Download 229_1.PDF

LHC Beam Loss Monitor System Design
B. Dehning*, G. Ferioli*, W. Friesenbichler*, E. Gschwendtner* and
J. Koopman*
*CERN, CH1211 Geneva 23, Switzerland
Abstract. At the LHC a beam loss system will be installed for continuous surveillance of particle
losses. The system is designed to prevent hardware destructions, to avoid magnet coil quenches and
to provide quantitative loss values. Over 3000 ionization chambers will be used to initiate the beam
abort if the loss rates exceed the quench levels. The time and beam energy dependent quench levels
require the acquisition of chamber currents in the range from 50 pA to 0.5 mA and an update of the
values every 89 [is. The acquisition and control electronics will consist of a front end electronics
near (< 400 m) to the ionization chambers and a threshold controller in the surface buildings. The
front end will include a charge balance converter, a counter and multiplexer part. The charge balance
converter is most suiteable to cover the large dynamic range. The introduced error is smaller than
few % in the required dynamic range. Six channels will be transmitted over one cable of up to 3
km length. The threshold controller will issue warnings and dump signals depending on the beam
energy and the loss durations.
INTRODUCTION
At the nominal energy of 7 TeV each beam in the LHC stores energy of up to 0.35 GJ.
The loss of only a fraction (10~8 ) of the beam can have a severe impact on the
smooth machine operation. Therefore the beam loss detection system must fulfil several
requirements, first: Protection: The magnets and other equipments must be protected
from damage due to beam losses. Second: Prevention: The super conducting coils of the
magnets could be come normal conducting by heat depostion from lost particles. In both
cases beam dumps are initiated before a dammage or quench will occur. Thired: Serve
as a beam diagnostic tool.
For protection and prevention the beam loss monitors trigger the beam dump via the
beam interlock system, whenever they detect beam losses above a certain limit. The
quench levels of the super-conducting magnets define this limit.
Detection of shower particles outside the cryostat or near to the collimators will be
used to determine the coil temperature increase due to particle losses. The relation between loss rate and temperature increase (quench levels) is based on shower simulation
and heat transfer considerations [1]. The expected particle flux outside the cryostat is as
well based on shower simulation.
CP648, Beam Instrumentation Workshop 2002: Tenth Workshop, edited by G. A. Smith and T. Russo
© 2002 American Institute of Physics 0-7354-0103-9/02/$19.00
229
1.E+15
quench levels [proton/s]
1.E+14
1.E+141.E+13
———^%J
total quench
jenehlevels
levelsatat450
450GeV
GeV ——
total quench levels at 7 TeV
.(/>
1.E+12
1.E+12I
£1.E+11
1.E+11 •
1.E+10
1.E+10-
heat
reserve of
cable
cable |
******
^
^*x**»»
*•
==
1 heat
heatflow
flowbetween
between
cable and He
—
—
1
1.E+09
He
Heheat
heatreserve
reserve
cooling
>olingflow
flow |=
———|He
Hec(
^"Xn,
1.E+08
1.E+08"""Biiniiiiim^
1.E+07
1.E+06
1.E-02
1.E-02
*******»«»
1.E-01
1.E-01
1.E+00
1.E+00
1.E+01
1.E+01
1.E+02
1.E+02
1.E+03
1.E+03
1.E+04
1.E+04
1.E+05
1.E+05
1.E+06
1.E+06
time
time[ms]
[ms]
FIGURE
FIGURE 1.1. The
The bending
bending magnet
magnet quench
quench levels
levels asas function
function of
of the
the loss
loss duration
duration for
for two
two different
different
energies.
energies.
QUENCH
QUENCH LEVELS
LEVELS
The
Thecoils
coilsof
ofthe
themagnets
magnetscan
canquench
quenchififaalocal
localenergy
energy deposition
depositiondue
duetotobeam
beam particle
particle
losses
losses increases
increases the
the coil
coil temperature
temperature to
to aa value
value were
were the
the conductor
conductor changes
changes from
from the
the
super-conducting
2]. In
super-conductingtotonormal
normalconducting
conductingstate
state[1,
[1,2].
Inaddition
additionto
tothe
theenergy
energy dependence
dependence
of
of the
the quench
quench levels
levels (see
(see Fig.
