Data from the current right septum setting shows similar feature as in the second right septum setting. However, if looking at Figure 1 (#23151), VertexZ correlates with tg_th, because reconstruction matrix just thinks all events bending same angle (about 6.85deg) when they go through septum. But in current case, up septum bends less, bottom septum more. Then events going through up septum are reconstructed as if they are from smaller Vz. This results in the spreading of Vz. So probably should not cut on Vz before tg_th and tg_ph calibrations. Without Vz cut, this septum current (711.4A, prediction value from old formula) gives a good acceptance, and center hole positions comparing to other septum current scan runs. Thus, we decided to use this current in following runs.
Figure 1
Figure 2