Review of Beam Instrumentation in CTF3

Review of beam instrumentation in CTF3 A. Dabrowski, 06/05/2010

Review of Beam Instrumentation in CTF3

Anne DabrowskiCERN BE/BI

On behalf of all involved


Outline• Overview of Instrumentation

BPMs ; Transverse Profile ; Longitudinal Profile ; Bunch Frequency Measurements

• Instrumentation activity in 2009

New Installations

Maintenance and improvements

Calibrations and Beam Based Measurements

• Instrumentation activity foreseen for 2010

Design

New Installations

Upgrades


Overview of Instrumentation installed (I)

Beam Position and Intensity Monitors

2 BPE’s + 46 BPM’s + 54 BPI’s + 6 Re-Cavity BPM (califes) + 16 BPS + 5 WCM’s

Juanjo Garrigós et al .Valencia Univ.

(based on M. Gasior BPM, scaled for reduced aperture – 24 mm)16 unitsresolution ~30um

Inductive Pickup (BPS) - TBL CTF3 has final 129 position / intensity monitors!

Main activities 2009

Final BPM vacuum installations—Installation of 16 BPS with aligned support - Univ Valencia, Built and tested at IFIC. —BPS Amplifiers: Univ. Politècnica de Catalunya (UPC)

Radiation damage to LAPP digitizing acquisition for TL2 / CLEX installed in the machine Preparation of “Linac / Combiner Ring type” acquisition system and calibration software for the BPMs in TL2 / CLEX


CLEX

DELAY LOOPCOMBINER

RING

LINAC

Overview of Instrumentation Installed (II)

TL2

Transverse Profile monitors

14 TV stations for OTR based emittance measurements

7 TV stations for OTR based spectrometry (energy)

7 TV stations for Synchrotron light (Chicane and Rings)


Cleaning Chicane Cleaning Chicane

Stretching Chicane Stretching Chicane

Injector, 1.5 GHz Bunch Spacing Injector, 1.5 GHz Bunch Spacing

Compression chicane - TL2Compression chicane - TL2

Combine Beam, 12 GHz Bunch Spacing Combine Beam, 12 GHz Bunch Spacing

• Linac ~ 1-7 ps

• Delay Loop and Combiner Ring > 8ps

• CLEX 1-2 ps

• Probe Beam (Califes) < 2 ps

σ= 4.5ps (1.4 mm)

σ= 8.9ps (2.7 mm)

Streak LabsDL, CR (now)CLEX (2010)

Overview of Instrumentation Installed (III)Longitudinal Bunch Profile

RF Deflecting Cavity, OTRStreak Camera Synch Light, or OTR

Bunch Length Form Factor (r.m.s.)BPR-W (Power @ 30 GHz)RF-pickup (Power 30 – 170 GHz)CDR Experiment

Bunch Combination EfficiencyBPR-S (Down-mixed 3 GHz)Phase monitor DLPhase monitor CRStreak Camera Synch Light, or OTR





New Installations




Design

New Installations

Upgrades


Instrumentation activity in 2009

New Installations 7 Transverse Profile Monitors

• TL2 emittance tank

• TL2’ emittance tank

• TBL emittance tank

• TBTs drive beam (Uppsala / Saclay mechanics & planning – CERN acquisition & control)

• TBTs probe beam (Uppsala / Saclay mechanics & planning – CERN acquisition & control)

• TBL final spectrometer line

• PHIN Virtual Cathode acquisition & control

PHIN- 2 Gated (5ns ; 100ns) Intensified Camera

- 4.8 degrees to the spectral reflection for OTR based profile measurements

- 20 Channel Segmented Dump for Time Resolved Profile Measurements designed, built and commissioned

Maintenance and Improvements Replacement of damaged screens Planning for “Linac / Combiner Ring type” BPM acquisition hardware TL2/CLEX BPM Calibration software development for TL2/CLEX

Activity 2009


Streak

Camera

–

Bunch

Length

Measurement

Activity 2009 – Beam Based Measurements

• Bunch Shape– A Skew Gaussian bunch shape

• Measure calibration factors– Vary Time Streak Trigger and calculate the

corresponding position of the peak of the bunch– 0.122 ± 0.004 ps/pixel (2 sigma) for 10ps/mm

• Measurement of the jitter in peak– eg 5.5 ps ± 0.2 (2 sigma)– Contribution from trigger and beam

• Slit size contribution to measurement - FWHM in focus 14.8±0.9 pixels (2 sigma)- Corresponds to ~350 μm

Propagate all error contributions …

Typical measurement error on FWHM is 4% (2 sigma)