Fig. 1),
1), aa strong
strong dependence
dependence on
on the
the duration
duration of
of the
the loss
loss isis
observed.
observed.At
At450
450GeV
GeVthe
the quench
quench limit
limit rate
rate isis 1110
101313 protons/s
protons/s corresponding
corresponding 0.9
0.9 10
1099
protons
s) and
protons for
for aa loss
loss duration
duration of
of 11 turn
turn (89
(89 µ/is)
and the
the state
state state
state rate
rate (t(t >> 100
100 s)s) isis
77 10
1088 protons/s.
protons/s.At
At77 TeV
TeVthe
the magnet
magnet quench
quench rate
rate isis 1110
101010 protons/s
protons/s corresponding
corresponding
toto0.9
0.910
1055protons
protonsper
perturn
turnand
andthe
the state
state state
state rate
rate (t(t >> 50
50 s)s) isis 8810
1066 protons/s.
protons/s.
The
The thermal
thermal conductivity
conductivity of
of the
the surrounding
surrounding helium
helium flow
flow determines
determines the
the maximum
maximum
loss
loss rate
rate atat long
long time
time intervals.
intervals. At
At short
short time
time intervals,
intervals, the
the heat
heat reserve
reserve of
of the
the cables,
cables,
as
aswell
wellas
asthe
the heat
heat flow
flow between
between the
the superconducting
superconducting cables
cables and
and the
the helium,
helium, tolerates
tolerates
much
muchhigher
higherloss
lossrates.
rates.The
Theheat
heatreserve
reserve of
ofthe
thehelium
heliumdetermines
determines the
the quench
quench levels
levels atat
intermediate
intermediatetime
timescales.
scales.
BEAM
BEAM LOSS
LOSS PARTICLE
PARTICLEDETECTION
DETECTION
At
At the
the positions,
positions, where
where most
most of
of the
the beam
beam losses
losses are
are expected,
expected, simulations
simulations of
of the
the particle
ticle fluences
fluences outside
outside the
the cryostat
cryostat induced
induced by
by lost
lost protons
protons at
at the
the aperture
aperture have
have been
been
performed
performed with
with the
the Monte
Monte Carlo
Carlo Code
Code Geant
Geant 3.21.
3.21. The
The geometry
geometry used
used in
in this
this simulasimulation
tioncorresponds
correspondstotothe
thedispersion
dispersionsuppressor.
suppressor.Calculations
Calculationsfor
forthe
the arc
arc have
have already
already been
been
presented
presentedinin[2].
[2].The
Thesimulated
simulatedshower
showerparticles
particlesproduced
producedby
bylost
lostprotons
protons are
are counted
counted in
in
230
counts
two
placed all
all along
along the
the cryostat
cryostat on
on both
both sides.
sides. Also
Also the
theenergy
energydeposition
depositioninin
two detectors
detectors placed
these
elements
is
calculated.
Figure
2
gives
a
typical
example
of
detector
signals
caused
these elements is calculated. Figure 2 gives a typical example of detector signals caused
700
:
'
TOO
|
600
600
|
500
5OO
400
'4OO
300
300
200
200
100
100
0
i
0
-4000
'
'
'
|
11
111
1
Jit
'
'
'
'
|
'
'
'
'
|
i
i
i
|
|
|
liU
^
'
'
'
'
=
E
E
^
=
::::::|l|i|::z
J
™ IL nJl.1 1 i MI nil IL mini
-3000
-4C>OO
ILII Lniilliiiui
-2000
-3OOO
ilniiii
.in.
n i ,
1=
-1000
0
O
location [cm]
location [cm]
-2OOO
-1OOO
MB10
MB10
Q10
Q10
MQML MC
MI:
MQML
BC
B(
BB
MBB10
MBB10
MBA10
MBA10
FIGURE
detector signals
signals of
of shower
shower particles,
particles, caused
caused by
byaapoint
pointlike
likeloss
lossininthe
themiddle
middleofof
FIGURE 2.
2. Simulated
Simulated detector
the
magnet MQML.
