Previous page	Up to the index	Next page

Of course, after the MVD is fully assembled, we can measure where every panel is, and feed that information to the vertex-finding software. In addition, a calibration program can take a set of known good events, and one by one 'move around' each panel to maximize its contribution to the found vertices, building up a position correction data base. (This may not work for rotations)

Using these calibrations, the algorithm will then perform as if each panel is perfectly placed, and the maximum vertex-finding efficiency will be recovered.

However, it is possible that at some point in the future, we might want to have a very simple vertex-finding algorithm that is small enough to fit in the DCM FPGA, such that we can deliver a vertex for the Level-2 trigger. In that case it is unlikely that we have room or time for calibration constants. This compact algorithm must assume that all panels are in their nominal positions.

The mechanical tolerances we specify in the following pages will therefore assume that we have to live without after-the-fact position calibrations.