Content uploaded by Eduardo Nebot del Busto
Author content
All content in this area was uploaded by Eduardo Nebot del Busto on May 08, 2014
Content may be subject to copyright.
CERN-ATS-2012-095
20/05/2012
EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH
CERN – ACCELERATORS AND TECHNOLOGY SECTOR
CERN-ATS-2012-095
STUDY OF THE RESPONSE OF LOW PRESSURE IONIZATION
CHAMBERS
Eduardo Nebot Del Busto, Bernd Dehning, Ewald Effinger,
Viatcheslav Grishin, Juan Herranz Alvarez, CERN, Geneva, Switzerland.
Abstract
The Beam Loss Monitoring System (BLM) of the Large Hadron Collider (LHC) is
based on parallel plate Ionization Chambers (IC) with active volume 1.5l and a
nitrogen filling gas at 0.1 bar overpressure. At the largest loss locations,the ICs
generate signals large enough to saturate the read-out electronics. A reduction of the
active volume and filling pressure in the ICs would decrease the amount of
charge collected in the electrodes, and so provide a higher saturation limit using the
same electronics. This makes Little Ionization Chambers (LIC) with both reduced
pressure and small active volume a good candidate for these high radiation
areas. In this contribution we present measurements performed with several LIC
monitors with reduced active volume and various filling pressures. These detectors
were tested under various conditions with different beam setups,
with standard LHC ICs used for calibration purposes
Presented at the International Particle Accelerator Conference (IPAC’12) –
May 20-25, 2012, N. Orleans, USA
Geneva, Switzerland, May 2012
STUDY OF THE RESPONSE OF LOW PRESSURE IONIZATION
CHAMBERS
Eduardo Nebot Del Busto, Bernd Dehning, Ewald Effinger,
Viatcheslav Grishin, Juan Herranz Alvarez, CERN, Geneva, Switzerland.
Abstract
The Beam Loss Monitoring System (BLM) of the Large
Hadron Collider (LHC) is based on parallel plate Ioniza-
tion Chambers (IC) with active volume 1.5l and a nitrogen
filling gas at 0.1 bar overpressure. At the largest loss loca-
tions, the ICs generate signals large enough to saturate the
read-out electronics. A reduction of the active volume and
filling pressure in the ICs would decrease the amount of
charge collected in the electrodes, and so provide a higher
saturation limit using the same electronics. This makes Lit-
tle Ionization Chambers (LIC) with both reduced pressure
and small active volume a good candidate for these high ra-
diation areas. In this contribution we present measurements
performed with several LIC monitors with reduced active
volume and various filling pressures. These detectors were
tested under various conditions with different beam setups,
with standard LHC ICs used for calibration purposes.
INTRODUCTION
The BLM system [1] is responsible for protecting the su-
perconducting LHC magnets from quench and damage due
to beam losses. A large fraction of the BLM monitors are
ICs with circular parallel plate electrodes of 8.9 cm diam-
eter separated by 0.5 cm gaps. The chambers, filled with
N
2
at 1.1 bar as the ionization medium, are 50 cm long and
they have an active volume of 1.5l. At the largest expected
LHC loss locations, where the ICs collect enough current
to saturate the read out electronics, the ICs are replaced
by Secondary Emission Monitors (SEM) which give much
lower signals and can therefore be used to increase the dy-
namic range. The SEM consist of only three electrodes,
the central one being of titanium due to its secondary emis-
sion properties while the volume of the detector is kept at
vacuum (10
−7
mbar). However, the sensitivity of the SEM
detectors, measured to be ∼ 7 · 10
4
times lower than for
ICs, was so low that they very rarely measured signals.
The active volume of several SEMs were filled with N
2
at various pressures (0.1 bar to 1.1 bar) to study their re-
sponse as ionization chambers. In the text we will refer to
these detectors as LIC prototypes. Finally, several ioniza-
tion chambers with three aluminum electrodes and a filling
pressure0.4barwere built and tested. In the textwe will re-
fer to these detectors as LICs. In this document we describe
the response of several LIC, LIC prototypes and IC under
different irradiation conditions. An experimental setup on
the dump line of the Proton Synchrotron Booster (PSB) al-
lowed the response of the detectors to be verified against
fast pulses (∼ 50ns) of protons. The CNRad facility (con-
nection tunnel to the neutrino beam target area, CNGS) [2]
was used to irradiate the detectors with a mixed field of
medium duration (∼ 10µs). The chambers were also tested
with continuous losses in the LHC betatron cleaning areas.