Review of beam instrumentation in CTF3 A. Dabrowski, 06/05/2010Activity 2009 – Beam Based Measurements

Bunch Length measurement along pulse train Streak Camera in Combiner Ring using MTV 0496 (zero dispersion point) 3 GHz uncombined beam, by-passing the delay loop 50 ns sampling 2 sigma error bars

Use Streak Camera measurement and this bunch length variation to cross calibrate other bunch length instruments


BPR with WR-28 waveguide port

•Power measurement at (30±4) GHz 3 db band poss

•For given beam current & position:

Maximise the signal Minimise bunch length

Relative Bunch Length measurement – BPRW

shape)bunch Gaussian (example2

22

22

2c

o

t

eqr

rP


Calibration of CR.BPRW0505W with Streak Data


- Data used: 04-12-2010- Beam conditions: 3 GHz 4 Amp beam- Use time resolved bunch length from Streak - Assume quadratic function for fit

222

21

0

22AmV2

2AmV10ps

/mV)(A ps 0.06p

/mVA ps 2.04p

ps 23.22p

))/(BPM(BPRW*p)/(BPMBPRWppσ

BPR and Streak in good agreement

Exercise should be repeated with different beams to study systematics and verify current and position normalization

Error in calibration large (40% error 2 sigma)

Measurement of BPM, BPR and Streak relevant for a good calibration

BPR’s used as Online bunch length measurement available today!

Application of Calibration:


Similarly Compare RF pickup (waveguide down mixing) to Streak

Power measurement in time domain

Correspond power (33 GHz) to bunch length Streak

Corresponding Frequency domain

Apply Band pass filter to isolate beam signal

Good agreement between RF pickup and Streak in the Steady state part of the pulse

Calibrated RF-pickup 33 GHz form factor



Bunch Combination Efficiency Measurements

odd buckets

even buckets

Delay Loop

RF deflector

Combination scheme

λo =20cm

Bunch combination efficiency

Streak Camera using Synch Light

BPR (3 GHz Beam signal mixed with ref. 3 GHz signal)

Phase monitorDL: Harmonics of 1.5 GHz and 3.0 GHzCR: Harmonics of 3 GHz [6, 9 12, 15 GHz]

CCD

Streak

Gated

Combiner Ring Multiplication

λo =10cm


Streak Camera measurements of 18th November 2009

Bunch spacing between Delay Loop and Combiner Ring:

Measurement conditions: 50 ps/mm, blue filter in

Turn DL late by

1 12.99 ± 3.32 ps

2 25.33 ± 3.44 ps

3 32.43 ± 3.50 ps

4 16.89 ± 3.36 ps

Example

Aurélie Rabiller


Online monitoring bunch spacing with the Phase Monitor


Frequencies

Delay loop: 7.5, 9.0, 10.5 & 12 GHzCombiner ring: 6.0, 9.0, 12.0 & 15 GHz

Power induced by the passing beam, picked up by 4 antennae


Measuring the bunch spacing with Phase Monitor (CR)


Simulation

beam with uniform current

15 ps FWHM Gaussian Bunch length uniform along the pulse

Turn 1, shows effect of bunch length

Data compared to simulation

3 GHz uncombined beam for hardware test (04-12-2010)

Raw signal corrected for electronic gains

Simulation includes the bunch length dependence along the pulse measured with the BPRW & current variation using BPM

Data compared to simulation, shows strong correlation

Data 3 GHz

Simulation 3 GHz + bunch length (t) + current (t)


Monitor bunch spacing with the Phase Monitor


Example of measurement 18 Nov 2009, Factor x8 combination current

DL ~ 20 ps late Bunch length variation along the pulse

train

Measurement difficult to interpret because of bunch length and bunch spacing variations

The way forward?

Instrumentation:

Use BPRW in CT line to measure bunch length along pulse train & normalize

Produce feedback for the operators

Re-design phase monitor to work at lower frequencies – less sensitive to bunch length variations ?? (see extra slides for potential upgrade Steve Smith)

Raw data CR


Spectrometery

Design

work

2009

• Extensive Fluka and GEANT4 studies on 5-150 MeV electrons interacting with matter

• Necessary for optimizing detector resolution, thermal effects, material choices and radiation damage for time resolved spectrometry

• Transverse and longitudinal shower energy deposition

Excellent FLUKA /

GEANT4 agreement


Segmented

dump for

PHIN commissioned

• Good correlation with Energy measured in the dump and RF

• CTF3 note 099

• Tuning of the RF beam loading with segmented dump to get a uniform steady state beam pulse

