next up previous
Next: Summary and conclusions. Up: note Previous: Random triggers

Noise in the other calorimeters.

The other DELPHI calorimeters have also been studied by using the random triggered sample. At angles above STIC the calorimeters do not see any of the off-energy electron background. Instead they suffer from noise showers and occasional showers created by cosmic rays. In the 1999A data, a noisy area in HCAL which created high energy showers was observed, but was removed in the 1999B processing (Figure 17). The left plot in Figure 18 and Table 9 gives the probability of a noise-shower in different calorimeters as a function of a cut on the shower energy.

Figure 17: The theta-phi distribution of noise showers in HAC in the 1999A data (left) and the 1999B data (right). The middle plot shows the energy distribution of the peak in the theta-phi distribution.
\begin{figure} \begin{center} \mbox{\epsfxsize 5.cm \epsfbox {hac1.eps}} \mbox... ...hac2.eps}} \mbox{\epsfxsize 5.cm \epsfbox {hac3.eps}} \end{center}\end{figure}

Figure 18: Left: The probability of a shower in different calorimeters as a function of acut on the shower energy. The data used were triggered by the random trigger and come from the 1999 B-processing. Right: Noisy areas in the FEMC calorimeter during the 1999 data taking.
\begin{figure} \begin{center} \mbox{\epsfxsize 7.cm \epsfbox {noise.eps}} \mbox{\epsfxsize 7.cm \epsfbox {femc_noise.eps}} \end{center}\end{figure}


Table: 9 The probability of a shower with energy higher than Emin in HAC, FEMC and the HPC.
Emin [GeV] $ \cal {P}$HAC[%] $ \cal {P}$FEMC[%] $ \cal {P}$HPC[%]
0.1 5.15$ \pm$0.18 4.47$ \pm$0.17 3.12$ \pm$0.14
0.5 2.21$ \pm$0.12 4.42$ \pm$0.17 0.05$ \pm$0.02
2.5 0.25$ \pm$0.04 0.05$ \pm$0.02 0.04$ \pm$0.02
5.0 0.10$ \pm$0.03 - 0.03$ \pm$0.01


The probability plots in Figure 18 are of course only useful for analyses which does not select events based on energy in the calorimeters. If one takes the FEMC as an example, the probability of a noise-shower with energy larger than 2.5 GeV on top of a physics events is completely negligible. If on the other hand single photon events are selected by triggering on energy in FEMC and by not requiring any signals in any other DELPHI detectors (which would confirm the event to be a genuine physics event), events caused by noise in FEMC are selected. In this way one can find hundreds of single photon events in FEMC caused by fake showers. This is illustrated in the left plot of Figure 18 which shows noisy areas in the 1999B FEMC data. In this plot, only noisy areas producing showers larger than 2.5 GeV which survive the Margoni offline noise algorithm are included.

next up previous
Next: Summary and conclusions. Up: note Previous: Random triggers
Andreas Nygren
1999-11-17