What you get ?Transverse and Longitudinal

distributionsEvian Workshop 2012

18-Dec-2012

F.Roncarolo, W.Andreazza, S.Bart-Pedersen, A.Boccardi, E.Bravin, B.Dehning, J.Emery, J-J. Gras, A.Guerrero, M.Kuhn, T.Lefevre, A.Nosych, M.Sapinski, G.Schneider, G.Trad, R.Veness, M.Wendt

Many thanks to: R.Jones, L.JensenOP teams (V.Kein, ….) A.Bertarelli, M.Garlasche & MME teamF.Caspers, E.Metral, B.SalvantG.Lanza, G.Bregliozzi and Vacuum Team(s)Many others ….

F.Ro

ncar

olo

- Evi

an W

orks

hop

Scope / Contents

SCOPEFocus on what did not work, what is missing, current limitations during the 2012 run. Present the changes to be made during LS1 and expected performance post LS1.

18-D

ec-2

012

2

CONTENTS Wire scanners (WS) Beam Gas Ionization monitor (BGI) Synchrotron light detector (BSRT)

Abort Gap Monitor (AGM or BSRA) Longitudinal Density Monitor (LDM)

Will not cover:Matching monitor, proposal for a new VELO-Like detector, WCM

F.Ro

ncar

olo

- Evi

an W

orks

hop

3

Wire Scanners – 2012 issues

Vacuum leaks- Bellow designed for ~ 10’000 scans, in 2012 1 system failed at ~10200

scans Wire breaking

- No evidence of breaking due to beam during normal operation (RF or direct energy deposition). Evidence of ageing due to sublimation.

- Wire damage can be traced back to power supply failure followed by server crash that left the wire in between IN and OUT position

Overall accuracy dependence on working point- PM voltage + filter settings on same beam give different beam sizes- Many studies in 2012 (see M.Khun’s talk)

SW, OP GUI- Judged as inefficient by OP (bunch selection, automatic scans, display)

Dumps due to BLM thresholds- Secondary shower amplitude depends on actual wire diameter, that

changes with ageing (see next slide)

18-D

ec-2

012

4

Wire Scanners – Dumps due to losses18

-Dec

-201

2F.

Ronc

arol

o - E

vian

Wor

ksho

p

22-Aug 12-Oct 22-Nov

System B1 H (aged wire, changed during intervention to fix bellow leak)

B1 H (new wire) B1 V (new wire)

Beam Intensity 4.29e12 No dump 4.18e12 Dump 3.54e12 Dump

BLM signal [Gy/s] 0.0091 0.0218 0.0335

Losses / proton@ downstream BLM [Gy/p]

5.4e-19 2.7e-17 2.4e-17

Aged wire partially sublimated smaller diameter lower losses

16 um

aged wire

34 um

new wire

F.Ro

ncar

olo

- Evi

an W

orks

hop

5

From beam profile to emittance18

-Dec

-201

2

Beta from modelBeta from beta beatingBeta from K mod

WS measurements during a test fill, with a high emittance and a low emittance bunch

Beta values during the ramp from linear interpolation 450GeV – 4 TeV

M.KhunG.Trad

6

WS - Upgrades

TS#4- Install 7 um wire on 1 system

- tests before the end of the run for assessing robustness and signal/losses

- Thinner wire more robust according to literature, but less material to sublimate before breaking

LS#1- Possibly thinner wires on all systems- Slightly higher speed (~10% max)- New bellows (aim at gaining a factor 5 in lifetime)- Improve OP GUI (OP+BI)- More system redundancy (big investment, under discussion)

LS#2 (?)- New fast (20 m/s) devices, following SPS prototype after LS#1- Possibly new detectors (e.g. diamonds) to replace scintillator + PM

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

F.Ro

ncar

olo

- Evi

an W

orks

hop

7

BGI – 2012 issues

Both beam 1 BGIs had MCP failures in early 2012 (MCPs were exchanged during winter TS). - Reasons of the failures are understood (Operational/Technical failure)- Protection measures in place (e.g. automatic HV shutdown).

Problem with remote camera gain control, important to provide repeatable beam size measurement.