MQML.
the quadrupole
quadrupole magnet
by
particles, which
which are
are induced
induced by
by point
pointlike
like losses
lossesof
ofthe
thebeam
beamin
inthe
themiddle
middle
by the
the shower
shower particles,
of
MQML (j8-function
(β -function maximum).
maximum). The
The shower
shower maximum
maximum isis about
about 11m
m
of the
the quadrupole
quadrupole MQML
after the beam loss location. The shower
shower width
width is
is 0.7
0.7 m.
m. For
For one
one lost
lost proton
proton of
of 77 TeV
TeV
−2
2
110
observed. The
The second
second shower
shower maximum
maximumisisdue
due
110~2 charged particles/proton/cm2 are observed.
to the gap between the quadrupole
quadrupole (MQML)
(MQML) and
and the
the dipole
dipole (MBB).
(MBB). The
The signals
signalsin
inthe
the
detectors on the opposite side of the
the cryostat
cryostat are
are reduced
reduced in
in amplitude
amplitudeand
andthe
thebeam
beamloss
loss
maxima from beam 1 and beam 2 are separated between
between 1 and 2 m
m [3].
[3]. Positioning
Positioningbeam
beam
loss monitors at the shower maxima locations
locations fulfils
fulfils the
the requirements
requirementsfor
forthe
thedistinction
distinction
between the two beams and
and for
for localising
localising the
thebeam
beamlosses.
losses.ItItisisforeseen
foreseentotoplace
placethree
three
detectors on either side of the cryostat
cryostat 3 at
at the
the optimal
optimal position
position for
for the
theobservation
observationofof
beam losses located in the middle of the
the quadrupole
quadrupole magnet
magnetand
andlocated
locatedat
atthe
thetransitions
transitions
between the quadrupole magnet and the
the bending
bending magnets.
magnets.
mm
- 4 4 - 3
-1
•
1
1
%
4
FIGURE 3. Location of the beam loss detectors
detectors near
near to
to the
the quadrupole
quadrupolemagnets
magnetsMQ.
MQ.
231
Beam
Beam
N2
High
HighVoltage
Voltage
m —ji—I
——jjfc
g;
*
Signal Currei
Current
Signal
—^- , I
1—*—————
\
\
Hi
Signal
SignalOutput
Output
mm
66mm
FIGURE4.4. Left:
Left: Schematic
Schematic structure
structure of
of the
the baseline
baseline ionisation chamber.
FIGURE
chamber. Right:
Right: Principle
Principle circuit
circuitdiadiagramfor
forthe
thefrontend
frontend electronic
electronic circuit.
circuit.
gram
Tunnel
lonization
Ionization
Chamber
Chamber
max. 400 m
Analog
Front end
Surface
Surface
Dump
Dump
Controller
Controller
max. 2 km
Analog Signals
Digital Signals
FIGURE 5. Schematic of the beam loss monitor readout chain.
FIGURE
5. Schematic of the beam loss monitor readout chain.
BEAM LOSS
LOSS DETECTORS
DETECTORS
BEAM
Ionisation chambers will be used as beam loss monitors. The baseline layout is a N 2
Ionisation chambers will be used as beam
loss monitors. The baseline layout is a N
filled cylinder with a surface of 80 cm2, a length of 19 cm and operated with a bias2
filled cylinder with a surface of 80 cm2 , a length of 19 cm and operated with a bias
voltage of V = 800 - 1800 V ( see fig. 4). The chamber electrodes consist of 31 parallel
voltage of V = 800 - 1800 V ( see fig. 4). The chamber electrodes consist of 31 parallel
plates. All the odd and even plates are connected together to increase the electrical field
plates. All the odd and even plates are connected together to increase the electrical field
volume, which has a strenght of up to 3 kV/cm. To stabilize the high voltage an low pass
volume, which has a strenght of up to 3 kV/cm. To stabilize the high voltage an low pass
filter (R = 1 MQ and C = 0.5/iF) is mounted on the chamber feed throughs.
filter (R = 1 MΩ and C = 0.5µ F) is mounted on the chamber feed throughs.
