Three straw tube arrays will be located as shown in Fig. 
.
The demands on the granularity of the arrays are mainly driven by the
desire to minimize pile-up of multiple interactions during the maximum
drift time in the tubes.
With a beam rate of 10 MHz and a 10% interaction length target,
events giving hits in the spectrometer occur at a rate of 1 MHz. About
10% of these events are central collisions, which yield the
hit multiplicities in Table . The occupancies for
less central events are similar. The use of conventional argon plus
ethane gas mixture results in maximum drift times of 40 nsec in arrays S1
and S2, and 80 nsec in S3, corresponding to pile-up rates of 4% to 8%.
We plan to investigate the feasibility of using faster gases, such as
CF 
[65], which could reduce the pileup rates by a factor of two.
Occupancies of a few per cent and acceptable pile-up can be achieved with straw tube dimensions similar to those described in the literature [66, 67, 68, 69].
Each straw tube array is comprised of six layers, as illustrated
in Fig. , with 2 layers each of x (vertical), u and v (stereo) tubes.
Figure: Contents of one straw tube array.
The overlapping double layers are needed to cover the dead region at tube boundaries.
To save on electronics costs, the straw tubes in S1 and S2 are read out as yes/no hits. In principle the spatial resolution, both in horizontal and vertical directions, could be improved by measuring drift times. The resulting improvement in momentum resolution is marginal, since Coulomb scattering dominates for most of the rare objects of interest.
Design parameters for S3 are shown in Table .
Table: Straw Tube design parameters.
