1. Resistive Plate Chambers performance with Cosmic Rays
in the CMS experiment
D. Piccolo a) - Laboratori Nazionali di Frascati dell’INFN
on behalf of the CMS RPC collaboration
a) On Leave of Absence from INFN Napoli
Resistive Plate chambers used as dedicated Trigger detector in both Barrel and End Cap:
480 RPC chambers in Barrel (about 4800 m2 of single gaps)
432 RPC chambers in EndCaps (about 3000 m2 of single gaps)
The total volume of gas is 18 m3
Gas flow of the system is 5 m3/h re-circulating in closed loop with a fresh mixture fraction of 10 %
At the end of 2008 a dedicated period of data taking of about one month has been performed by the CMS experiment to test the full system as a whole and to learn as much as possible about detector operations and performance before LHC start-up.
at CRAFT start
at CRAFT end
Good stability in terms of dark current during the data taking period.
Dark current of a complete chamber (about 10 m2 of single gap RPC)
below 1.5 mA on average and not more than 9 mA for any chamber.
Chamber temperature have been stable at about 22 ºC.
CMS RPCs are double gap chambers working in avalanche mode with the following gas mixture:
C2H2F ,2 %
i-C4H ,5 %
SF ,3 %
They are operated at a voltage of at least 9.2 kV
Strip signals are discriminated and shaped into 100 ns LVDS pulses by the front-end board with 220 mV threshold corresponding to about 120 fC of induced signal charge.
RPC barrel
RPC dark current stability
Average Dark current vs time
Barrel muon chamber DT
RPC
+ HV
time
- HV
Drift Tube segment information is used to study RPC performance.
DT segment is extrapolated on adjacent RPC planes (only one plane for the two external layers) searching for RPC hits in a +- 2 strip range centered on segment impact point.
RPC Position measurement capability is studied looking at the difference between DT segment extrapolated impact point and cluster center.
RPC residual studies
RPC trigger synchronization
Bunch crossing
Distribution:
Cosmic Data
Bunch Crossing distribution YB+2
Entries
Mean
RMS
Details of residual distributions as a function of different layers and cluster size
Layers of different sectors have misalignments of the order of 0.3 cm in respect to DTs, not corrected in these plots
Time adjustment to assign the correct bunch crossing to
RPC hits. Cosmic data are used to synchronize the detector
taking into account time of flight and cable lengths
Residual distribution integrated over the full Barrel:
the plot is a convolution of residual distributions
in all layers (with different strip pitch), and with no selection on cluster size.
DT vs RPC trigger timing
RPC cluster multiplicity
RPC Layer
Strip pitch (cm)
Average cluster size
RB1in
2.3
1.62
RB1out
2.5
1.58
RB2in
2.8
1.48
RB2out 3
1.45
RB3
3.5
1.35
RB4
4.1
1.27
DT bunch crossing assignment
RPC bunch crossing assignment
Cluster Size (strips)
RPC of different layers have different strip pitch.
Going from the innermost layer (RB1in) to the outermost (RB4), the average cluster size and the probability to get more than one strip fired decrease.
Synchronization between RPC trigger and DT trigger.
RPC efficiency studies
Y (cm)
X (cm)
Single gap working region
Single gap chambers
Swapped cables
CMS preliminary
Preliminary maximum efficiency distribution for the RPCs of three wheels of the barrel.
The efficiency is measured extrapolating the DT segment on the RPC plane and looking at fired strips in a small region around the impact point.
No optimization of the threshold and of the effective working voltage has been performed up to now, so to have an idea of the distribution of RPC efficiency the measurements are repeated at different working voltage and the maximum efficiency is obtained by a fit to the efficiency vs High Voltage trend.
The tail in the distribution is due to problematic chambers working in single gap mode or affected by swapped cables in the readout that have been fixed during the shut down.
Systematic effects due to the fact that the distribution is obtained with cosmic muons and not with collision muons are under investigation and tend to lower the efficiency respect to expected value, moreover no fiducial regions are selected to remove the edge of the RPCs.
RPC detector efficiency vs impact point measured extrapolating the DT segment on the RPC chamber. The lower efficiency points in step of 10 x 10 cm2 are due to the dead regions induced by spacers.
The efficiency reduction is more visible for y coordinate around 55 cm where, due to the chamber construction, the RPC works in single gap.