Assembly Plans and Expected Performance

Final assembly is in progress at the time of this writing for most of the silicon system components described above. Delivery of the silicon sensors and assembly and testing of hybrids was completed by mid 2000, and construction of the ladders was completed shortly thereafter. Assembly of these ladders into barrels is currently in progress. A partially configured ``fourth'' barrel including components of all three subsystems was installed into the CDF central detector in the summer of 2000 to provide a test of hardware performance and to exercise the data acquisition and trigger systems. This test barrel will be replaced by the full silicon system after applying any lessons learned from this early installation.

The expected impact parameter resolution of the full Run II CDF tracking system for minimum-ionizing particles is shown in Figure 10 versus transverse momentum of the normally-incident track. The grey shaded region shows the range of estimated resolution without the innermost layer, depending on whether or not the track passes through the region of the SVX II that contains its hybrid readout electronics. The line shows the improved resolution provided by the innermost layer. Without misalignments, the estimated intrinsic impact parameter resolution should improve from an average of $\sigma_{\rm IP}=(9\oplus50/p_{\rm T})~\mu$ m for tracks at normal incidence without Layer 00 to an average of $\sigma_{\rm IP}=(6\oplus25/p_{\rm T})~\mu$ m with the addition of this layer[10], for $p_{\rm T}$ in GeV/c. The values shown in Figure 10 include 10 $\mu$ m in quadrature as the estimated contribution of mechanical misalignments to the expected final resolution.

**Figure 10:** Calculated impact parameter resolution at the location of the beam line for the CDF II silicon system.
$\begin{figure}\centerline{\epsfxsize=132mm \epsffile{cdfii_ipresvspt_afs.eps}}\end{figure}$

Physics simulations of the full system show that the number of tracks and the signal-to-noise in displaced-vertex tags are also improved. The resulting efficiency for finding displaced vertices in

jets in top events is estimated to be 49% for single tags and 25% for double tags[10]. These numbers, which include geometric acceptances and are averaged over all top events, compare very favorably to the corresponding values of 25% and 8.0% for single and double tags with the SVX $^{\prime}$ vertex detector in Tevatron Run I, respectively.