STRUCTURE OF THE BEAM LOSS MONITOR READOUT
STRUCTURE OF THE BEAM LOSS MONITOR READOUT
CHAIN
CHAIN
The first evaluation of this signal is done by the analog front end. Due to radiation
The
first evaluation
of this tunnel,
signal is
the analog
end. Dueand
to the
radiation
exposure
in the accelerator
thedone
cablebylength
betweenfront
the chamber
front
exposure
in thebetween
accelerator
tunnel,
theincable
length
theand
chamber
and m
theinfront
end can vary
several
meters
the arcs
of between
the tunnel,
up to 400
the
end
can vary
between
several
meters
in the radiation
arcs of the
tunnel,
up to
m in
the
straight
sections
(see fig.
5). The
increased
load
in theand
tunnel
(up400
to 10
Gray
straight sections (see fig. 5). The increased radiation load in the tunnel (up to 10 Gray
232
How to measure 100pA to 1mA
Directly
Log Amplifier
+ wide dynamic range
+ constant accuracy
Integrating
Transimpedance Amp
ADC
+ no further electronic
Switched Gain
Cascaded Amp
+ high SNR
+ low offset
+ fast I/U Conversion
+ stable system
- slow I/U conversion
- stability problems
- difficile treshold
adjustment
- slow I/U Conversion
- requires timing
- LSB about µV
Passive
- poor accuracy
- not for wide dynamic
range
Switched
Simple Reset
+ charge lost during
reset
- poor SNR
- high offset
- overshoots
Reset / Hold I
Reset/Hold
+ no charge loss
- 2 samples per cycle
2 samples per cycle
- no termination
no termination
achievable
achievable
BalancedCharge
Charge|
Balanced
+ no following ADC
required
+ wide dynamic range
+ wide dynamic range
- needs fast switches
- amplitude dependent
-amplitude dependent
conversion rate
Ion rate
FIGURE 6.
6. Overview
Overview of
of the
the applicable
applicable principles
principles to
to measure
measure electrical
electrical currents.
currents.
FIGURE
per year are expected)
expected) necessitates
necessitates a simple
simple and
and robust
robust design
design of
of the
the front
front end
end circuitry.
circuitry.
per
final evaluation
evaluation and the
the beam
beam dump
dump decision
decision is
is made
made by
by the
the threshold
threshold controller,
controller,
The final
surface building. The threshold
threshold controllers
controllers are
are located
located at
at eight
eight
which is located in a surface
locations where they
they are connection
connection to the
the beam
beam abort
abort system.
system. The
The cable
cable lenght
lenght between
between
the frontend
frontend electronic
electronic and the threshold
threshold controllers
controllers varriers
varriers between
between 1.8 and 2 km The
threshold controller itself
itself calculates the
the beam
beam losses
losses for
for different
different loss
loss durations
durations and
and
beam energies and compares them
them with
with the
the appropriate
appropriate treshold
treshold values.
values. A
A warning
warning or
or
beam dump signal is generated if the tresholds are exceeded.
FRONT END ELECTRONIC DESIGN
DESIGN
quench levels are varying
varying between
between 600 pA and 500 /lA.
µ A. To ensure high sensitivity
The quench
extended at least one order of magnitude from
from
at low losses, the dynamic range must be extended
the lower
lower limit
limit was set
set to
to 50
50 pA,
pA, which
which results
results in
in aa dynamic
dynamic
600 pA down to 60 pA. Thus the
dB.
range of 140
140dB.
The integrated loss rate
rate has
has to
to be
be acquired
acquired in
in an
an interval
interval between
between 89
89 /is
µ s and
and 100.
100. The
The
minimum time limit is given by the
the fastest
fastest beam
beam dump
dump possibility
possibility after
after one
one revolution
revolution
(89 /is)
µ s) and the longest interval is set by the temperature
temperature measurement
measurement system
system of
of the
the
He of the
the magnets
magnets to
to 100 s. Simulation studies revealed [1] that the quench level
±50 %. Consequently,
Consequently, the
the maximum
maximum allowable
allowable error
error of
of the
the
prediction is accurate to ±50
±10 %.
readout electronics was set to ±10
chamber leads
leads to
to the
the measurement
measurement of
of aa particle
particle loss
loss rate
rate
The usage of the ionization chamber
design considerations
considerations to
to acquire
acquire this
this current
current have
have
equivalent electric current. Several design
overview of the different
different methods that
that have
have been
been
been made [4]. Figure 6 shows an overview
taken into account.
