commissioning of blm system
DESCRIPTION
Commissioning of BLM system. L. Ponce With the contribution of C. Zamantzas, B. Dehning , E.B. Holzer, M. Sapiensky. Outlines. Overview of the system signal available in the CCC Strategies for the thresholds settings hardware commissioning commissioning with beam conclusions. - PowerPoint PPT PresentationTRANSCRIPT
MPWG 9 March 2007 1
Commissioning of BLM system
L. Ponce
With the contribution of C. Zamantzas, B. Dehning, E.B.
Holzer, M. Sapiensky
MPWG 9 March 2007
Outlines
Overview of the system
signal available in the CCC
Strategies for the thresholds settings
hardware commissioning
commissioning with beam
conclusions
MPWG 9 March 2007
Quench protection system
(damage protection)
BLM system damage
protection, no redundancy
BLM for machine protection
Arc Dipole Magnet
The only system to protect LHC from fast losses (between 0.4 and 10 ms)
The only system to prevent quench
MPWG 9 March 2007 4
Detector
about 3800 ionisation chambers + 320 Secondary emission detectors measure the secondary shower outside the cryostats created by the losses
MPWG 9 March 2007 5
BLMS Signal Chain
Channel 1
Mul
ti-
plex
ing
and
doub
ling
Optical TX
Channel 8
Detector
Digitalization
Optical TX
VM
E B
ackp
lane
Optical RX
Optical RX
Sup
erco
nduc
tin
g m
agne
tSecondary particle shower generated by a lost
Demulti-plexing
Demulti-plexing
Sig
nal
sele
c-ti
on
Thr
es-h
olds
co
mp-
aris
on
Cha
nnel
sel
ecti
on a
nd
beam
per
mit
s ge
ner-
atio
n.
Status monitor
FE
E 2 Beam Energy
TU
NN
EL
SU
RF
AC
E
FE
E 1
Unmaskable beam permitsMaskable beam permits
Back End Electronics (BEE)
Front End Electronics (FEE)
MPWG 9 March 2007
Thresholds and interlocks
12 running sums for 32 energy levels for each channel, 16 channels per card, 345 surface cards. → table of 2 millions values!
Any of this signal over the thresholds generate a beam dump request via the BIC
28
Tunnel Card BLMCFC
1 23 . . .
28
Ionisation Chamber
Patch Box
Surface Card BLMTC (DAB64x)
GOH
GOH
Surface FPGA
MEMORY 16 bits
16 bits
16 bits
Anti-Fuse FPGA
Control
Control
16 bits
Control
Mezzanine
SRAM (1)Acquisition Data
SRAM (2)Acquisition Data
SRAM (3)Post-Mortem
data
data
data
Power PC
Control
Data
VMEInterface
Control
Data
Control
Control
Control
Control
1 23 . . . 8
2|||||||||||||
|||||||||||||||
|||||||||||||||
|||||||||||||||
CFC Analogue
. . .
Control
TLKOptical receiver
16 bits
Control
TLKOptical receiver
16 bits
Control
TLKOptical receiver
16 bits
Control
TLKOptical receiver
16 bits
Control
GOL
GOL
Tunnel Card BLMCFC
GOH
GOH
16 bits
16 bits
Anti-Fuse FPGA
Control
Control
1 23 . . . 8
2|||||||||||||
|||||||||||||||
|||||||||||||||
|||||||||||||||
CFC Analogue
. . .
Control
GOL
GOLIonisation Chamber
Combiner CardBLMCOM
DUMP (x3)
BEAM ENERGY
BEAM PERMIT
Beam Interlock Controller
BIC DUMP (x2)
BEAM PERMIT
Beam Energy Tracker
BETBEAM ENERGY
optical fibres
VM
E6 4
x B
us
Bac
kpla
ne
DUMP (x3)
BEAM ENERGY
BEAM PERMIT
Control
Data
Tunnel max 2km Surface
max 300m
Ionisation Chamber
... ...
