowcurrents.

• During March to June 2016, we mostly operated with 4 or 5 main cavities at 270 kV each.

• In May –June, the phase and amplitude loops weremade to regulate on the cavity fields without problems.

• The machine studies mainly concentrated along twopaths: 1) Lattice/Optics investigations (low current), and 2) Instability studies with/without Harmonic Cavities(current up to 160 mA).


3 GeV ring, collective multibunch effects

• Possible to store >120 mA without feedback and without harmonic cavities. Predicted RW thresholdwas only ~ 40 mA !

• HOM driven longitudinal motion is evident at a fewmA in uniform fill.

• Harmonic Cavities not fully tested yet. Optimal settings at 150 mA and above.

• Preliminary BBB feedback tests using a short stripline showed a longitudinally stable beam up to 35 mA.

• Longer striplines for BBB feedback were installed in february.

• Longitudinal Actuator (cavity) under design.Åke Andersson, SLRI Thailand, 15 August, 2016

3 GeV ring, Vacuum lifetime, May 2016

The beam was not stable longitudinally, nor vertically. Touschek lifetime fairly largeGas lifetime at 140 mA must be larger than 30 h!


Thank You for your attention!


Extra slides

72

Digital Low Level RF


• The DLLRF is based on the Perseus FPGA platform from Nutaq. Two units is in operation in the 3 GeV ring controlling two cavities each. The third will be taken into operation soon.

• It is possible to implement two independent loops besides the tuning loop. One controlling the amplitude of the cavity field and one the phase of the forward power. Either I / Q or polar loops can be selected.

• It has a fast data logger for post-mortem analysis.

Design by Angela SolomGUI by Antonio Milan


Chopper for Ring Injection • Has two identical vertical kickers.• The kickers consist of a 15 cm long stripline pair

with a characteristic impedance of 50 Ω for odd TEM modes.

• Both electrodes are fed by RF• An aperture is located downstream. The

unwanted bunches will be dumped here. • The aperture can be selected so the wanted

bunches either passes a 1 mm iris, a 2 mm iris, or over an edge.

Aperture

0.84m0.24m

If φ1=-φ2 → Zo=49.9ΩIf φ1= 0 → Zo=63.8ΩIf φ1= φ2 → Zo=88.2Ω

2 D design

333 ps

10 ns


Kicker system for ring injection

50 Ω

300 MHz~1 kW

100 MHz~400 W

Combining network

700 MHz~300 W

50 Ω

300 MHz~400 W

50 Ω

100 MHz~1 kW

Crosstalk <-31db

333 ps

10 ns

The MAX IV thermionic pre-injector will be covered by the talk of David Olsson


MAX IV LinacThe linac should be used as an injector for both the 1.5 and 3 GeV storage rings and the SPF (Short Pulse Facility)• 18 klystrons• 18 SLEDS• 39 linac structures

Operating frequency 2998.5 MHzMaximum rep. rate 100HzMaximum RF power 35 MWRF pulse length 4.5µsLinac length 250 m

• Two Electron sources1. One klystron (7.5MW) feeding a

thermionic RF gun used for ring injections

2. A photo cathode gun for the SPF fed from the first linac klystron

Operating beam energy 3 GeV

Max. on-crest beam energy 3.6 GeV

44% RF power redundancy. Part of this has been reduced due to arcing at

some of the RF Power Units at high power. For safe operation two RF units

will be added later in 2016-2017. The linac tunnel is prepared for this change.


MAX IV linac

Photo Perry Nordeng 130903

K00+K00TG K01 K02K19

• RF conditioning did take longer time than anticipated despite that everything except the waveguides is preconditioned by RI. Problems with the subsystems have limited the time for conditioning

• Only minor impact on the Linac commissioning time schedule. The personal safety system PSS was changed so that it is possible to accelerate electrons up to the first bunch compressor while RF conditioning could continue in the rest of the linac.

• 3 GeV was reached for the first time February 9, 2015

Two more RF units willbe added for safeoperation, August 2016

ppt-mall 2 - accelerator · ppt-mall 2 med linje Åke andersson, slri thailand, 15 august, 2016 1...

Documents