Use of CCDs in the Time Projection Chamber

Abstract

The Time Projection Chamber produces 3-dimensional information about the multiple tracks of reaction products produced by e+ e-collisions at the center of the chamber. Two dimensions (r,ø) result from processing signals from proportional wires and induction pads (a total of 17,000 signals) mounted at the end planes. The third (axial) dimension is obtained by measuring the drift time of ionization from the particle tracks to the end planes. To achieve this, all wire and pad signals (suitably shaped to about 500 ns width) are stored in Charge Coupled Devices (CCDs) which, at any moment, hold 45.5 μs of signal history clocked in at a sampling rate of 10 MHz. About 40 μs after a master trigger, the history is frozen by stopping the clocking action, and a slow clock mode (20 kHz) is initiated to read out the stored samples which are then digitized (9 bits) and processed by an on-line computer. A total of 17,000 channels × 455 samples (each digitized to 9 bit accuracy) is stored in the CCDs (i.e., about 70M bits) for each event. This paper describes the pulse shaping philosophy, the design of the CCD units, the drivers and the testing procedures.