Cathode strip detector

The performance test was carried out using cosmic rays before the CDC was installed into the Belle detector. Data were taken with a simple trigger system with scintillation counters located in the inner cylinder to select cosmic rays that passed through the beam pipe area inside CDC. Without any special optimization, the anode wire data provide an extrapolation resolution of 160 $\mu$m in the $r$-$\phi$ plane and 3.6 mm in the $z$ direction to the CDC cathode part. Anode-wire tracks were used to estimate the detection efficiency of the cathode readout.
The charge distribution on the cathode strips depends on the anode-cathode gap and the spread of the avalanche at the anode. When an incoming particle is inclined with respect to the normal incidence, the electron avalanche is spread along the anode wire, making the induced charge distribution broader. This effect was made minimal owing to the gate expansion and the limitation of the charge integration time to 150 ns.
Consecutive cathode hits are recognized as a "cluster". At least three strips are required to form a cluster. If two local charge maxima are separated more than three strips, they are regarded as separate clusters. The matching of a track to a cathode-hit cluster is done by requiring that the extrapolated $\phi$ is within the $\phi_{min}$ and $\phi_{max}$ region covered by the cathode strip hits and that the extrapolated $z$ is within the $z$ region covered by the strip hits. Fig. shows the cathode efficiency as a function of incident angle for $V_{th}$ = 50 mV and an anode high voltage at 2.3 kV. The cathode strip efficiency is defined as the ratio of the number of matched tracks to the number of sampled tracks. As the Belle tracking trigger requires at least two hits in the three cathode layers for each track, the trigger efficiency by the cathode strips is better than 99%, even for the worst case of the incident angle of $90^o$ where the cathode charge is smallest due to gas gain saturation.
The spatial resolution of the cathode readout was obtained with the self-tracking method in the $z$ direction using cathode data alone. Fig. shows measured results of the intrinsic cathode resolution as a function of incident angle. The data for the three layers are consistent. The data measured in a beam test using a prototype detector is also shown [4].

Figure: Angular dependence of the cathode detection efficiency.

Figure: Angular dependence of the cathode resolution.

To evaluate the improvement in the three dimensional tracking provided by the cathode data, each cosmic ray track that passed through the entire CDC was treated as two separate tracks (up and down) going outwards from the beam pipe area. The tracking performance was evaluated by checking the mismatch of these two track segments, $\Delta z = z_{up} - z_{down}$, at the closest approach to the beam axis. Figs.  (a) and (b) show $\Delta z$ distributions respectively (a) without and (b) with including the cathode cluster information in three dimensional tracking with the axial and stereo anode wire information. The $\Delta z$ resolution was improved from about 3 mm to 0.64 mm with the cathode data. We note that $\Delta$z obtained in the present analysis is equivalent to $\sqrt{2}$ times of the track extrapolation resolution of CDC to the interaction point.

Figure: The effect of the cathode readout information on $\Delta z = z_{up} - z_{down}$ for cosmic ray tracks: (a) without and (b) with the use of the cathode information in tracking, respectively.