Ionisation Chamber
Post-Mortem
DataLogging
Data
MPWG 9 March 2007 7
Loss pattern given by R. Assmann team (C. Bracco, S. Redaelli, G. Robert-Demolaize)
GEANT 3 simulation of the secondaries shower created by a lost proton impacting the beam pipe
simulation of the detector response to the spectra registered in the left and right detector (M. Stockner with G4)
500 protons same z position and same energy
impacting angle is 0.25 mrad
longitudinal scan performed to optimize the BLM location
1. Principle of the simulation
MPWG 9 March 2007
Definition of the thresholds
Loss pattern given by R. Assmann team (C. Bracco, S. Redaelli, G. Robert-Demolaize)
GEANT 3 simulation of the secondaries shower created by a lost proton impacting the beam pipe
simulation of the detector response to the spectra registered in the left and right detector (M. Stockner with G4)
500 protons same z position and same energy
impacting angle is 0.25 mrad
longitudinal scan performed to optimize the BLM location
MPWG 9 March 2007 9
Geometry description
3 transverse positions of impact outermost, innermost and top
MPWG 9 March 2007 10
Typical result
Maximum of the shower ~ 1m after impacting point in material
increase of the signal in magnet free locations
factor 2 between MQ and MB
z (cm)
MPWG 9 March 2007 11
Particle Shower in the Cryostat
catch the losses:
MB-MQ transition
Middle of MQ
MQ-MB transition
minimize uncertainty of ratio of deposited energy in the coil and in the detector
B1-B2 descrimination
Position of the detectors optimized to:
MPWG 9 March 2007 12
2. Position in the ARCS Example of topology of Loss (MQ27.R7) Peak before MQ at the shrinking vacuum pipe location (aperture limit
effect) End of loss at the centre of the MQ (beam size effect)
More simulation are needed to get better evidence (higher populated tertiary halo)
MPWG 9 March 2007 13
Particle shower in the detector
Addition of all the weighted signals from the previous locations
Positions chosen for the arcs also optimum for the DS.
MPWG 9 March 2007
Mainly BLMQI at the Quads (3 monitors per beam) + cold dipoles in LSS
Beam dump threshold set to 30 % of the quench level (to be discussed with the uncertainty on quench level knowledge)
Thresholds derived from loss maps (coll. team), secondaries shower simulations (BLM team), quench level simulations and measurements (D. Bocian)
BLMs for the arcs
beam 1
beam 2
MPWG 9 March 2007 15
MPWG 9 March 2007 16
BLMs for warm elements
beam 1
beam 2top view
BLM in LSS :at collimators, warm magnets, MSI, MSD, MKD,MKB, all the masks…
Beam dump threshold set to 10 % of equipment damage level (need equipments experts to set the correct values
MPWG 9 March 2007 17
Mobile BLM
use the spare chambers
use the spare channels per card : 2 in the arcs at each quad, a bit more complicated in the LSS.
electronics is commissioned as for connected channel
a separate Fixed display for non-active channels is planned : to be discussed
detection thresholds: ???
MPWG 9 March 2007
Generation of threshold table
Quench and damage level threshold tables will be created for each family of BLM locations.
They will be assembled together into MASTER table.
For every location a threshold for 7 TeV will be calculated.
Table will be filled using parametrized dependence on Energy and Integration time.
MASTER table, MAPPING table (BLM location vs electronic channel) will be stored in safe database.
MPWG 9 March 2007 19
Calibration and Verification of Models
Shower code(prediction error large
for tails)
Magnet quench(2 dim, energy, duration, large
variety of magnet types)
Threshold table
Detector(particle - energy spectrum
dependence)
Detector model (Geant) =
(CERN /H6)
Magnet model (Geant)=
HERA beam dump(tails of shower measurements)
Magnet model (SQPL)(heat flow, temp. margin, …)
= fast loss: sector test
slow loss: SM18
Calibration needed for: verification:
MPWG 9 March 2007 20
Reasons to change the thresholds? How often?
1) to check Machine Protection functionalities of BLMs (interlocks): decrease the thresholds in order to provoke a dump with low intensity
frequency: during the commissioning, after each shut-down (?),
for a set of detectors
2) study/check of quench levels (“quench and learn” strategy?):
implies dedicated MD time, post-mortem data analysis, could be related
to check the correct setting of the thresholds
Frequency : ? Probably during shut-down
For HERA, only one change since the start-up
MPWG 9 March 2007 21
1) commissioning of individual systems (MSI, LBDS, collimators) : to get a loss picture of a region, to give “warning” levels
adjust thresholds after studies of the systems to optimize the operational efficiency vs. the irradiation level
frequency : 1 or 2 iterations after determination of the thresholds and localized in space (injection region, IR7…)
2) To match quench level during commissioning (operational
efficiency):
Probably few iterations
some flexibility would help operation but is not an absolute need
MPWG 9 March 2007
Systematic Uncertainties at Quench Levels
about 1 % Radiation & analog elec.
Electronic calibration< 10 %Electronics
sim., measurements with beam (sector test, DESY PhD)
< 10 - 30 %
fluence per proton
Simulations
measurements with beam (sector test), Lab meas., sim. fellow)
Source, sim., measurements
Correction means
?Topology of losses (sim.)
< 200 %Quench levels (sim.)
< 10 – 20 %
Detector
relative accuracies
B. Dehning, LHC Radiation Day, 29/11/2005
MPWG 9 March 2007
3. Proposed implementation
Threshold GUI Reads the “master” table Applies a factor (<1) Saves new table to DB Sends new table to CPU
CPU flashes table if allowed (on-board switch)
Thresholds are loaded from the memory on the FPGA at boot.
Combiner initiated test allows CPU to read ‘current’ table.
SIS receives all tables
Compares tables Notifies BIS (if needed)
DETECTOR 16 – RS 12
DETECTOR 16 – RS 11
...
