next up previous
Next: Introduction

Also available as

  • latex file
  • dvi file
  • postscript file(104 KB)


  • A Fully Software-Programmable Pipelined Trigger-Processing Module

    G. D. Barr, F. Formenti, W. Iwanski tex2html_wrap_inline185 , P. Kapusta tex2html_wrap_inline185 , H. Wahl and M. Ziolkowski tex2html_wrap_inline185
    CERN, 1211 Geneva 23, Switzerland

    Abstract:

    We describe a general purpose 40MHz pipelined trigger processing board containing six look-up tables and the logic needed to programmably route the signals between them. A series of identical boards are connected in cascade in the NA48 CP-violation experiment at CERN, accept digital inputs generated from calorimeter signals and are used to perform the different computations necessary for the trigger reconstruction. The board architecture is designed to be flexible so that a single basic hardware design can perform all the different algorithms we require and so that the trigger algorithms can be adapted if needed. The computation latency of the combined network of eight boards is approximately 3 tex2html_wrap_inline191 s. We present examples of some of the functions which have been implemented. We discuss the automatic reprogramming software which has been developed so that a new design can be implemented easily.





    Paolo Calafiura
    Thu May 2 16:58:44 MET DST 1996