The ALICE TPC Inner Readout Chamber: Results of Beam and
Laser Tests
U. Frankenfeld 1, G. Augustinski 1, P. Braun-Munzinger 1, H. Daues 1, J. Fiess 2,
C. Garabatos 1, J. Hehner 1, M. Ivanov 1, R. Renfordt 2, H. Sann 1, H.R. Schmidt 1,
H. Stelzer 1 and D. Vranic 1
1
GSI Darmstadt, 2 University of Frankfurt
The ALICE TPC readout chambers, – based on the design
of the NA49 and STAR TPC's –, are optimized to be able to
handle the expected high particle load at √s=5.5 ATeV
Pb+Pb collisions. Up to now TPC readout chambers have
not been operated at high particle load and at high gain; it is
thus a prioi not fully evident whether they can be operated
stably at all.
current was reaching up to 80 A during a spill, which is
significantly above the expected ALICE value. During the
test the chamber operated stably even at the highest loads
and no signs of deterioration in performance has been
observed.
Figure 2
Online picture of a single laser track crossing
the chamber at an angle of about 5° relative to the 31 pad
rows.
Figure 1
Signal on an individual pad at 1280 V.
“Negative” signals levels due to baseline shifts are readout
from the ADC as zero. The insert shows the occupancy
distribution for three different settings of the anode HV.
An inner readout chamber, taken from the ongoing
chamber series production, has been exposed to secondary
particles from a GSI heavy-ion beam 12C (800 AMeV). The
particle rate, – controlled by varying the target thickness
and/or the beam intensity –, was adjusted to create typical
ALICE conditions, i.e., an occupancy in pad-time up to 50%
and an average anode current ≥ 10 A. Figure 1 shows a
typical time distribution of signals, – produced by particle
tracks of one event –, arriving at one of the readout pads.
The anode voltage is 1280 V, which gives the nominal gain
of 2  104. This comparatively high gain is necessary to
achieve a good signal/noise ratio (S/N >20).
The insert of Figure 1 gives the occupancy distribution for
signal above threshold. As can be seen the occupancy
reaches values up to 50%. The corresponding average anode
Besides the behavior under high load, design parameters
like pad response function (PRF), position resolution, gating
efficiency, noise and cross talk have been investigated
employing a UV laser to generate well defined tracks in the
chamber drift volume. As an example, the image of a laser
track in the chamber is shown in Fig. 2. The intensity of the
laser has been adjusted such that a minimum ionizing
particle (MIP) is simulated. The image reveals that, – as
desired –, 2-3 pads respond to a hit, which is sufficient to
determine the track position with the desired spatial
accuracy at low occupancy.
The PRF is measured by moving the laser beam relative
to a (fixed) reference pad, whose pulse height is recorded as
a function of the position of the laser beam and normalized
to the sum of all pads in the region of the track. The result,
as shown in Figure 3, has slightly non-gaussian tails, which
are expected based on the definition of the PRF [1]. Thus the
gaussian fit to the data has been done only for the central
part of the distribution. At the same time the data is fit with
the single parameter Gatti-formula [2], which describes the
PRF correctly. The contribution of the diffusion is estimated
by measuring the PRF for several drift fields and
extrapolation to infinite field, i.e., drift time equal zero. The
resulting PRF, – contributions from laser width and
diffusion properly subtracted –, is 2.06 mm and is in very
good agreement with the estimated value of 2 mm [1].
that no intrinsic chamber property limits the achievable
position resolution.
The determination of the gating efficiency is based on the
measurement of the laser intensity in the chamber via a
reference diode. The magnitude of high intensity laser
beams cannot be determined by the TPC itself due to
saturation of the electronics. The strategy to measure the
gating efficiency is thus as follows: an intense laser beam
corresponding to 2104 ADC channels is sent through the
chamber with gate closed; any residual intensity seen on the
pads should be due to gating inefficiency. The insert in
Figure 5 shows the quadratic response of the chamber vs the
reference diode. Deviations from the quadratic behavior
starts at 500 ADC channels and indicates the onset of
saturation in the preamp.
Figure 3
Signal height in a given pad as function of the
laser beam position relative to the pad yielding the pad
response function. The solid line is a gaussian fit to the data
in the region from 4 to 12 mm, while the dashed curve is a
fit employing the Gatti-formula.
Figure 5
Gating efficiency: the dashed line indicates the
signal with gate open, the full line shows the remaining
signal with gate closed. The insert shows the output of the
reference diode vs. the response of the chamber.
Figure 4
Width of the residual distribution as function
of the pulse height. The error bars indicate fluctuations of
the laser intensity.
The position accuracy is quantified employing the
“residual” method, i.e., the deviations from a straight line fit
to the track. The obtained resolution depends on the signal
height and should scale like 1/√N, as verified experimentally
and shown in Figure 3. For a MIP signal one expects an
azimuthal position resolution of about 400 m. This value,
however, represents the resolution without deterioration due
to diffusion, i.e., reflects geometrical chamber properties as
well as noise and statistics only. It should be noted that the
numbers given in the TPC Technical Design Report [3]
include the average diffusion due to a drift path of up to 250
cm and are of the order of 1000 m. Thus, the test proves
Figure 5 shows the measured signal - about 10000 events
– on the pads while the gate is closed (black line). The
dashed line indicates where to expect a signal in case of
gating inefficiencies. The measured signal is below 0.02
ADC-Channels. From this measurement an upper limit of
the gating inefficiency, i.e., the signal ratio with gate closed
and open, is deduced to be < 10–6.
References
[1]
[2]
[3]
Alice Internal Note, ALICE-INT-2002-030 and
references therein
E. Gatti et al., Nucl. Inst. and Meth 163 (1979) 83
TPC Technical Design Report, CERN/LHCC 2000001