Corrections to CTD hit position

The position measurement of CTD hits takes the form of a drift distance in the Planar Drift Approximation. This approximation is very effective but the drift is not in a plane with constant velocity, and corrections to drift distance are required. Corrections for the following effects have been implemented at the beginning of the track fit:

Angle of track from drift plane (PsiPrime)
Lorentz angle dependence on position
Drift in z (Lorentz angle out of x-y plane)
Drift velocity dependence on position
Distance near end of drift cell
Distance near sense wire
Unsaturated pulse height
Saturated pulses (number of saturated time bins)
Stretch stereo drift distances into the x-y plane

The dependence of Lorentz angle and drift velocity on position is caused by the variation of the magnetic field. The first two effects listed introduce differences between positive and negative tracks, and also have the largest effect on hit position. For example, a hit is moved by about 0.5 mm when the angle between the track and the drift plane is 45 degrees. Many hits have a very small correction but a few are moved by more than 1 mm.

The purpose of the corrections is to remove systematic effects on drift distance. Improvements are also observed in the position resolution of hits, there are more hits on tracks, and more tracks are attached to the main vertex. In short, the tracks are better measured.

The effect on the track parameters is usually relatively small. This is because hits are shifted towards or away from the sense wires, and every time a track crosses a plane of sense wires or a cell boundary the effect on the track tends to cancel. The greatest effect is on positive tracks with transverse momentum less than 200 MeV, where the track becomes parallel to the plane of sense wires; and on short SuperLayer1 tracks with extreme theta values, where the Lorentz angle becomes rather different from the nominal value. In these two cases the reconstructed track can be moved significantly by the corrections.

The corrections package takes raw drift time as input, returns the corrected planar drift distance, and is layered with a simple interface. This enables its use more than once, such as in the pattern recognition too, if required. Switching the corrections on can have a small effect on the calibration of time-zero, drift velocity and Lorentz angle. The corrections for time-of-flight of the particle to the sense wire, and propagation delay along the wire are unchanged.

Last update: 19 October 2000 by John Lane (UCL) email: jbl@hep.ucl.ac.uk

http://www.hep.ucl.ac.uk/~jbl/CTD/corrections.html