Drift Chamber Tracking with a Vlsi Neural Network

Government nor any agency thereof, nor any of their employees, makes any warranty, ezpress or implied, 07 assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, OP process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. We have tested a commercial analog VLSI neural network chip for finding in real time the intercept and slope of charged particles traversing a drift chamber. Voltages proportional to the drift times were input to the Intel ETANN chip and the outputs were recorded and later compared off line to conventional track fits. WC will discuss the chamber and test setup, the chip specifications, and results of recent tests. We'll briefly discuss possible applications in high energy physics detector triggers. 1. Introduction Neural networks implemented in VLSI offer the capability of very fast pattern recognition. The first generation of such chips can process an entire multilayer network in microseconds and could provide on line processing. In high energy physics experiments. an important on line application of VLSI neural networks would be the selection of events written to tape. (Here an event refers to a collision of two particles and the subsequent signals induced in the detectors from the reaction products.) This on line selection process, called triggering, is a crucial part of an experiment. Collisions of particles can occur at rates of hundreds of kIIz and the rate will include background interactions such as stray beam particles striking beam pipe material. Usually the most interesting events occur only at a small fraction of the total rate and also the recording the detector signals onto tape for later off line processing is typically limited to tens of Hz. So the trigger system must select only those types of events of greatest interest, while efficiently rejecting the backgrounds. The first step or level in a typical trigger system makes cuts on simple raw signals such as the amplitude of the signal from a calorimeter. The second level of the trigger transforms the raw …