CMS Bulletin

MUON DETECTORS: RPC

During data-taking in 2010 the RPC system behaviour was very satisfactory for both the detector and trigger performances. Most of the data analyses are now completed and many results and plots have been approved in order to be published in the muon detector paper. A very detailed analysis of the detector efficiency has been performed using 60 million muon events taken with the dedicated RPC monitor stream.

The results have shown that the 96.3% of the system was working properly with an average efficiency of 95.4% at 9.35 kV in the Barrel region and 94.9% at 9.55 kV in the Endcap. Cluster size goes from 1.6 to 2.2 showing a clear and well-known correlation with the strip pitch. Average noise in the Barrel is less than 0.4 Hz/cm² and about 98% of full system has averaged noise less then 1 Hz/cm². A linear dependence of the noise versus the luminosity has been preliminary observed and is now under study.

Detailed chamber efficiency maps have shown a few percent of chambers with a non-uniform efficiency distribution. This could be a clear sign of chambers that are not working in a plateau region. A special calibration was planned, during the first day of 2011 data-taking, to perform an efficiency scan as function of the high voltage in order to determine in a precise way the effective working point of all the chambers.

During the winter shutdown many problems have been fixed and in particular the power system has been completely recovered and re-calibrated (more than 1000 channels). Few cable swaps, found with the detailed 2010 data analysis, have been fixed. About 98.5% of the electronic channels are properly working. The present status of problematic chambers is: 6 chambers out of 912 are “disconnected” and 28 are working in “single-gap mode”. Most of the problems are due to broken high-voltage connectors or electronic failures that cannot be solved until the long shutdown foreseen for 2013-2014.

At the same time many other aspects of the RPC project have been improved in the last few months in view of the 2011 run. A new set of efficiency tables has been included in the Monte Carlo simulation in order to have a better performance modelling of the system. The 6 dead and the 28 single-gap chambers will be described as well in the MC.

A new trigger algorithm for the barrel region has been deeply studied and now implemented. Muon candidates are generated if at least three out of six (it was four out of six) chambers are fired, following a selected combination of chambers in order to keep constant the trigger rate. Another important improvement has been the development of a new procedure to configure the electronic front-end boards. All the thresholds and widths are now loaded automatically by the database and provide a very easy way to fine-tune the thresholds and improve the speed of the configuration. This was the last step to have the full RPC detector automatically configured and this will certainly further reduce dead time compared to less than 2% from 2010.