INFN

jSpecView: A Graphical Tool for Data Analysis

Email Us
 
For more information contact Călin A. Ur
Last updated: March 29th, 2013


General Description
Download Area

General Description

The online analysis programs have to take into account the particular way of data taking. In a standard acquisition system the energy and time values are extracted from the detector signals by the frontend electronics and the DAQ has only to collect, store and/or histogram them.A sampling system like MD 2S provides directly the digitized version of the detector signals for a selected time interval around the event trigger; these waveforms are collected by the DAQ, saved on hard disk and passed through digital filters to determine the energy and time values. Provided one is able to handle large volumes of data, this way of proceeding is very powerful and gives the possibility to optimize the parameters of the filters for each 'detector-sampling chain' pair so as to get the best results for the specific detector and application (energy and time).

To satisfy these requirements, the online GUI must be characterized by a large flexibility. It must allow for a detailed analysis of the waveforms i.e. exponential decay , baseline , start of the signal, pulse rise form, pile-up, etc. The GUI must allow the use of any type of digital filters, and one needs to have the possibility of inspecting the results of each step by viewing the wave forms and collecting parameter spectra. Besides the usual spectrum-viewer actions, the online GUI provides an oscilloscope mode for the visualization of the waveforms prior to any treatment and after application of the digital filters. It also allows to perform chained operations on waveforms in order to determine the exponential decay constant of the signals, the features of the baseline, the starting point of the signals and the criteria for removal of pile-upped events.

We developed such a graphical analysis GUI derived from the SpecView code used in GASP and implemented in the Java language. A screenshot of a typical analysis configuration is shown in Figure 1. Gamma-rays from Eu and Am sources were recorded with a small x-ray detector. Waveforms of 2048 samples digitized at 64 MHz were stored on disk. The upper panel of Figure 1 shows such a waveform. The second panel illustrates the distribution of the signal decay-times obtained from a fit with an exponential function. In order to extract the energy value corresponding to the signal a digital trapezoidal filter is applied and the result is shown in the third panel. A discrete Fourier transform of the input signals shows the frequency domain transmitted through our system. Since the full range shown in the fourth panel correspond to the sampling frequency of 64 MHz it can be noticed that the analog conditioning circuit was optimized to cut-off frequencies higher than 30 MHz in order to avoid aliasing phenomena. The last panel shows the energy spectrum obtained by histogramming the the results of the trapezoidal  filter.

dcsdcs

Figure 1: Example of jSpecView

Download Area