Direct monitoring techniques
techniques provide
provide an output
output signal
signal that
that is
is proportional
proportional to
to the
the
input current. A transimpedance amplifier,
amplifier, e.g.,
e.g., transforms
transforms the
the input
input current
current into
into aa
sampled with
with an ADC at the
the required
required
proportional voltage. The output voltage can be sampled
233
Integrator
C
Treshold
comparator
iin(t)
U
Vtr
Iref
∆U
One-shot
∆T
UTr
Reference
current
∆U
T
fout
V-
FIGURE
FIGURE7.7. Left:
Left:Circuit
Circuitdiagram
diagramofofthe
thecurrent
currenttotofrequency
frequency converter.
converter. Right:
Right: Output
Output voltage
voltage of
of the
the
operational
operationalamplifier
amplifierversus
versustime
timeapplying
applyingaaconstant
constantinput
inputcurrent.
current.
frequency,
frequency,which
whichisisdetermined
determinedby
by the
the evaluation
evaluation time
time of
of the
the particle loss rate. It is not
possible
possibletotoachieve
achievewith
withthis
thismethod
methodaadynamic
dynamicrange
range of
of more
more than
than approximately
approximately 60 dB
without
withoutintroducing
introducinglarge
largeerrors.
errors.Thus
Thusseveral
severalgain
gainstages
stages have
have to
to be
be incorporated,
incorporated, which
have
havesome
someother
otherfeatures
features(see
(seepros
prosand
andcons
consindicated
indicated at
at the
the figure:
figure: 6).
6).
Integrating
Integratingtechniques
techniquescompared
comparedto
todirect
directmonitoring
monitoring provide
provide only
only average
average loss rates,
but
butoffers
offers unique
unique properties
properties inin terms
terms of
of dynamic
dynamic range
range and
and the
the signal
signal to noise ratio.
The
Themost
most simple
simple way
way toto realize
realize such
such aa circuit
circuit would
would be
be an
an ADC that offers
offers a high
dynamic
dynamicrange.
range.Several
Severalcomponents,
components,especially
especially for
for audio
audio applications,
applications, have been found
but
butnone
noneoffered
offered aasufficient
sufficient low
lowinput
inputcurrent.
current.The
Theswitched
switched integrator
integrator isis aa widely
widely used
used
circuit
circuittotomeasure
measuresmall
smallcurrents.
currents.An
Anelectric
electric current
current charges
charges the
the feedback
feedback capacitor of
an
anoperational
operationalamplifier
amplifier and
and the
the output
output voltage
voltage isis equal
equal to
to the
the integral
integral of the current
overtime.
time.However,
However,the
thedynamic
dynamicrange
range isis only
only 60
60 dB.
dB. This
This range
range can
can be
be extended,
extended, if a
over
certaintreshold
tresholdvoltage
voltageisisintroduced.
introduced.IfIf the
the output
output voltage
voltage reaches
reaches the
the treshold within
certain
thetime
timeinterval,
interval,the
theintegrator
integrator will
will be
be reset.
reset. At
At the
the end
end of
of the
the integration
integration interval, the
the
actualoutput
outputvoltage
voltageisissampled
sampledas
as before,
before, but
but also
also the
the number
number of reset actions within
actual
theinterval
intervalare
aretaken
takeninto
intoaccount.
account. As
As long
long as
as the
the reset
reset time
time isis much
much shorter
shorter than
than the
the
integrationtime,
time,the
theintroduced
introducederror
error can
canbe
be neglected.
neglected.
integration
Thecurrent
currenttotofrequency
frequency converter
converterfunctional
functional principle
principle (CFC)
(CFC) is
is shown
shown in
in figure
figure 7. It
The
consistsofofan
anintegrator,
integrator,whose
whosecapacitor
capacitorisischarged
chargedby
by the
the signal
signal current and discharged
consists
byaafixed
fixedreference
referencecurrent.
current.The
Theconcept
conceptof
ofaaCFC
CFC was
wassuccessfully
successfully improved
improved and
and used
used
by
byH.
H.Reeg
Reeg[5,
[5,6].