Camera failures - intensifier reaching Mean Time To Failure

Difficult cross-calibration - no intensity overlap WS/BGI during p-p runs, BSRT B2 problem

Overall results interpretation still difficult

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

8

BGI – Results example

WS vs BGI during p-Pb MD (R. Versteegen, see CERN-ATS-2012-094 MD)- 13 Pb bunches, 7e9 charges/bunch

- could find good calibration w.r.t. to WS despite low BGI signal

18-D

ec-2

012

During p-p runs: Calibration more difficult Some evidence of

dependence on bunch length

F.Ro

ncar

olo

- Evi

an W

orks

hop

9

BGI - Upgrades

Will dismantle and re-surface the vacuum sealing surfaces to avoid leaks (some troubles in 2012 to have them leak-tight)

MCP refurbishment Optical system upgrade

- To cope with high brightness beams at 7 TeV HV system upgrade

- To ensure a more stable operation Camera refurbishment Low level SW re-design from scratch

What is the allowable gas budget during a year? Can we run continuously with gas injection?

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

10

BSRT - Introduction18

-Dec

-201

2

Gated camera (BSRTS)

DC CameraAbort Gap Monitor

(AGM)

Long. Density Monitor (LDM)

Optical delay line

Neutral filtersColor filters

Proton/Ion beam

90 %

10 %

60 %40 %

40 %60 %

F.Ro

ncar

olo

- Evi

an W

orks

hop

11

BSRT – 2012 Issues

Heating with high intensity beam- Mirror coating and mirror support failures

Absolute and relative calibration (even more difficult in 2012 , affected by heating)

Software- Fast scan on demand (expert GUI) working from early 2012- Fast scan server (communicating with OP-GUI - V.Kain ) tests started in

October, to be validated during p-Pb run Overall reliability – robustness affected by

- Extraction mirror heating / failures- FESA server automatisms failures

- Steering following heating and energy ramp- Camera gain adjustment following injections and energy ramp

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

12

BSRT – Calibration Example18

-Dec

-201

2

B1 with new optical line (focusing lenses instead of mirrors), 450 GeV during MD period

Excellent agreement BSRT – WS over a wide emittance range, after applying - Magnification within 10% w.r.t. nominal- PSF ~20% smaller than typical values with old optics

Scraping

B1 Vertical

Similar examples exist at 4 TeV

1.3 mm1 mm 0.8 mm

F.Ro

ncar

olo

- Evi

an W

orks

hop

13

BSRT – Heating – Findings after B2 mirror removal18

-Dec

-201

2

Mirror coating blistered

Mirror clamps deformed

Later Removal of B1 mirror evidenced similar effects- Both mirrors were: silicon bulk + dielectric coating

TS#3: replaced both mirrors- B1: glass bulk + metallic coating – OK only at low beam intensities- B2: silicon bulk + polishing (no coating) – not usable for imaging

F.Ro

ncar

olo

- Evi

an W

orks

hop

14

BSRT – RF simulations18

-Dec

-201

2 Longit. wake impedance of BSRT with and without Ferrite damping

Measured LHC bunch power spectrum. A 650 MHz resonance is very dangerous.

LHC bunch power spectrum

F = 650 MHzP_loss = 10-50W

Credits: T.Mastoridis,P.Baudrenghien/CERN

B-filed of the beam in Time Domain. Red = Hot (bigger current density)Blue = Cold

Mirror Back

Mirror front

E-field of a dominant resonating mode at 650 MHz. (Q = 1263 / Rsh = 25841 Ohm)

From January:RF laboratory measurements on spare tank

F.Ro

ncar

olo

- Evi

an W

orks

hop

15

BSRT – Upgrades

TS#4 – Intervention on B2 only Change mirror: dielectric bulk no coating glass bulk + dielectric coating 6 temperature probes in vacuum, to validate RF – Thermomechanical simulations

BASELINE for after LS#1 New optics New light parasitic shielding Extraction mirrors: likely glass bulk + dielectric coating

- Need high intensity test to validate it before end of this run ! Modified tank minimizing RF coupling Operational fast scan server

Studies during LS#1 Expected performances (resolution/accuracy) at 6.5-7 TeV

- Monitor lower wavelengths to reduce diffraction ? Novel tank design, much less sensitive to RF coupling / heating

- Reflective tapered pipe with view-port on the side instead of on the bottom?

