We have found great improvement of the efficiency by setting cuts on the values dx/dz and dy/dz of the tracks. The plots in this page show the distribution of these values for different runs. The results are worse for those runs with magnetic field and angle, such as run 1692.

Our aim has been to find the maximum efficiency achievable ( for the best reconstructed tracks ) and to set the cuts so that the efficiency is reduced no more than 1% from its maximum value, without loosing too much statistics.

For most of the runs, this condition is satisfied requiring dx/dz and dy/dz to be less than 0.002 (even higher values give good results). This value keeps the efficiency close to the maximum by loosing only a few events.

Only for runs with magnetic field and angle, we recommend a lower value. A cut on dx/dz and dy/dz less than 0.0002 satisfies the condition above, but loosing 40% statistics. For higher values of the cut, the number of events increases slowly while the efficiency keeps decreasing at a high rate (shown for run 1692).

If higher statistics is necessary, it can be raised to 75% by setting the cut to 0.001, which reduces the efficiency 3% from the maximum value.