Beam loading compensation





New Installations




Design

New Installations

Upgrades


Bunch

Length

Measurement in

CLEX -

2010

- Bunch Length Measurement in CLEX

Activity 2009-2010

Cleaning Chicane Cleaning Chicane

Stretching Chicane Stretching Chicane


Acceleration 3.0 GHz Acceleration 3.0 GHz

Compression chicane - TL2Compression chicane - TL2

Combine Beam, 12 GHz Bunch Spacing Combine Beam, 12 GHz Bunch Spacing

Streak LabsDL, CR (now)CLEX (2010)

Design & Installation foreseen (2010)

Long Optical lines and New Streak Lab for Streak Camera measurement in CLEX FESCA200 Streak Camera (300 fs resolution) arrive in July 2010 2 Streak Cameras aid beam studies in two machine locations at the same time

Non-destructive high frequency RF based bunch length monitoring measurement Waveguides and diode components ordered (CTF3 & NWU) Assembly & commissioning for 2010


Ti

me resolved

energy

measurement

TBL

Activity 2009-2010

ΔE ~10-60%

High intensity (28 A) e⁻ beam of 150 MeV

High energy transient

Nominal pulse length: 140 ns

New detector design

•Design:32 channel transverse faraday cup10° spectrometer line includes full FLUKA simulation realistic beam profile from PLACET3 mm segments, 400 μm slits collimatorSingle shot measurement of the steady state

•Implement in Machine TBL Fall 2010

5% measurement on energy spread


New

BPM

BI

Calibration

Software for

TL2 and

CLEX

• Java user interface in development to perform the calibration for BPMs in TL2 and CLEX• Software:

– Communicate to calibration controller crate– Calibration of all BPMs for 2 different gain ranges and attenuation ON/OFF settings– Save calibration coefficients to an archive file

• Operators can load those calibration coefficients into database for use with the XenericSampler applications

• Dedicated beam time for June startup needed for BPM calibration– Machine should be a “nominal” setting, with only the start beam trigger disabled, for calibration

132 devices with raw data: 44 pickups, 3 devices per each pickup: H/V plane + sigma

132 devices with raw data: 44 pickups, 3 devices per each pickup: H/V plane + sigma


BPM TL2/CLEX Calibration expert GUI


BLM general layout

400 x 400 SPAD array

Active surface 1mm2

Recovery time ca. 4 ns

1 photon detection

CMOS technology

Low cost detector

Quantum efficiency 20% in blue range

Immunity to external magnetic fields

Beam Loss Monitor for TBL: a general layout

Cerenkov cone

α

β

electronFiber core

Cerenkov photon

Physical principle

Silicon Photomultiplier

Angela Intermite


Beam Loss Monitor for TBL: SiPM Dark Noise Characterization

Goal: Identification of best sensor for detection of Cerenkov Light generated by particle losses.

Investigation into:• Dark noise• Optical cross talk• Dynamic range of noise

As a function of:• Overvoltage• Temperature• Number of pixels• Pixel arrangement• Optical trench between pixels

Installation at CTF3 in 2010.

Dark count rate for different samples

1 pixel fired

2 pixels fired

3 pixels fired

Dark count rate as a function of the overvoltage

1 MHz/mm2 @ 32V

Angela Intermite


Conclusion

Activity 2010

• A lot of BI activity in 2009• All Vacuum installation for BPMs complete• All MTVs for transverse profile measurements installed• Full diagnostics spectrometer line for PHIN has been designed, installed commissioned• Robust Bunch Length Measurement with Streak Camera• Calibration of non-destructive RF bunch length measurements using Streak Results

– BPR 30 GHz waveguide port– RF pickup

• Measuring of the bunch spacing with the Steak and Phase monitor well mature– Systematic corrections due to bunch length variation & long bunches add additional complications to

phase monitor measurement– Proposal to use 1.5 GHz down mixing technique (See extra slides Steve Smith)

• New BPM calibration software for different gain / attenuation settings – for 2010• “Linac /CR type” BPM electronics for CLEX will be ready for 2010 run• Design for time resolved spectrometry for TBL mature – ready for manufacturing• Design of bunch length measurement for CLEX has started

– Long Optical lines to New Streak Camera Lab with new FESCA 200 Streak Camera– Non destructive RF based bunch length measurement techniques

• Bunch spacing measurement for PHIN (phase coding)– BI will give support where possible (manpower/BI priorities)

Review of beam instrumentation in CTF3 A. Dabrowski, 06/05/2010Activity 2010

Thank you for all you contributions & to the operators for the beams!