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

16

Abort Gap Monitor

The weakest point remains the steering of the BSRT: “no spot no AG monitor”

A signal presence check could be implemented in the sequencer, but this doesn’t eliminate the need to verify that the telescope is not completely out of steering- BI is preparing a list of self checks to be implemented if possible during

LS1, but at the cost of dead times in the measure (less than 1% of availability loss)

Calibration was kept reasonably well updated- Re-calibration each time optical path changed (e.g. due to heating)

The new BSRT optics- Doesn’t use additional delay line to pass from Undulator to D3 imaging

- eliminates the need to compensate for light loss at the moment of the delay line insertion

18-D

ec-2

012

AFTER MANY DISCUSSIONS, WE GOT IN THE END A RATHER RELIABLE SYSTEM – WHEN THE BSRT IS ‘WORKING’. J. Wenninger – BI Day Dec.2012

F.Ro

ncar

olo

- Evi

an W

orks

hop

17

Longitudinal Density Monitor (LDM)

The LDM remains an expert tool- The SW is still under development - Artefacts linked to the detector behaviour still need an expert to correct

for in normal fills (no impact during VdM scans) - dead time and afterpulse seems to change with filling pattern, to

be checked after LS1- Some reflections can affect the measure and need to be recognized to

avoid misinterpretation: particular care was placed in the new optic setup to solve this for B1

It can measure satellites and ghosts up to 10^-4 but is not good to verify bunch shape at a fine level

18-D

ec-2

012

WITH 25 NS BEAMS WE HAVE TO GO BACK TO MAIN-MAIN COLLISIONS, THE INTEREST WILL THEN SHIFT BACK TO SATELLITES AT INJECTION AND LUMINOSITY CALIBRATION RUNS. J. Wenninger – BI Day Dec.2012

Noted …

F.Ro

ncar

olo

- Evi

an W

orks

hop

18

From 4 to 7 TeV

Smaller Beam Sizes- WS: less points per sigma, could need to apply multiple-scan or multiple

bunches overlaps to increase resolution (as done @ SPS Flat Top)- BGI, BSRT: will adapt optical imaging to have ~mm/pix as @ 4 TeV- BSRT: higher contribution from diffraction

- Can correct for it, after quantifying it precisely- Can think about going to lower wavelength detectors

Lower BLM thresholds - Limits WS scans to ~ 1e12

AGM, LDM- More photons, will adjust optical filters if needed

18-D

ec-2

012

Energy Limit Reason450 GeV 2.7e13p Wire damage4 TeV 3.6e12 p BLM threshold6.5 TeV ~1e12 p BLM threshold

F.Ro

ncar

olo

- Evi

an W

orks

hop

19

From 50 to 25 ns

WS cross-talk on electronics- ~ 10 % between 25ns slots (under study)- With a reduced wire diameter, can we scan 288b at

injection? BGI:

- no evident impact BSRT: 25ns would

- Increase the time to loop over all bunches (currently ~7min for 1380 bunches)

AGM, LDM:- No evident impact

18-D

ec-2

012

20

Conclusions - Overview

need development for improving the overall performance and reliability after LS#1- For each device we have a list of issues to study, improve, develop for 7 TeV and 25 ns

Fundamental checks before LS1- BSRT heating- WS thresholds

SW/Control- We need to profit of LS1 for reviewing all instruments low and high level SW individually

Transverse Profile Workshop, CERN, April 2013- Experts from GSI, DESY, FermiLab etc … presenting experience with BGI, Sync. Light etc …

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

Transverse

Longitudinal AGM reliability depends on BSRT optical line robustness. MP workshop will

trigger again interlocking? LDM needs some work to make it operational, efforts in LS#1, but will re-start

likely as expert tool (need for development and relative resources)

Transverse and longitudinal diagnostics allowed optimizing and safely running the LHC (WS as references, BSRT bunch-per-bunch, AGM)

Need high intensity run in Jan-Feb 2013 !!

