Introduction

From data taken during parasitic running in the summer of 1998, the Radphi experiment has demonstrated its capability to trigger on and reconstruct multi-photon final states up to the mass of the $\phi(1020)$. The most significant obstacle that we encountered to our readiness for our first physics run is the rate limitations of our trigger and data acquisition. The rates expected for the looked-for $\phi$ radiative decays are very low. When running at $5\cdot 10^7$ $\gamma$/s, near the highest photon beam intensity compatible with tagging, we expect of order 1 event per hour per rare decay channel. At this beam intensity, the total rate of hadronic events coming from the target is calculated to be about 170KHz. Scaling measurements taken during 1998 parasitic tests up to this beam intensity yield a total rate of $8.5\cdot 10^7$/s charged particles within the acceptance of the detector.

The experiment is looking for final states of $\mbox{\rm n}\gamma$ plus a recoil proton, so most of the background rate can be excluded fairly easily. The basic trigger design is to look for a single charged track at large angle that corresponds to the kinematics of the recoil proton, and then look in the forward calorimeter for a large energy deposition. The trigger has gone through several revisions, based upon experience gained during parasitic running, and we now have a design capable of collecting data efficiently at our design beam intensity of $5\cdot 10^7$ tagged $\gamma$/s. Following a brief commissioning of the new trigger hardware, we would be ready now to take physics data with the existing detector as soon as we can get a place on the schedule.

The interval from the time that an approved experiment is ready until the time it can receive beam as primary user is presently a year or more. In view of this, the Radphi collaboration decided to proceed immediately with its plans to upgrade the barrel region of the detector. The upgrade extends the photon coverage from the $28^o$ back to $90^o$ in the lab by adding a barrel calorimeter, and also increases the angular acceptance of the recoil proton trigger. This was accomplished by replacing the recoil proton detector [RPD] shown in Fig. 1 with the barrel calorimeter and scintillator assembly from the former Jetset experiment (PS-202 CERN-LEAR) shown in Fig. 2.

Figure 1: Drawing of the Radphi recoil proton trigger scintillators that are being replaced by a new barrel assembly during the barrel upgrade.

Figure 2: Drawing of the new barrel assembly being installed in the Radphi experiment during the barrel upgrade.

This upgrade significantly improves the quality of the experiment, increasing the acceptance for $\phi$ radiative decay channels to the $a_0$ and $f_0$ channels by more than a factor of 2, and significantly improving our ability to discriminate these reactions from background. It also represents a major change to the configuration under which we have run in the past, and so will require some additional commissioning time.

This point was noted by the Jefferson Lab review committee during our readiness review that was held on March 10, 1999. We asked to be placed on the schedule nevertheless, based upon the fact that the upgrade mainly adds new capability without affecting present function (apart from the trigger) and that this apparatus has been used successfully in a prior experiment. The committee went along with our request to place us on the tentative schedule for the year 2000, but made some conditions related to commissioning the upgraded detector during summer 1999 parasitic running. In particular, they requested that we calculate the rates that we expect in the detector trigger elements under the new configuration in advance of test beam. Subsequent measurement of rates during 1999 parasitic running in agreement with these predictions would confirm our claim that we understand the apparatus and operating conditions of the experiment.

The most important risk in the upgrade, as pointed out above, is the change that it makes to the trigger. Instead of covering a narrow annulus between $40^o$ and $60^o$, the barrel scintillators now subtend the region from $32^o$ to just beyond $90^o$. This increased acceptance, particularly towards forward angles, may be expected to result in an increased rate of level-1 triggers and introduce new dead time to the system. The major conclusion of this study is that the total level-1 rate is expected to decrease, rather than increase, in the new configuration, and at the same time running with higher acceptance in the recoil proton by a factor of 2.