EXTRA SLIDES


Position Monitors – CTF3 & R&D CLIC

C. Simon, CEA Saclay6 units

Reentrant Cavity BPM - Califes

Inductive Pickup - EuroTev BPM

Lars Soby et al. @ CERN 3 units

Juanjo Garrigós et al .Valencia Univ.

(based on M. Gasior BPM, scaled for reduced aperture – 24 mm)16 units

Inductive Pickup (BPS) - TBL

CTF3 Sept. 2009, Operation2 BPE’s + 54 BPI’s + 46 BPM’s + 6 Re-Cavity BPM + 16 BPS + 5 WCM’s


Procedure

Extracting the

Bunch Length


Single bunch profile – Skew Gaussian

Extract FWHM from measured parameters

Error of FWHM (in pixels) given by:

Calibrate and subtract quadratically contributions from jitter and PSF

Peak = µ

where:

)()()( 222

px

pscalpxPSFJitterBLpsBL FWHMFWHM

Propagate all error contributions … Typical measurement error on FWHM is 4% (2 sigma)


BPM

TL /

CLEX

Acquisition system

Activity 2010

Lars Soby


New

BPM

Calibration

Software for

TL2

and

CLEX

Activity 2010

• New BPM acquisition Software with Java user interface in development• Compliments the exisiting sampler software used by CTF3 OP• Software will

– Communicate to calibration controller crate– Calibration of all BPMs for 2 different gain ranges and attenuation ON/OFF settings– Save calibration coefficients to a database log

Setup a gain and attenuator. 2 devices, one per rack. FECs:• cfv-2010-actfrfmt• cfv-2010-bpmtl2tbts

Setup a gain and attenuator. 2 devices, one per rack. FECs:• cfv-2010-actfrfmt• cfv-2010-bpmtl2tbts

According to the given settings, uploading operational calibration coefficients to 396 devices linked to pickups and to 6 devices of a WCM

According to the given settings, uploading operational calibration coefficients to 396 devices linked to pickups and to 6 devices of a WCM


Measurement

of the

Phase

Coding in

PHIN

Activity 2010

• Phase coding for the PHIN laser will be implemented & tested late 2010/2011• Phase coding of the Laser verified with the Streak Camera

– BI/PM will provide support (depending on other priorities)– Design optical line with only a small fraction of laser photons

• Measurement phase switch on the electron beam to be designed– Proposal

• Generate Cherenkov photons with a sapphire Chrystal• Hardware exists from CTF2 –compatibility with PHIN beam parameters (beam size,

bunch charge) to be checked• Build an optical line to transport photons from PHIN to the laser room• Image these photons with the Streak Camera

– BI/PM will provide support (depending on other priorities)


Proposal for phase

monitor for the

DL based on 1.5

GHz signal (S.

Smith)

Activity 2010

• To remove the bunch length dependency from phase measurement propose a phase measurement based on a lower frequency

• For DL Loop 1.5 GHz rate in 3.0 GHz rate out• Ideal output is periodic at 3.0 GHz• Path length error yields signal periodic at 1.5 GHz

– gives rise to 1.5 GHz component in signal– So does slow intensity modulation in input beam

• For the DL, S. Smith proposes to use a reference 1.5 GHz signal, and to downmix it with the beam signal hence measuring directly any residual 1.5 GHz beam component that indicate a poor combination after the Delay Loop

• Simulations show sensitivity to < 1ps shown in his simulations• Much of the BPR-S pickup and electronics can be reused, only e need an (unlocked) RF source

for the 1.425 MHz LO signal needed and 1.5 GHz mixer– Block Schema – see extra slides

• For measuring the bunch combination efficiency in the combiner ring, no simple schema available

– Needs more work to find a bunch length independent schema

Steve Smith


Simulation of 1.5 GHz down-mixing phase monitor

• Simulate ±1ps delay loop timing error. – Modulated at 25 MHz to make it stand out

in the simulation, • Add ±10% charge variation in alternating (3

GHz) buckets. – modulated at 10 MHz (for visibility) – expect errors quasi-static in real machine

• The simulated LO is phased to make the timing error show up as (almost) purely real.

• find a scale of 10 mV/ps timing error. • demodulate the 1.5 GHz to:

– real component (timing error) – imaginary component (amplitude mis-

match)• The amplitude of the 1.5 GHz signal is

completely dominated by the amplitude mis-match signal

– one can still cleanly extract the timing error signal.

– The timing error signal contaminated ~0.5ps level by the charge variation

Steve Smith


Systematic: 1.5 GHz phase monitor

• Resolution is not limited by the signal strength– but by systematic like charge variation present on drive beam.

