(5/16/02)
There is usable depth information in the 200ns shaped strip signals but only in the top half of the depth (near the anode). The strip signal amplitude can place the event in either the top or bottom half and for events in the top half it could resolve things farther.
This plot shows 200ns shaped strip height vs depth. It is from a top illuminated Cs 137 flood run. This run gives events fairly evenly distributed over all depths. The histogram selects only the events with high pixel values (the high energy peak of the spectrum). This gives a large signal and minimizes the effect of several error sources.
The red spot at depth 10 is a test pulser peak. This peak width gives some indication of the electronics noise width. The test peak is orders of magnitude higher count than the data max so the noise FWHM is smaller than its width on this plot.
The depth was calculated from cathode/pixel which isn't perfect but should be close enough for this analysis. The depth vs pix plot shows the depth "calibration" and the range of events used. Strip collection events indicated by negative strip signals have been removed.
Since the peak width isn't dominated by electronics noise, adding strips together can increase the signal without adding significant noise. The sum of all strips shows an improved shape with some hope of resolving depths in the low half.
Further improvement should be possible by using 1usec strip shaping. This results in less signal at 0 depth but it has a better slope (increasing signal vs depth) in the lower half of the depth range.
The width of the strip pulse height distribution increases for lower energy events. There are several error sources. Things such as gain and offset differences between different pixels and between different strips could be calibrated out. This data does include some correction for offsets but none for gains.
Despite the problems, the distribution of summed strip height using events from all energies still looks usable.