Since the silicon detector segments have (varying) dark currents, the integrated signal output is dependent on the integration (i.e., dwell) time. To remove the influence of this dark contribution, a time dependent pedestal is subtracted from the detector output.
To perform the calibration with the IDL routine det_cal.pro, you need to collect a few images with the detector in dark conditions.
To do this
sm_script_detcal @detcal_script.datwhich runs the script file that reads
# Line 1 is repeated before every scan (why?) # Line 2 is repeated before every scan (why?) scan,XPZT,0.001,1000,YPZT,0.001,10,1.00 scan,XPZT,0.001,1000,YPZT,0.001,10,2.00 scan,XPZT,0.001,1000,YPZT,0.001,10,5.00 scan,XPZT,0.001,1000,YPZT,0.001,10,10.00
Start an IDL session on the microscope computer. Start the detector calibration routine:
IDL> det_cal, n_darkfiles=4The program will perform a linear fit to the dark signal, characterize the noise by taking a Gaussian fit of voltage distributions (hopefully showing the spread expected by the statistics of the number of electron-hole pairs created per photon), and do a Fourier analysis of the noise to see what time frequencies have significant noise. If your first file is x1ais_08feb2004_0025.sm, as an example, it will make files with that same file header but extensions of _histogram.ps, _dark_noise.ps, _dark_signal.ps, and _fourier.ps. It will also create the file /mnt/x1a/sm_gui_disp_settings_stxmi.dat, which contains the latest best estimate of detector voltage scalings and offsets.