F.Ro

ncar

olo

- Evi

an W

orks

hop

21

18-D

ec-2

012

SPARES

F.Ro

ncar

olo

- Evi

an W

orks

hop

22

BSRT – Optical Line18

-Dec

-201

2

Old Optics

New Optics

Entrance (steering) mirror

F.Ro

ncar

olo

- Evi

an W

orks

hop

23

Wire Scanners – Present Performances

Integration- 40 MHz sampling of PM integrator allows bunch per bunch measurements

- 50 ns ok- 25 ns cross-talk being studied

Repetition Rate- Ideally ~0.2 Hz, at cost of system lifetime (wire, bellows)

Dynamic range- From pilot bunch to ultimate intensity per bunch, but:

- Limits on total beam intensity

Future: faster WS (20 m/s?) - can allow higher intensities at the cost of

- multi-scans on a single bunch (go faster few points/sigma) need to overlap multi-scans with sampling position offsets

- single scan, combine NN bunches to have enough points/sigma

18-D

ec-2

012

Energy Limit Reason450 GeV 2.7e13p Wire damage4 TeV 3.6e12 p BLM threshold6.5 TeV ~1e12 p BLM threshold

F.Ro

ncar

olo

- Evi

an W

orks

hop

24

WS - Accuracy Resolution

- limited by minimum wire speed vs protons revolution frequency- 1 m/s 89 um between two consecutive wire position acq. ( profile points)

- Can be improved overlapping multi-scans (or single scan combining NN bunches) with sampling position offset (as being tested now @ SPS)

- Present wire position resolution limited by potentiometer noise (some 20um)- New WS: aiming for 2um resolution (independent of speed)

Accuracy - With proper PM and filter settings, absolute accuracy proved to be 1% for the SPS

linear WS- Accuracy of LHC WS under study

- theoretically equal to SPS linear WS- At the moment: evidence of dependence on working point (PM gain + filter

settings SLIDE/PLOT ON THIS?)- Plan for different secondary shower detector (diamond), related to SPS prototype to

be tested after LS1- Improve dynamic range- Get rid of filters avoid dependence on working point- PHD on electronics

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

25

BGI – Operational Specs18

-Dec

-201

2

Gating/Integration- Gated camera- Need to gate over multi-bunches to have enough signal (see dynamic

range) Repetition Rate

- 50Hz, limited by image digitalization (BTV)

Dynamic range- With a “fresh” MCP:

- 10 proton bunches with gas injection 10-8mbar- Single Pb ion bunch with gas injection 10-8mbar- A bit better at 4TeV due to denser beam

- MCP aging rather quick

F.Ro

ncar

olo

- Evi

an W

orks

hop

26

BGI – Results Example II

Typical p-p high intensity fill

18-D

ec-2

012

emittance “decrease” at the beginning of the ramp (related to bunch length?)

variations of calibration parameters from fill to fill

emittance “decrease” right after the ramp…

signal amplitude decrease during the fill (more then expected from intensity)

Measurement at injection, investigating bunch length influence on beam size seen by BGI.

Bunch length from 1.4 to 1 ns ns(NO RAMP)

F.Ro

ncar

olo

- Evi

an W

orks

hop

27

BGI - Performances

Resolution- Present optics gives 0.115 mm/pixel

Accuracy- Optics magnification validated to 1% by

- Beam orbit local bumps- Reference wire-grid calibration

- Needs cross calibration w.r.t WS and BSRT- For the moment not better than 20%, degrading with MCP aging- Many studies on going to understand ultimate resolution/accuracy

LS1: - Replace MCPs- Second camera with better performances

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

28

BSRT Operational Specs18

-Dec

-201

2

Gating- Intensified camera gating down to 25ns with a 12.5ns gating resolution

Repetition rate- Max 200 Hz (limited by intensifier trigger rate)- Present image digitalization (BTV) 50 Hz- Present control + acquisition SW ~12 Hz Can do bunch per bunch @ ~12Hz Can do single bunch single turn but not on consecutive turns