6].The
Theauthors
authorsinvestigations
investigationsare
are based
based on
on the
the work of E. G. Shapiro [7].
by
Figure7,7,right
rightdepicts
depictsaatypical
typicalanalog
analogoutput
outputvoltage
voltageof
of the
the CFC
CFC with
with aa constant
constant input
Figure
currentthat
thatisisalmost
almostatatthe
themaximum
maximumvalue
valueof
of the
the operating
operating range.
range. If
If the
the output voltage
current
rampsnegative
negativeand
andreaches
reachesthe
thetreshold
tresholdvoltage
voltageUUTrrr,,the
theone-shot
one-shotswitches
switches the
the reference
reference
ramps
currentsource
sourceIre
\ref r totothe
the summing
summing node
node for
for aa fixed
fixed time
time interval
interval ∆T.
AT. As the reference
current
currenthas
hasthe
theopposite
oppositepolarity
polarityand
andisislarger
larger as
as the
the signal
signal current,
current, the integrator output
current
voltageisisincreasing.
increasing.Later
Later IIre f * isis disconnected,
disconnected, the
the output
output ramps
ramps again
again down
down until it
voltage
reachesthe
thethreshold
thresholdpotential.
potential.Assuming
Assuming an
an ideal
ideal operational
operational amplifier,
amplifier, with
with no
no input
input
reaches
234
1.E+07
1.E+07
1%
0%
1.E+06
-1%
1.E+05
out
put
1.E+04
fre
qu
en 1.E+03
cy
[Hz
1.E+02
]
-2%
-3%
-4%
-5%
-6%
1.E+01
-7%
1.E+00
1.E-02
1.E-11
1.E-11
-8%
Frequency [Hz]
foff_corr [Hz]
Lin Error [%]
1.E-01
1.E-10
1.E-10
1.E-09
1.E-09
1.E-08
1.E-08
1.E-07
1.E-06
1.E-07
1.E-06
input current [A]
1.E-05
1.E-05
1.E-04
1.E-04
-9%
1.E-03
1.E-03
-10%
-10%
1.E-02
1.E-02
input current [A]
FIGURE
FIGURE8.8. CFC
CFCoutput
outputfrequency
frequencyand
andlinearity
linearityerror
errorversus
versusinput
inputcurrent
currentfor
foraamaximum
maximumcurrent
currentofof22
mA.
mA.
current
re f ==Ire
current and
and no
no offset
offset voltage,
voltage, aa fixed
fixed charge
charge QQ*
I f*· -∆T
AT isisextracted
extractedfrom
from the
the
summing
node
at
the
operational
amplifier
during
the
reset
period
∆T.
This
charge
summing node at the operational amplifier during the reset period AT. This charge isis
equal
equal to
to the
the the
the total
total signal
signal input
input charge
charge during
during the
the period
period T.
T.Because
Because the
the signal
signal charge
charge
and
and the
the capacitor
capacitor charge
charge are
are equal
equaltotoQQrereff ==IIreref · •∆T
ATduring
duringthe
thetime
time∆T
ATand
andsince
sincethe
the
accumulated
charge
of
the
capacitor
during
the
period
∆T
and
T
−
∆T
are
equal.
By
accumulated charge of the capacitor during the period AT and T — AT are equal. By
counting
counting the
the number
number of
of reset
reset actions
actions the
the input
input current
current isis convertted
convertted into
into an
an frequency,
frequency,
were
were one
one count
count represents
represents the
the average
average integrated
integrated charge
charge in
in the
the period
period T.
T. The
The relation
relation
between
between current
current and
andfrequency
frequency isisgiven
givenby:
by:
1
f=
I
(1)
(1)
Ire f ∆T in
The
The output
output frequency
frequency isis not
not depending
depending on
on the
the integration
integration capacitor.
capacitor.However,
However,itithas
hastoto
be
selected
carefully
because
it
determines
the
amplitude
of
the
integrator
output
voltage
be selected carefully because it determines the amplitude of the integrator output voltage
but
but also
also the
the sensitivity
sensitivity to
to noise
noise and
and to
to the
the charge
charge injection
injection from
from the
the current
current switch.
switch. The
The
relative
error
on
the
frequency
depends
only
on
the
square
root
of
the
quadratic
relative error on the frequency depends only on the square root of the quadraticsum
sumof
of
the
the relative
relative errors
errors on
onthe
the reference
reference currents
currents and
andon
onthe
thereference
reference time.
time.Since
Sincethe
therelative
relative
conversion
conversion error
error should
shouldbe
bebelow
below0.1
0.1this
thiscould
couldbe
bereached
reachedwithout
withoutany
anycalibration.
calibration.
