Deadtime Calibration

EPMA Probe

 

StartWin > Description

Deadtime Calibration

To calibrate the deadtime constants for each spectrometer, StartWin may be utilized. StartWin is capable of producing a data set that is easily copied and pasted into Excel for use with Paul Carpenter's deadtime spreadsheet template.

         

First, set a sufficient count time (for example 60 seconds) for a precise measurement of intensities. If your interface supports sample (absorbed) current measurements, select this option also. Turn off the beam drift correction to acquire "raw" intensities. Then tune all spectrometers to the same element, for example Ti Ka on LiF and PET or Si Ka on PET and TAP. Use pure metal standards for high count rates and acquire 5 points at 10 different currents from about 10 nA to 200 nA.

 

Copy and paste the data set from the StartWin log window into a blank Excel spreadsheet and duplicate the beam current column if the absorbed current column is missing. See DEADTIME.DOC for more information on the data format required by DEADTIME.XLS. Finally create a column of the count time in the proper position (typically 1 second for data acquired by StartWin).

 

When the data is properly formatted, simply copy and paste the data set into the DEADTIME.XLS spreadsheet. The calculations on the new data set will be automatically updated and plotted. Use these new figures to update the values in the SCALERS.DAT file. Note that the actual calibrated deadtimes are entered on line 13, while the constant integer deadtime constants used to set the PHA interface, are specified on line 35 for the Cameca interface only.

 

Be aware however that deadtime is actually not a constant and can vary with the line energy of the x-ray being measured. For this reason, it is best if a pulse "stretcher" circuit can be placed in front of the counter-timer board to ensure that a "forced" deadtime is used to mask the actual deadtime of the spectrometer (as is done with the Cameca PHA hardware). If the pulse stretcher is adjusted to produce pulse widths greater than the worst deadtime seen on the spectrometer than the deadtime of the pulse stretcher (which is constant) can be used for the software deadtime correction. This will result in a much more accurate deadtime correction even at very high count rates.