CRDC calibrations
Mask calibration
CRDC
mask calibrations are used to relate channel numbers and true physical
distances in mm for the position-sensitive cathode readout drift chambers. A
mask with a distinct and well-determined pattern of slits and holes can be
inserted remotely in front each CRDC. The particles passing through the holes
leave the pattern of the mask in the position spectrum of the detectors.
First-order polynomials are used for the calibration. In x direction, where the
position is determined from the charge induced in the pads, the slope of the
first-order polynomial is fixed by the geometry of the detector to 2.54 mm/pad.
The offset has to be set to center the "beam hole" at 0 mm. Since the
drift time depends on the drift voltage, the gas composition and pressure, the
y slope will change from experiment to experiment.
How to take data for the mask calibration:
- Set the trigger to S800 singles and d/s 1.
- On PanelMate, open page 4 in S800DRIV.
- Click on "upstream mask" and hit the "insert" button.
This puts the mask in front of CRDC1.
- Start a run without taking to disk and look in the S800 SpecTcl at
the 2d xy position spectrum of CRDC1
(s800.fp.crdc1.x,s800.fp.crdc1.tac). If it starts to look like the
spectrum shown below (the screenshot corresponds to mask 2, mask 1 looks
inverted, in the sense of the triangle of dots pointing downwards), start
taking data to disk until it has about the quality of the screenshot
shown. If only a small fraction of the S800 focal plane is illuminated
(dispersion-matched optics), consult with the S800 group to sweep the beam
across the focal plane by changing the Brho of segment 8.
- Remove the "upstream mask" with PanelMate.
- Repeat the procedure with CRDC2 by inserting the "downstream
mask" on the same PanelMate page.
For
past experiments, we recommend to use SpecTcl 2.1 for the analysis of the
mask-calibration spectra and to use the "map to real parameters"
option in xamine to display the channel number of the TDC that used to
processes the drift time. s800.crdcn.drift.raw was a 16bit parameter
and 16bit spectra are too big to be handled by SpecTcl/Xamine. In SpecTcl 2.1,
the spectrum can be binned to a reasonable size and the mapping option then
recovers the true channel numbers of the TDC. Consequently, in older versions
of SpecTcl, the channel number for the y calibration must be scaled to consider
the 16 bit.
Pad Calibration