Peter McIntoshSTFC and Cockcroft Institute

CI SAC Review1 – 2 November 2010

SRF and RF Activities

2

Outline

• International Cryomodule• COOL-IT Cryogenics• EMMA RF System• Digital LLRF• Working with UK Industry• Conclusions

3

International ERL CM Collaboration

• International collaboration initiated in early 2006:– ASTeC (STFC)– Cornell University– DESY– FZD-Rossendorf– LBNL– Stanford University– TRIUMF

• Fabricate new cryomodule and validate with beam.

• Dimensioned to fit on ALICE:– Same CM footprint– Same cryo/RF interconnects

• Scheduled for installation in April 2011

Target Cryomodule Specification

4

SRF Cavity Design• 2 x 7-cell superstructure

cavities provided by DESY.• End groups re-designed by

LBNL, ASTeC and Cornell:– large b-p HOM absorbers,– larger variable coupler.

• Modifications performed and validated by Cornell.

5

Cavity Fabrication and Qualification

1.0E+09

1.0E+10

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0

Qo

Eacc (MV/m)

Before He vessel weldingAfter He vessel welding

ALICE OperationalTarget

Design Target

1.0E+09

1.0E+10

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0

Qo

Eacc (MV/m)

Before He vessel welding

After He vessel weldingDesign Target

ALICE OperationalTarget

Cavity #1

Cavity #2

6

Cavity Processing

• Cavity #1– 21/8/09

• 15µm BCP, HPR and 48hr 110C bake

– 25/9/09• 25µm BCP and HPR

– 8/3/10• 340µm BCP, HPR,

5µm micro-BCP and 48hr 115C bake

– 6/5/10• He vessel weld and

HPR

• Cavity #2– 6/12/09

• 80µm BCP and HPR– 21/1/10

• 80µm BCP, HPR and 15-20µm micro-BCP

– 12/3/10• 48hr bake at 115C

– 26/3/10• 320µm BCP and HPR

– 15/7/10• He vessel weld and

HPR

7

Input Coupler Design• Utilised Cornell ERL injector

coupler as original design.• Cold section of the Cornell

injector coupler too long to load into the cryomodule.

• Removed 80 K intercept ring and two bellows convolutions.

• Reduced the 2 K to 5 K transition tube.

• Shortened the coupler cold section by 15 mm and modified 80K skeleton to allow cavity string insertion.

Original

Modified80K Intercept

removed

2K5K

80K 300K

Cold section shortened by

15mm

Original

Modified80K Intercept

removed

2K5K

80K 300K

Cold section shortened by

15mm

8

Input Coupler Conditioning

Vac 1

Vac 2

Vac 3

9

HOM Absorbers• Cornell ERL injector HOM absorber

utilised for high current operation (up to 100mA).

• Experience at Cornell however has identified that TT2 ferrite bulk resistivity increases considerably at T<80K, resulting in significant charge build-up. • Net effect is to

deflect and distort beam at low energy.

• Modification to remove TT2 ferrite tiles and rely solely on ceramic tiles to damp HOM power.

Beamlet distortion throughCornell injector cryomodule

10

TT2 Tile Removal & Cold Tests

11

Tuner Development

• Employed a modified Saclay-II tuner assembly:– Wider aperture– Low voltage piezo cartridges

Saclay-II

Modified Saclay-II

• Dual cams precision aligned and pinned.

12

Cavity String Assembly

13

Outer Cryomodule Assembly

14

Cavity String Integration

15

Final Cryomodule Configuration

3 Layers of Magnetic Shields

16

System to COOL to Intermediate Temperatures

• The radiation shields and other thermal intercepts in the existing ALICE SRF Cryomodule are cooled with LN2.

• The boiling of liquid nitrogen generates microphonics which detune and destabilise the operating frequency.

• The new cryomodule will use Cold Ghe for cooling radiation shields and other thermal intercepts to avoid microphonic generation as there is no boiling.

• COOL-IT, developed indigenously by ASTeC, provides the necessary cooling power at intermediate temperatures at 80K and 5K .

COOL-IT

New Cryomodule

Existing Cryo-system for ALICE

17Concept Design Build

Concept to Commissioning

18

• Instrumentation & Controls developed in-house• Manufactured with the help Local UK industry • Off-line tests at the factory conducted successfully

COOL-IT Verification

19

TCF 502K BOX 1500 L Dewar LINACBOOSTER

COOL-IT

HEX

COOL-IT

Transfer Lines

COOL-IT Integration on ALICE

Ready for Commissioning....

20

EMMA – World’s 1st NS-FFAG

Injection Line

Diagnostics Beamline

Frequency (nominal)

1.3 GHz

No of RF cavities

19

Repetition rate 1 - 20 Hz

Bunch charge 16-32 pC single bunch

Energy range 10 – 20 MeV

Lattice F/D Doublet

Circumference 16.57 m

No of cells 42

Normalised transverse acceptance

3π mm-rad

21

EMMA Complete – Aug 2010

22

RF System Requirements

• Voltage:– 20 - 120 kV/cavity essential for serpentine

acceleration, based on 19 cavities– Upgrade possible to 180k

• Frequency:– 1.3 GHz, compact and matches the ALICE RF

system– Range requirement 5.6 MHz

• Cavity phase:– Remote and individual control of the cavity

phases is essential – 19 waveguide phase shifters

23

The EMMA RF SystemRF Cavities

Waveguide Distribution System

High Power RF Amplifier System

Libera LLRF System

24

EMMA LLRF System• Instrumentation

Technologies Libera LLRF system provides:– Initial cavity setting

conditions

– Precise control of the cavity amplitude and phase to ensure stable acceleration

• Diagnostic monitoring:– Cavity pick-up loops

– Forward and reverse power monitoring to each cavity

– IOT power levels before and after the circulator

• Novel synchronisation of the accelerators:– A 200µs beam pre-trigger

used to reset LLRF phase accumulators every beam pulse.

