Studies with Cathode Drift Chambers for the GlueX Experiment at Jefferson Lab
Lubomir Pentchev for the Forward Drift Chamber group and the GlueX collaboration
(pentchev@jlab.org)

THE DETECTOR
The GlueX experiment at JLab (Virginia, USA) uses tagged polarized photon beams from the recently upgraded 12GeV Continuous Electron Beam Accelerator Facility (CEBAF) to search for exotic hybrid mesons (quantum numbers not allowed as quark-antiquark configuration) as predicted by lattice QCD. GlueX detector - hermetic coverage (to enable partial wave amplitude analysis) for charge and neutral particles using calorimetry and tracking within 2T-solenoid; high trigger (up to 200 kHz) and data acquisition rate (up to 1 Gby/sec). 

12,672-channel drift chamber system consisting of 24 1m-diameter chambers, grouped in 4 packages and inserted in the magnet bore. 
Pipelined data from 10,368 flashADC (125MHz) channels and 2,304 TDC channels

Each drift chamber consists of two cathode planes divided into strips facing  the wire plane. 
Wire plane - sense (20 mm) and field wires (80 mm), 5mm apart. 
Cathodes - coopper strips with a pitch of 5mm. Coupling of the 2mm-thin strips to the pre-amp connector challenging.
Light frame made, mostly of Rochacell, to allow detection of low energy photons by outside e.m. calorimeters. 

Resolution

Same avalanche is seen in three projections (both cathodes and the wire).
Such redundancy allows  to study uniquely the charge induction process and the strip resolution.
 
Cathodes used to reconstruct avalanche (and hit) position along the wires.

Each chamber gives a point in space of the track.

Top right - wires (avalanches) reconstructed by the cathode strips in one chamber. The widths of the projections on the x-axis (top left in blue)  are used to estimate (left plot) strip resolution after correction for the strip angles. Tracking is used to estimate the strip resolution along the wire (y-axis) (red points). The difference between resolutions along and perpendicular to the wire - due to transverse diffusion or finite avalanche extent as discussed on the right panels.

High Cathode Strip Resolution and 3D Avalanche Reconstruction - allows unique studies of avalanche  development around wire  using 55Fe source

Cell  10x10 mm

The two cathode planes "see" the avalanche position in 3D:
perp. to cathodes (Z) from ratio of the total charges on the cathodes.
Perp. to wire (X) from strip crossing
Along wire (Y) from strip crossing
2D transverse distribution (right plot) of avalanches  around wire (XZ) from 55Fe source  at bottom side.

Plotted is the apparent avalanche position  as "center-of-gravity" seen by the strips. Corrections to the reconstructed positions as a result of the induced charge distortion due to the proximity of the avalanche to the wire not applied.

Garfield simulations predict the cathode charge asymmetry that depends on the position of the avalanche around the wire - blue lines in the plots below.

The cathode charge ratio dependence on the total charge (plot above) can't be reproduce within Garfield only. A possible explanation  is the azimuthal extent of the avalanche mediated by UV photons, that increases with the total charge. 
Plot on the left shows the azimuthal extent of  the avalanche vs the charge estimated by distributing electrons around the wire (to simulate the photons) and then propagating the avalanches with Garfield microscopic simulations. Examples of such simulations shown on the top plots.    

Cluster Counting Studies - space charge effects within the same track; longitudinal avalanche extent

One chamber has been modified with 4cm drift gap (left). High single electron efficiency and big drift times allowed to count the number of clusters for each track using fADC (right). 
The number of clusters turned out to be ia strong function of the track angle w.r.t. wire (bottom). For tracks perpendicular to the wire the electrons travel almost the same paths (bottom left) encountering higher space charge at the wire from the previous electrons. This can explain the lower number of cluster for such tracks. From the plot we estimate a lateral extent of each cluster to be of the order of few hundred microns.

Summary and conclusions: The cathode drift chamber tracking system for the GlueX experiment has been successfully operated, giving already physical results.  Redundancy in avalanche reconstruction allows operation at low threshold and precise calibration thus reaching resolution of 150 microns with beam tracks at high efficiency. The resolution might be limited by the finite extent of the avalanche.  Higher resolutions have been obtained with 55Fe source demonstrating for a FIRST TIME the possibility to reconstruct the position of the avalanche  w.r.t. the wire in TWO DIMENSIONS (one dimensional reconstructions have been done before). The dependence of the cathode charge ratio on the total charge can be explained by photon-mediated azimuthal extent of the avalanche - the dependence of the angular extent on the charge has been estimated for Ar/CO2 gas mixture.  Cluster counting studies with 4cm-gap modified chamber show sharp decline in the cluster number  for tracks perpendicular to the wire. This demonstrates  the effect of the space charge along the wire within the same track. The uncertainty of the longitudinal extent  of the avalanche may contribute to the difference between the cathode resolutions along and perpendicular to the wire.