PSB MEASUREMENTS
The response of one LIC at a pressure of 0.4 bar was
tested on the dump line of the Proton Synchrotron Booster
(PSB). An LHC IC was located in a neighbouring location
for comparison. The set up is shown in Figure 1. The two
detectors were situated on a movable device that allowed
them to be moved in to intercept the beam. A ceramic plate
placed upstream of the LIC detector allowed a radiation-
hard videocamera to verify that the beam was hitting the
chambers. The irradiation was produced by proton bunches
of 50 ns length and intensities ranging from 3.0 · 10
9
−
2.2 · 10
10
protons with kinetic energy E
kin
= 1.4 GeV.
Figure 1: Detector setup in the PSB dump line. The beam
direction is from bottom to top intercepting first the ce-
ramic plate, then the LIC and finally the IC.
Figure 2 shows the response of both detectors in a time
window of 1 µs were the charge collection due to electrons
dominates. Figure 3 shows the response in a time window
of 100 µs where the ion collection is observed. Note that in
both cases the LIC detectors collect the charges faster than
the IC. This is consistent with the fact that the mobility of
both ions and electrons in a gas is inversely proportional
to its pressure [3]. The time response of the detectors was
determined as the Full Width at Half Maximum (FWHM)
of the electron induced peak and was measured to be 75 ns
(120 ns) for the LIC (IC) detector.
Figure 4 shows the total number of charges collected
in a window of 700 ns. The response of the IC is lin-
ear with intensity collecting from 0.1 up to 0.6 µC at a
rate of 2.93 · 10
−17
µC/proton. For intensities lower than
Time (s)
0 0.2 0.4 0.6 0.8 1
-6
10×
Voltage (V)
-10
0
10
20
30
40
50
60
70
LIC
IC
Figure 2: Electron induced signal in the ionization cham-
bers for a beam of 8.2 · 10
9
protons.
Figure 3: Ion induced tail in the ionization chambers for a
beam of 7.5 · 10
9
protons.
1 · 10
10
protons, the LIC detector also collects charges
that increase linearly from 0.008 to 0.02 µC at a rate of
2.86 · 10
−18
µC/proton. In this region, the LIC detectors
collect approximately a factor 10 less charges than the IC.
Due to the geometry of the setup both detectors do not re-
ceive the same dose and monte carlo simulations are re-
quired for an absolute calibration factor. With higher inten-
sities a different linear behaviour of the LICs is observed,
where extra charges are collected (larger slope). Previous
measurements with a LIC prototype with filling pressure
0.1 bar showed pulses (in the absence of irradiation) with
duration of a few milliseconds that were attributed to the
formation of sparks. It is suspected that the read out of
large currents favors the formation of such sparks.
Intensity (p)
2 4 6 8 10 12 14 16 18 20
9
10×
Charge (C)
0
0.1
0.2
0.3
0.4
0.5
0.6
-6
10×
2.86e-18 x + 1.32e-09
1.20e-17 x + -5.17e-08
2.93e-17 x + 1.87e-09
Figure 4: Charge collected in 700 ns vs intensity for IC (red
circles) and LIC (blue triangles) detectors.
CNRAD MEASUREMENTS
The CNRad facility is located downstream of the CNGS
target area, where a neutrino beam is produced by hitting a
graphite target with a beam of 400GeV/c protons. The pro-
tons are provided by the Super Proton Synchrotron (SPS)
in two spills of 10.5µs separated by 50 ms. The average
proton intensity per spill was 1.818 · 10
13
protons. Eight
different ionization chambers were located in a metallic
cross as shown in figure 5, in a tunnel 50 m downstream
of the target chamber and perpendicular to the beam di-
rection. Three LIC detectors at a filling pressure of 0.4
bar and two CERN LIC prototypes at filling pressures of
0.1 and 1.1 bar were installed for verification while three
standard LHC ICs were installed for calibration purposes.
The signals were integrated over 40µs via a Current to Fre-
quency Converter (CFC) [4] and sent to the surface elec-
tronics using optical fibers for further processing. The sur-
face electronics consists of a laptop and a test system for
the BLM acquisition cards [5]. The device kept a history
of the signals received and computed twelve running sums
which correspond to the integrated signal in twelve differ-
ent integration windows spanning 40 µs to 83.4 s.
Figure 5: View of the detectors on the metallic cross in
TSG45.
The signals integrated over 40 µs during CNGS extrac-
tions in all 8 detectors are presented in figure 6 for a period
of about a month. We observe channels 2, 4 and 6 (ICs)
recording signals in the order of 100k-120k ADC counts.
