D : perpendicular distance from the beam axis to first strip
Z : position along the beam axis
Phi : rotation around the beam angle
Theta : rotation around horizontal axis (X)
Eta : rotation around vertical axis (Y)
Yoff : vertical offset of forward modules
The stereo angle of all barrel modules is compatible with 40 mradians (as measured by the difference in Eta of the independently aligned planes of each modules). The variation between two alignments of the value obtained is some 5 mradians. Attempts are going on to improve this and use H8 as a metrology machine. The effect of the magnetic field on the alignment is also under study.
In the case of forward modules there is an ambiguity between the vertical offset and the angle Eta (both affect the effective pitch). These values can thus not be known absolutely. Besides, of all angles only relative values can be obtained.
May 2000
All May runs with telescopes (up till run number 820) use the alignment of run number 802. Only 2X2 telescope planes were available, resulting in a maximum of two space points to reconstruct the track. This makes it very hard to distinguish between real tracks and ghost tracks between noise hits. The fake tracks can be (nearly) eliminated from the sample by a simple cut on the deviation of the track with respect to the beam axis (dX/dZ and dY/dZ). These numbers are available in the DST. Rejecting all tracks with either of the deviations larger than 0.00025 yields good results. See Gabriela's page.
June 2000
Alignments were done for the four June 2000 scans:
Note: Runs 1912 to 1924 form part of scan C
During the June runs, the telescope planes were not functioning stably. The
efficiency is varying strongly and generally rather low. The performance
seems to become worse when the magnet is ON, an effect which is not
understood until now. As a consequence the mininum number of space points
that defines a track had to be lowered to 2, resulting in the ambiguities
explained for the May alignment. A simple cut rejecting the most divergent
tracks can reduce the effect on efficiency to within the errors without
loosing too much statistics. See Gabriela's page.
Reasonably good alignments were obtained in the four cases. As an example, the figure shows residuals for single hit clusters (red) and double hit cluster for Anchor 2 at 1 fC. Gaussian fits give 23.2 and 10.6 mu, respectively.
August 2000
Alignments are available for 13 scans at different angles and/or magnetic field. The three first scans had to be realigned as halfway through the scan the granite table was moved to get the full beam in one detector. Total 16 alignment files:
December 2000
The alignment of all runs from run II was done using run 3507.
Three pairs of new (smaller, but very efficient) telescope planes were used. The order of modules and telescopes is: mod0 : mod1 : tel0 : mod2 : mod3 : tel1 : mod4 : mod5 : tel2 : mod6.
The situation at KEK is rather different from that at H8, as the beam energy is much lower (4 Gev/c protons). The probability for a particle to scatter while traversing the test beam box is now not negligible (although attempts were made to minimize scattering by running without cover plates). Particles that undergo scattering while within the telescope (between tel0 and tel2) can be removed efficiently by cutting on the track quality parameters (chi-squared, dx/dz and dy/dz). Particles that scatter outside the telescope but inside the test beam box are much harder to remove, resulting in a slightly worse residual for the outer modules (mod0, mod1 and mod6). Finally, the OK flag - set to 1 whenever either of the reference planes is efficient - is very effective in removing bad tracks. Using a combination of the above recipes, a large sample of clean tracks can be obtained, allowing for a high precision study of interstrip effects in the 4 central modules (mod2,3,4 and 5). The precision for modules 1 and 6 are slightly worse, but does not introduce a significant error in the efficiency determination.
Last update September 12, 2000