Progress in understanding and reducing dead-time in the Radphi trigger

Many aspects of the Radphi experiment make use of proven technology. For example, the performance of the lead glass calorimeter has been well understood from previous experiments using a similar detector. The major area in which Radphi is breaking new ground, from the point of view of technique, is in the high photon beam intensity at which it is foreseen to operate. This aspect of the experiment has been viewed with some skepticism by some with experience with tagged photon beams, and it is fair to say that we have yet to prove that we can make effective use of the beam at the intensities called for in our proposal. Considerable progress towards demonstrating this capability has been made since we saw our first beam in the summer of 1997, primarily in the area of shielding. During our 1998 commissioning we were able to raise the beam intensity to projected operating levels for the first time, and observe the factors limiting effective operation at these rates. Several of these were relatively easy to fix during the run, others will require some study offline to evaluate the most cost-effective way to deal with them. In this study I present the measured rates at a variety of beam intensities and compare them to a calculation that relates the rates to one another and a small set of basic parameters describing our trigger. Three parameters emerge as the limiting factors in the capability of our present setup to run at high rates: data acquisition dead time, trigger processor (level 2) dead time, and the width of the CP veto window.