Dynamic Range- Protons: From pilot at injection (single turn, every 220 turns) to average

over all bunches at flat top- Ions: From ~30 bunches at injection to average over all bunches at flat

top

F.Ro

ncar

olo

- Evi

an W

orks

hop

29

BSRT – Heating vs bunch length18

-Dec

-201

2

Clear evidence that heating is due to RF coupling with high intensity beams

Temperature gradient changes at each bunch

length step

F.Ro

ncar

olo

- Evi

an W

orks

hop

30

BSRT - Performances

Resolution- Present optics 0.1 mm/pix, next: 0.05 mm/pix

Relative bunch per bunch accuracy <= 5%- 5% on single shot, dominated by reproducibility affected by noise (airflow,

optical elements vibration, fit accuracy, etc …)- 1% averaging on multi-shots

Absolute accuracy: - Optics magnification validated to <= 5%

- Calibration target- Beam orbit local bumps

- Ultimate accuracy dominated by aberration / diffraction- Need cross calibration w.r.t. WS - calibration factors accuracy <=10% after calibration

- Calibration factors not stable- Possible drifts due to mirror coating aging (heating)

18-D

ec-2

012

F.Ro

ncar

olo

- Evi

an W

orks

hop

18-D

ec-2

012

31

F.Ro

ncar

olo

- Evi

an W

orks

hop

18-D

ec-2

012

32

F.Ro

ncar

olo

- Evi

an W

orks

hop

33

BSRT - Heating

B2 Temperatures end of August

18-D

ec-2

012

Last two fills before putting mirror to OUT position and then removing it.

+ UFO activity ….

F.Ro

ncar

olo

- Evi

an W

orks

hop

34

BSRT – Heating – B2 Mirror 28-Aug-2012 18

-Dec

-201

2

1- Beam (spot not in the ‘right’ place vertically)

2- Beam dump

3- Close Vacuum sector and retract mirror to OUT

4-No beam, no motors movement, mirror movedAccess to remove mirror

F.Ro

ncar

olo

- Evi

an W

orks

hop

35

BGI - Introduction Collect electrons from beam-gas ionization

- Dipole B field to avoid drift from ionization location to MCP- MCP electron multiplication- Phosphor coupled to MCP output for electronphoton conversion- Imaging of phosphor output

Designed for heavy ions Enough signal from protons by injecting local pressure bumps or high

intensity Can monitor average relative beam size variation during the ramp

18-D

ec-2

012

M.Patecki, M.Sapinski

F.Ro

ncar

olo

- Evi

an W

orks

hop

36

Wire Scanners - Introduction

Reference device for transverse profile measurements- 1 H + 1 V per beam (+ a spare for each)- 30 um Carbon wire flying at 1 m/s

Scan on-demand Dynamic range controlled by PM gain and optical filters Can be used up to a maximum intensity that depends on beam

energy- Above such maximum intensity: wire damage and/or quench downstream

magnets ( BLM thresholds to dump before reaching quench limit)18-D

ec-2

012

Energy Limit Reason450 GeV 2.7e13p Wire damage4 TeV 3.6e12 p BLM threshold6.5 TeV ~1e12 p BLM threshold

F.Ro

ncar

olo

- Evi

an W

orks

hop

37

BSRT (and BGI) - Calibration18

-Dec

-201

2

sm [pixels] = Measured Sigma

M [mm/pixels] = System magnification (optical line + pixel size)

• determined by optical line adjustment with energy (UND D3)

• reference target vs closed orbit bumps give up to 10% differences

spsf [pixels] = measurement error due to • aberration, diffraction• extraction mirror aging/deformation

During 2012 heavily affected by extraction mirror heating (deformation + coating damages)Difficult to find a stable calibration w.r.t. WS

F.Ro

ncar

olo

- Evi

an W

orks

hop

38

Longitudinal Density Monitor (LDM) - Upgrades

After LS1: - new BSRT optics, new light shielding less parasitic reflections- new detectors under investigation (e.g. fast PM)

18-D

ec-2

012

B1: new optics B2: old optics Reflection on BSRT filter wheel, at a possible satellite locationSatellites 5ns

from SPS

5 ns

Different colors = different bunches

WITH 25 NS BEAMS WE HAVE TO GO BACK TO MAIN-MAIN COLLISIONS, THE INTEREST WILL THEN SHIFT BACK TO SATELLITES AT INJECTION AND LUMINOSITY CALIBRATION RUNS. J. Wenninger – BI Day Dec.2012

After L

Noted …