SSD tracking

Finding tracks with the SSD (see spec:ssd) detector is performed in three stages. First, hit strips are grouped into clusters of hits. Second, projections of clusters are found in each of the three measurement directions. Finally, these projections are combined into tracks.

Clusters of up to three hit channels are grouped into hit clusters. Using the ADC information, it is possible to determine if one or two²⁰ traversing particles created a single cluster by comparing the total ADC counts present to the number expected for a single Minimum Ionizing Particle (MIP). For multi-strip, single-MIP hits, enhanced position resolution is achieved by fitting the ADC values in the cluster with a pulse height sharing algorithm.

Projections are found in each of the three SSD measurement directions by taking all combinations of hits in the four SSD planes of a given direction and fitting those hits to a straight line. Projections are rejected if $\chi^2$ per degree of freedom () > 3.0. Projections must contain hits from at least three of the four planes; hits are allowed to be shared among 3-plane projections and in the first plane of 4-plane projections.

Tracks are formed by intersecting all combinations of three projections (one from each view) and requiring that $\csdof < 8.0$ for these combinations. Shared projections among tracks are arbitrated based on the lowest and groups of tracks with nearly identical parameters are reduced to a single equivalent track.

In E687, the spatial resolution of a track in the high resolution region of the SSD was
$$\begin{align}\sigma_x &= 11.0~\micron \sqrt{1 + \left(\frac{17.5~\gevc}{p}\right... ...y &= 7.7~\micron \sqrt{1 + \left(\frac{25.0~\gevc}{p}\right)^2} \, , \end{align}$$
while the resolution of a track entirely in the low resolution region of the SSD was about twice as large. In FOCUS, this resolution is slightly better because the pulse height sharing algorithm is being used. The constant term results from the resolution of the detector while the momentum dependent term is dominated by in the silicon planes. The resolution in y is better than that in x because all three silicon planes provide y information, while only two provide information in x.

The Target Silicon, or TSSD, is not used in finding the initial tracks in the target region.