The
output
frequency
of
the
CFC
was
measured
over
the
whole
dynamic
The output frequency of the CFC was measured over the whole dynamic range
range (see
(see
Fig.:
8)
[4].
The
measured
frequency
and
the
linearity
error
are
shown
in
figure
Fig.: 8) [4]. The measured frequency and the linearity error are shown in figure 8.8.The
The
straight
straight line
line isis set
set equal
equal to
to the
the data
data atat 55 kHz.
kHz. The
The CFC
CFC shows
shows aa deviation
deviation from
from the
the
straight
A. At
straight line
line of
of less
less than
than ±± 0.5
0.5 %%over
over an
an in-put
in-put range
range of
of 500
500pA
pAtoto100
100µ/lA.
Atlow
low
input
input currents,
currents, the
the error
error increases
increases mainly
mainly because
because of
of the
the total
total leakage
leakage current.
current. Above
Above
100
µ
A,
the
one-shot
works
partly
in
the
instable
mode,
which
leads
to
a
100 /lA, the one-shot works partly in the instable mode, which leads to a larger
larger error.
error. IfIf
the
A, the
the current
current exceeds
exceeds 200
200 µ/lA,
the CFC
CFC isis overloaded
overloaded and
andbecomes
becomes non-linear.
non-linear.The
TheCFC
CFC
235
is well within the specified error margin of ±10 %. The absolute error depends on the
actual value of I * and T but is constant over the whole dynamic range.
CONCLUSION
Longitudinal beam loss distribution studies show that losses concentrate on locations
with high j8-functions (centre of quadrupole magnets) or where mechanical limitations
of the aperture can be assumed. From shower simulations at the different loss locations
we see that a set of six detectors around the quadrupoles is sufficient for localizing
the beam losses and to distinct between the two beams. lonization chambers will be
used as shower particle detectors, which convert the particle loss rate into an electric
current. The expected ionization chamber current, equivalent to the quench levels, varies
between 500 pA and 500 /lA depending on the different loss distributions and detector
positions. Introducing a ten times higher sensitivity a dynamic range of 140 dB is
needed. Several possibilities to measure the current of the chamber are discussed. The
current-to-frequency converter is the most appropriate solution that is able to cover the
wide dynamic range with low effort. It produces an output frequency proportional to the
input current and simplifies the data transmission from the tunnel to the surface, where
the beam loss signal is evaluated. Tests revealed an excellent linearity of the current to
frequency converter. The frequency will be counted in intervalls of 40 /is and a serial
word will be transmitted together with 5 other channel over the cable to the location of
the threshold controller.
REFERENCES
1. J.B. Jeanneret et al., Quench levels and transient beam losses in LHC magnets, LHC Project Report
44,CERN,(1996).
2. A. Arauzo-Garcia et al., LHC Beam Loss Monitors, CERN-SL-2001-027-BI, CERN, (2001).
3. E. Gschwendtner et al., The Beam Loss Detection System of the LHC Ring, 8th European Particle
Accelerator Conference, (2002).
4. W. Friesenbichler, Development of the Readout Electronics for the Beam Loss Monitors of the LHC,
CERN-THESIS-2002-028 (2002).
5. H. Reeg and O. Keller, Linearer Strom-Frequenz-Konverter, Patent No. DE 195 20 315, Deutsches
Patentamt, (1996).
6. H. Reeg, A Current Digitizer for lonization Chambers, SEMS with high resolution and fast response,
Proceedings of the 4 th Workshop on Diagnostics and Instrumentation in Particle Accelerators DIPAC,
Daresbury, UK, pp. 140-142, (1999).
7. E. G. Shapiro, Linear Seven Decade Current-to-Frequency Converter, IEEE Trans. Nuclear Science,
Vol. 17, pp. 335-344, (1970).
236