next up previous
Next: Bibliography Up: note Previous: Noise in the other

Summary and conclusions.

VSAT:

The VSAT Bhabha sample has been used to estimate the probability of an off-energy electron in VSAT on top of a genuine physics events such as an untagged $ \gamma$$ \gamma$ event. The statistical precision of this measurement is unbeatable. In an independent analysis a muon sample was selected and the probability to have an off energy electron in VSAT calculated. The result agreed with the VSAT Bhabha measurement within large statistical errors. Finally, a STIC Bhabha sample was used. All features of the background were the same in the STIC and VSAT Bhabha samples. Comparing the probabilities in the four modules by taking the ratio $ \cal {P}$VSAT - $ \cal {P}$STIC/$ \cal {P}$VSAT for Emin=20 GeV -0.02$ \pm$0.01, -0.22$ \pm$0.03, +0.03$ \pm$0.01 and -0.12$ \pm$0.02 are obtained, i.e. a significant difference in the measurement of probability in the outer modules.

The best way of removing the background is by a cut on the measured y-coordinate since the off-energy electrons are concentrated in the horizontal plane.

STIC:

For STIC, a Bhabha sample has been used to measure the background of off-energy electrons. It is limited to electrons with an energy larger than 10 GeV and has therefore been supplemented by an analysis of random triggered events with which the low-energy off-energy electrons can be studied. At 10 GeV the difference $ \cal {P}$Bhabha - $ \cal {P}$Random/$ \cal {P}$Bhabha is -0.23$ \pm$0.15 .

The most effective way of removing the background is to discard any STIC showers with a polar angle less than $ \sim$ 3o.

FEMC:

The probability of a FEMC shower with energy above 0.5 GeV is sizeable (4.6%) but it drops off quickly with energy and for Emin > 2.5 GeV there is no need to take the detector noise into consideration (except in problematical analyses like single photon analyses which select noise events).

HPC:

The energy spectrum due to noise is very steep and with an energy cut of Emin > 0.5 GeV the probability to have a noise shower in an event is at the level of 0.05%.

HAC:

The hadron calorimeter is noisier than the electromagnetic calorimeters and certain noisy areas can produce showers with energies of up to 10-15 GeV. However, by a cut of Emin > 2.5 GeV the probability of a noise shower is reduced from 5% (without the cut) to 0.3% (with the energy cut).


next up previous
Next: Bibliography Up: note Previous: Noise in the other
Andreas Nygren
1999-11-17