• Expect that to get the timing correct to 1 degree of 3 GHz (1ps) one needs the current the same to 1% over the delay of the delay loop

• However at 1.5 GHz this signal is in quadrature to the timing error signal– can in principle be separated.

• The charge difference signal is in phase with the 1.5 GHz bunches, where the time error signal is 90 degrees out of phase,– that is it comes from the alternating short and long gaps between bunches

and is phased with the center of the short gap. • Guess: reduces sensitivity to charge variation by x100• Could probably tolerate 10% charge variation over the train and still measure

delay loop timing errors of <1 ps.

Steve Smith


Path-Lenth Diagnostics Conclusion

• Looks straightforward to measure timing to <1ps. Most of the hardware already exists.

• We need a mixer, a couple of filters and possibly an amplifier or two and probably a couple of pads.

• We need an (unlocked) RF source for the 1.425 MHz LO. • And software!

Steve Smith


Extension of same principle to RF-pickup


Measurement principle:1. Measure the amplitude of the beam harmonic (30-172 GHz) of interest2. The correlation between amplitude and bunch length depends on the bunch

shape3. Normalize the power to changes in the charge and the position squared in the

cavity

shape)bunch Gaussian (example2

22

22

2c

o

t

eqr

rP

Example for Talk:1. 33 GHz beam harmonic (since bunches rather long during calibration)2. ADC is 2 GS/s, typically use 4000 points, 2 micro second time window,

delta t = 0.5 ns (X10 faster than BPRW sampling)3. LO can be chosen to have an IF that gives the best sampling of the bunch

length variation along the pulse

Beam acceleration

Beam harmonic #

Beam harmonic

Fixed first Mixing

Variable Mixing

IF IF (measured)

2.99855 GHz 11 32.984 GHz 26.5 GHz 7.2 GHz 716 MHz 735 MHz


Implement band-pass filtering in the time domain

Power measurement in time domain

|power| [au]

Corresponding Frequency domain

Apply Band pass filter to isolate beam signal

Calibrate the Streak form factor vs. RF pickup power measurement

Use linear function

pfg(x) = a*x+b

a = 0.04571 (0.03752, 0.05389)b = 0.1536 (0.104, 0.2032)

Good agreement between RF pickup and Streak

Calibrated RF-pickup 33 GHz form factor


Energy gain and Bunch LengtheningDue to the pulse compression system, phase sag along the Klystron pulse ~ 5-15° (see talk of A. Dubrovskiy)

not all bunches see same RF phase Difference energy gain of one bunch with respect to another Within a single bunch, the head and tail of bunches to have different energy

Example of RF amplitude and phase for MKS03

Blue 30 GHz power signal before injector chicaneGreen 30 GHz power signal after injector chicane

Bunch length variation along the pulse train is a feature in CTF3 (to a greater / lesser extent depending on RF)

Time resolved bunch length diagnostics essential


OTR screens for

PHIN spectrometer

opti

mised

• Surface current on Aluminized mylar screen imaged

• CTF3 note 099 Replaced by Al foil works well

PHIN camera at 4.8 deg to vertical for optimal light acceptance

Maylar screen

Al screen works better


Ti

me resolved

energy

measurement

TBL

Activity 2009-2010

First measurement at TBL (November 2009)

Installed single slit segmented dump (2009)

Slit dump already used to understand TBL slit 1mm wide Length 100 mm iron


Instrumentation activity in 2009 New Installations

7 Transverse Profile Monitors• CC.MTV0532 – TL2 emittance tank• CC.MTV0970 - TL2’ emittance tank• CB.MTV1070 – TBL emittance tank• CBS.MTV0100 – TBL spectrometer line• CM.MTV0590 – TBTs drive beam side (Uppsala / Saclay mechanics & planning – CERN

acquisition & control)• CA.MTV0790 – TBTs probe beam side (Uppsala / Saclay mechanics & planning –

CERN acquisition & control)• CF.MTV0100 – PHIN Virtual Cathode

PHIN- 2 Gated (5ns ; 100ns) Intensified Camera- 4.8 degrees to the spectral reflection for OTR based profile measurements- 20 Channel Segmented Dump for Time Resolved Profile Measurements designed, built

and commissioned

Maintenance and Improvements Replacement of damaged screens Planning for an Linac / Combiner ring type BPM acquisition hardware &

calibration software for CLEX

Activity 2009

Review of Beam Instrumentation in CTF3

Documents

Transcript of Review of Beam Instrumentation in CTF3