Status of C9
Prepared for the April 1996 Collaboration Meeting
by David White
- Under-carriage. C9 now rests on an under-carriage which permits movement in both the X and Z directions.
- H9 (hodoscope). This summer, C9 will be refitted with a new rear gas window and a new hodoscope "H9". The window will be made of Kevlar and Tedlar. The hodoscope will consist of 32 segments (16 horizontally by 2 vertically) of 1/4" plastic scintillator. The scintillators will be sloped at roughly 30 degrees from vertical, to match the C9 mirrors. See figure.
- Magnetic shielding tests. One real PM tube was tested in situ, where the field is strongest (almost 400 gauss, parallel to the tube, when the MPS is at full field). For 3/4 field, the PM behaved normally with the nominal shielding. For full field, about 44% of the photoelectrons are lost. When one doubles the amount of shielding material around the tube, this fraction of lost photoelectrons is reduced to 25%. However, C9 prefers to run at less than full field for another reason -- read on...
- More concerns about the MPS field. The transverse momentum kick induced by the MPS field interferes with the focusing of the Cerenkov light onto the Winston cones. For low-momentum pions, C9 sees a virtual target that is displaced from the real target. See figure. Effectively this increases the momentum threshold of C9. This problem is further complicated by the fact that C9 will be further from the target than designed. Therefore a systematic bias is introduced: low momentum positive pions will be lost on one side of C9, and low momentum negative pions will be lost on the other side. The 96 mirrors are glued onto 2 planes -- the mirrors cannot be adjusted individually to counter the effect. For E910, which is running at 1/2 field and a longer distance to the target, the effect is such that C9 is borderline usable. The effect on E852 is difficult to estimate until we align the mirror planes for our target position.
- Gas. If the E910 recovery system works, then we will likely use Freon 114; otherwise we will more probably use C4F10. For a description of the gas system plus the recovery system, see the E910 web pages.
- Trigger rate studies. The reconstructed tracks from 1994 data were passed to SaGen, which has a crude simulation of C9 and H9. The data sample may be biased by DEA cuts, CsI cuts, and cuts on the number of charged tracks. As a trigger device, C9 would demand at least one K+/K- candidate; i.e., a hit in H9 with no hits in any of the C9 mirrors that overlap the H9 scintillator. In addition, a "perfect" RAM is allowed to reject all tracks with less than 2.8 GeV/c. Disclaimer -- the simulation is rather crude, and does not account for effects such as the defocusing of light from tracks with large transverse momentum kick (see above). Therefore the trigger rates with C9 shown below should be regarded as a lower limit.
| Trigger | Typical Rate | Approx. Rate | Approx. Rate |
| in 1995 | with C9 | with C9 plus |
| (per 10^6 beam) | | "perfect" RAM |
| 1-1-1 | 3000 | 45 | under 10 |
| 1-1-1-eta | 300 | 42 | under 10 |
| 1-3-3 | 600 | 186 | under 10 |
| 1-3-3-eta | x | x/3 | under 10 |