Previous page	Up to the index	Next page

The global placement accuracy of the MVD is a tricky issue. We would like to specify a position relative to the interaction piont. The original vertex finding algorithm operated with the assumption that the transverse beamsize would be negligable on any scale of interest, and that the beam would always be steered to be in the center of Phenix, and therefore exactly on the MVD z-axis. The plot below (taken from John's note) shows that, with these assumptions, the algorithm is efficient in finding the z-vertex even when the vertex is several hundred microns off-axis..

There are 3 contributions to the off-axis component of an event vertex:

• bunch spread
• mean fill position
• central magnet position

Even though once a fill is established, the mean beam position is steady, where this mean is can change from fill to fill, by as much as a few hundred microns [confirm this].

The MVD is mounted to the central magnet. When the magnet is rolled out for service, and back in again, it may be displaced relative to it's formet position by as much as 1-2mm. Even though the MVD mount on the magnet allows for repositioning, just finding out where the MVD is relative to the nominal beam position is difficult. A simple survey is not likely to give an answer closer than 0.5mm.

Therefore, we need to make the vertex-finding algorithm tolerant to large (few mm) off-axis vertex displacements. In the process, we will produce the x and y position of the vertex relative to the MVD.