The recorded signals were found to be higher for detec-
tors near the tunnel walls and they decrease by 0.5 %/cm
when moving further away from it. This is attributed to
the variation of the neutron flux. The 1.1 bar LIC proto-
type (connected to channel 8) showed signals in the order
10k ADC counts. One LIC prototype at pressure 0.1 bar
recored signals in the order of 700 ADC counts but gave
erratic behaviour with continuous spikes (presumably due
to sparking). Note that the situation became worse at the
end of data taking. Channels 3, 5 and 7 correspond to three
LIC detectors filled at 0.4 bar showing signals in the order
of 2k ADC counts. The sensitivity of the LIC detectors was
studied by comparing the recorded signals with the LHC IC
readings. Table 1 summarizes the charges collected by the
five LIC detectors normalized to the charges collected by
their closest IC neighbour for three integration windows.
The prototype LIC at 1.1 bar shows a roughly constant
ratio. Note that from geometrical considerations and as-
suming a perfectly homogeneous radiation field a factor
30 reduction in the sensitivity was expected. However, as
mentioned before this is not a realistic approximation since
large variations were observed depending on the detector
position. The three LICs at 0.4 bar present a lower ratio in
the 40 µs integration window due to the higher mobility of
the charges at lowergas pressure. This effect is also present
in the response of the LIC prototype at 0.1 bar. However,
for long integration times the result is dominated by a large
leakage current in this particular detector. A factor 3 − 3.4
reduction in sensitivity is achieved by decreasing the filling
pressure from 1.1 to 0.4 bar.
Table 1: Ratio of IC to LIC integrated signals in RS01 (40
µs), RS04 (640 µs) and RS08 (655.3 ms)
P (bar) RS01 RS04 RS08
0.1 152.0 193.0 10.5
1.1 14.0 16.5 16.3
0.4 45.9 58.9 51.2
0.4 44.6 56.8 49.4
0.4 43.9 56.6 50.1
Figure 6: Signals in 40µs vs time.
LHC BETATRON CLEANING AREAS
The response of two CERN LIC prototypes and one LIC
detector were tested against steady-state losses in the beta-
tron cleaning insertion region of the LHC. The three detec-
tors, with filling pressures 0.1 bar, 0.4 bar and 1.1 bar, were
located downstream of a graphite secondary collimator and
below the beam line. In this area, protons from the tails of
the LHC beamsor protons scattered by primary collimators
intercept the collimator jaw producing continuous particle
showers throughout an LHC fill. Figure 7 presents the sig-
nals observed during 5 hours in which an 1.8 · 10
14
proton
beam was injected into the LHC and put into collision. The
signals collected in 1.3 s for the three LIC detectors are
plotted versus the signals observed by the IC during the
same time. Due to the complicated geometry of the setup
some of the detectors are partially shielded from the show-
ers and an absolute calibration would require monte carlo
simulations. However, a very good linearity is observed
between LIC and IC for relativelly high losses. A turn-on
effect is observed for signals lower than 1.0
−4
Gy/s in the
IC were the LIC detectors are not sensitive enough to fully
detect the losses produced.
Signal IC (Gy/s)
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
Signal LIC (Gy/s)
-6
10
-5
10
-4
10
-3
10
BLMEL.06R7.B2I20_TCSG.A6R7.B2
BLMEL.06R7.B2I21_TCSG.A6R7.B2
BLMEL.06R7.B2I22_TCSG.A6R7.B2
Figure 7: Signal in LIC detectors in 1.3 s vs signal observed
in IC detector during the same time.
CONCLUSIONS
Several LIC prototypes, LICs and IC detectors have been
tested under various irradiation conditions. As expected,
the LIC detectors systematically showed lower charge col-
lection efficiency but monte carlo simulations are required
to establish an absolute calibration factor. When irradiating
with proton beams a non linear effect was observed in the
LIC for very large intensities, attributed to the formations
of sparks. The reduction of the signal for lower gas filling
pressure was verified, however, the formation of sparks in
the chamber appears to occur at a higher rate for low pres-
sures. Investigations on gas purity and the introduction of
a mixture of gases to avoid this effect is ongoing.
ACKNOWLEDGMENT
The authors of this document would like to thank Ana-
toli Larinov and Vladimir Seleznezthe from IHEP for their
careful and diligent work in the production of the LICs.
REFERENCES
[1] B. Dehning et al. “LHC Beam Loss Monitoring system de-
sign”. AIP proceedings. Beam Instrumentation Workshop
2002 (P 229).
[2] http://cngs-rad-facility.web.cern.ch.
[3] D.H. Wilkinson. , Ionization chambers and counters, (Cam-
bridge at the University press, 1950).
[4] E. Effinger, 12th Workshop on Electronics For LHC and Fu-
ture Experiments, Valencia, Spain, 25 - 29 Sep 2006, pp.108-
112.
[5] J. Emery, ”Functional and linearity test system for the LHC
beam loss monitoring data acquisition card”. 12th Workshop
on Electronics For LHC and Future Experiments, Valencia,
Spain, 25 - 29 Sep 2006, pp.447-450.