– The LLRF synchronises itself on every trigger pulse, preserves the relationship between ALICE 1.3 GHz and EMMA offset frequency.

25

First High Power Commissioning• Excellent cavity control stability

(up to 40 kW so far):– 0.007% rms voltage– 0.027o phase

• Many issues with tuner and phase shifter motors, communication errors, slipping motor shafts etc.

• Ability to ‘ignore’ bad cavities from the GVS.

• Further work planned during shutdown to understand and fix all motor problems.

• Libera LLRF will then take full system control with updated control software.

17/08/2010

26

EMMA Frequency Tuning

• Changing cavity frequency takes – 30 minutes.

• Currently using EPICS system to move motors in open loop.

• Using centre frequency and bandwidth controls ‘sweep analysis’ locates resonance of each cavity in system.

• A new centre frequency can then be set and the tuner motors driven.

• Calc detune shows new resonance of cavity.

• Low reflected power response used to fine tune each cavity.

27

EMMA Synchronisation• Yellow = ALICE 1300GHz• Blue = EMMA 1301Ghz• Trigger on laser pre

injection pulse.• Measured on 12GHz

scope while varying GVS phase and amplitude:– Can see that the global

phase of EMMA being moved while maintaining lock during this simple test.

– Beam based analysis of the synchronisation will be demonstrated soon.

28

Beam with RF• RF cavity phase set to 154

degrees separation (ToF).• Result on RF voltage on beam

is half of expected change in ToF:– Cavity phases not set

optimally.• RF buckets around transition

momentum still separated – not enough voltage for serpentine acceleration.

• Seen RF bucket & synchrotron oscillations inside it.

• Next step adjust each cavity phase separately use beam as diagnostic.

Synchrotron OscillationsObserved

29

Beam Commissioning• Zero cross of each cavity to

find optimum phase angle.• During recent experiment

beam loading effects could be seen on Libera LLRF.

• Possibility to zero cross each cavity, tune for max acceleration - needs testing.

• Close loop on Libera system and find the correct phase – phase accumulator reset during sweep.

• RF acceleration essential goal before shutdown in Nov - Dec 2010.

30

EMMA Milestones

Project start Apr 2007Design phase Apr 2007 – Oct 2008Major procurement contracts May 2007 – Aug 2009Off line build of modules Oct 2008 – 15th Jun 2010Installation in Accelerator Hall Mar 2009 - Sep 2009Test systems in Accelerator Hall Jul - Oct 20091st Beam down the Injection line 26th Mar 20101st Beam through 4 sectors 22nd Jun 20101st Circulating beam in EMMA 16th Aug 20101st Accelerated beam in EMMA Sep/Oct 2010

(Underway)ALICE & EMMA shutdown Nov – Dec 2010EMMA Experiments Jan 2010 – Mar 2011Basic Technology Grant complete Mar 2011

31

Digital LLRF Developments• The LLRF4 board (developed

by Larry Doolittle at LBNL) is used as the basis for the design.

• FPGA software has been written using VHDL, Matlab and simulink.

• Supervision and control of the system is performed by a Labview VI, which also implements adaptive feed forward for beam-loading compensation.

• Labview system interfaces with the ALICE EPICS control system.

• System developed and implemented in < 18 months!

32

Operational Performance• The Digital LLRF system has

been operated on the ALICE NC buncher cavity for a period of 2 weeks.

• The system was set up and locked within 10 minutes:– Short term stability better than the

existing analogue system (better than 0.04 degrees rms phase error)

– Long term stability is limited by temperature drifts within the analogue front end

– Some non linear behaviour has been observed

• The Adaptive feedforward system has been operated with beam.

• Found to be an effective way of reducing beam loading effects.

Feed Forward Table

33

Digital Piezo Tuner Control• SRF cavities experience strong detuning during pulsed operation

due to Lorenz forces, causing the cavity to dynamically detune during the pulse.

• Detuning causes the LLRF system to work harder and increases RMS phase and amplitude errors.

• Piezo tuners can be used to dynamically apply tuning forces to the cavity within a pulse. Hence reducing the detuning problem

• A National Instruments controller has been sourced which provides :– 16 Bit I/O channels running at 100kHz– An FPGA on which control loops will run

• Programming is done within the Labview environment, reducing development costs.

• Development is at an early stage.• Hardware specified / designed• Hardware purchased• Programming has not yet begun

34

Working with UK Industry• STFC Innovations (Mini-IPS)

funded activity with Shakespeare Engineering to fabricate a single-cell 1.3 GHz validation structure.

• Fabricated by Shakespeare and using STFC SRF processing and testing facilities to validate:– ISO 4/5/6 Cleanrooms– BCP chemical polishing– HPR rinsing– Vertical testing

• Validation by end Oct 2010.• First bulk Nb SRF cavity to

have ever been fabricated in the UK.

• Full IPS proposal now issued to build a 9-cell SRF cavity.

35

Fabrication at Shakespeare

36

Conclusions• All hardware for our international cryomodule is now

available:– Full integration and assembly preparations are underway.

• The new COOL-IT cryo-system has been verified and installed and is ready for commissioning when new cryomodule is installed.

• EMMA RF system has been successfully demonstrated – eagerly awaiting acceleration verification.

• Made outstanding progress in developing and implementing digital LLRF solutions:– other applications looming – MICE RF @ 200 MHz

• Hope to be able to nurture UK industry to demonstrate complete SRF cavity fabrication capability within the next 3 years.