DETECTOR 01 – RS 03
DETECTOR 01 – RS 02
DETECTOR 01 – RS 01
DETECTOR 16 – RS 12
DETECTOR 16 – RS 11
...
DETECTOR 01 – RS 03
DETECTOR 01 – RS 02
DETECTOR 01 – RS 01
Master Table
Applied Table
Database
VME CPU
FLASH Threshold
Table
BLM Threshold Comparator
FLASH MEMORY
FPGA
SWITCH
VME Bus
READ Threshold
Table
BLM Combiner & Survey
SoftwareInterlock System
Threshold Preparation
GUI
CO
MP
AR
E T
HR
ES
HO
LD
S
FPGA
READ
APPLY FACTOR
SEND TOSYSTEMS
BIS
BIS
*
*
* *
*
*
* Secure transmission using MCS
FirmwareThresholds
MaskingOther Note: possible upgrade by adding a
comparison with master table on the board BUT feasibility has to be checked
MPWG 9 March 2007 24
Consequences on the reliability of the system?
Flexibility given by changing remotely the thresholds has to be balanced with the loss of reliability of the system
The proposed implementation allows both possibilities
But the remote access will have to be validated by machine protection experts when more detailed implementation of MCS and comparator are available (by the beginning of summer?).
MPWG 9 March 2007
System ready for LHC and fulfill the Specifications?
Hardware expected to be ready for LHC start-up Threshold tables (calibration of BLM) based on
simulations. Analysis effort of BLM logging and post-mortem
data (LHC beam data, “parasitic” and dedicated tests) to be started in 2006!
Calibration of threshold tables Interpretation of BLM signal pattern
Extensive software tools for data analysis essential to fulfill the specifications! Start now to specify and implement
MPWG 9 March 2007
BLMS Testing Procedures PhD thesis G. Guaglio
Radioactive source test (before start-up)
Functional tests
Barcode check
HV modulation test (implemented)
Double optical line comparison (implemented)
10 pA test (implemented)
Thresholds and channel assignment SW checks (implemented)
Beam inhibit lines tests (under discussion)
DetectorTunnel
electronics
Surface
electronicsCombiner
Inspection frequency:
Reception Installation and yearly maintenance Before (each) fill Parallel with beam
Current source test (last installation step)
Threshold table beam inhibit test (under discussion)
MPWG 9 March 2007 27
Commissioning Procedures - Steps
Environmental test: temperature dose & single event
Steps:
Elec. tunnel, 20 year of operation & “no” single event effects Elec. tunnel, 15 – 50 degree
Functional test: before installation during installation during operation
All equipment, LAB, current and radioactive source
Connectivity, current and radioactive source
Connectivity, thresholds tables
Calibration: before startup after startup Establishing model (detector, shower, quench behavior)
a: no beam abort , no quench, no actionb: use loss measurements and models for improvements
Calibration
Functional testEnvironmental test
Beam energy
detector LBDSBICsurface elec.tunnel elec.magnet
Particle shower
MPWG 9 March 2007
Hardware commissioning
complete detailed procedure documented in MTF
functionalities linked with Machine protection will be reviewed in the Machine Protection System Commissioning working Group
validation of the connectivity topology: registration in database of the link position in the tunnel-channel identification-thresholds
MPWG 9 March 2007
Commissioning with beam see presentation of A. Koshick in CHAMONIX
Motivation of the test: Establish thresholds = establish the correlation
between quench level and BLM signal = Calibration! Verify or establish „real-life“ quench levels Verify simulated BLM signal (and loss patterns) In particular: What BLM signal refers to the quench level of
a certain magnet type?
=> Accurately known quench levels will increase operational efficiency!
MPWG 9 March 2007 30
How to do the quench test: Implementation
Initial conditions & requirements: Pilot beam 5x109
Clean conditions, orbit corrected (to better +/- 3 mm?). BPM data/logging available Trajectory BLM data/logging available Additional “mobile” BLMs at the chosen locations
Set optics (3-bump)
Vary intensity5x109 – max. 1x1011
+logging all relevant data (BPM, BLM,BCT,emittance …)
Magnet quench
if circulating beam, possibility to check steady state losses quench limit. If sector test, fast losses quench limit
Simple idea: Steer beam into aperture and cause magnet quench
MPWG 9 March 2007 31
How to do the test: What we want to learnBeam Parameters Emittance Intensity Momentum spread
Optics Parameters -function Dispersion Trajectory/Orbit
Impact Length Impact Position
Proton density that caused quench in SC magnet
BLM signal
Determination of quench level Calibration
MPWG 9 March 2007 32
Conclusions
Controlled, defined test to establish Absolute quench limits BLM threshold values Model and understanding of correlation of loss pattern,
quench level, BLM signal
This test is essential for an early calibration of the BLM system, even if beam time consuming
It has to be done before increasing intensity