TRACKER

The Tracker has continued to operate with excellent performance during this first period with 7 TeV collisions. Strips operations have been very smooth. The up-time during collisions was 98.5%, up to end of May, with a large fraction of the down-time coming during the planned fine-timing scan with early 7 TeV collisions. Pixels operations are also going very well, besides problems related to background beam-gas collisions where the particles produced generate very large clusters in the barrel modules. When CMS triggers on these events, the FEDs affected overflow and then timeout. Effort was mobilised very quickly to understand and mitigate this problem, with modifications made to the pixel FED firmware in order to provide automatic recovery.

With operations becoming more and more routine at P5, Pixels have begun the transition to centrally attended operation, which means that the P5 shifters will no longer be required to be on duty. The strip-Tracker is also planning to make this transition at the end of June. Operation and monitoring of the Tracker subsystems will then be done by a combination of the central shift crew and on-call Tracker experts. Data-quality and detector performance will continue to receive proper attention from the central-DQM and Tracker offline shift crews, supported by the Tracker DQM experts.

The Tracker infrastructure continues to perform very stably. Only seven problems required intervention in the strips power system during the first five months of operation in 2010, equivalent to a failure rate of ~1% of power supply units per year. In contrast, there were three times more interventions during the last five months of 2009.

The cooling systems all continue to run stably, with no significant, unplanned down-times so far in 2010. There have been a few small increments in the leak rate of the SS2 (silicon-strip-2) system in 2010, all coinciding with interventions on the system. These increases are under investigation and detailed leak rate measurements are being made in order to assess the contributions from specific parts of the system.

The combined strips and pixels DCS system was recently modified to provide a ‘locked-off’ state. This was put in place to satisfy the requirement for centrally-attended operations, specifically to avoid that the detector is switched on whilst the LHC is in an unstable mode of operation. Unlocking the locked-off state requires permission from the strips and pixels on-call experts. The locked-off state is also expected to assist in avoiding extra, unnecessary thermal cycles of the detectors. The default operation condition of the Tracker is now ‘Standby’ (LV on, HV off), though prudence is still maintained with regards to machine development and LHC conditions considered to be unstable. In particular, the Tracker DCS will automatically switch off (and lock) the detector if the Accelerator Mode goes to ‘Machine Development’, or Beam Mode goes to ‘Inject and Dump’ or ‘Circulate and Dump’.

by K. Gill