Recent advances in DXP triggering algorithms have greatly improved performance in the sub-1keV spectral regime. This is of particular interest to Electron Microscope users and anyone working with light elements. Moreover the proper handling of partial ionization events improves spectral integrity in vacuum and non-vacuum environments alike.

The data in figure 1. was collected using a high quality SiLi detector from Princeton Gamma Tech with the DXP Saturn digital x-ray spectrometer. A pure Boron sample with a thin layer of Carbon was analyzed at 64 microseconds peaking time. Of interest is the complete separation of the Boron K line from noise.

Data integrity partially depends on the ability of the processing electronics to detect and reject piled-up pulses. This is most evident when working at high count rates, where triggering circuits that rely on low-pass filtering tend to fail. The result is a combination of degraded energy resolution, creeping background levels and anomalous peaks.

The DXP's inherent performance at high rates has recently been further improved with new pileup inspection algorithms

The data in figure 2. was collected at maximum throughput (63% dead time), using an HPGe Iglet detector from Ortec and the DXP Saturn digital x-ray spectrometer running both the old (revision G, red) and new (revision J, blue) filtering algorithms. Note the dramatic improvement in pileup rejection evidenced by lower background levels and smaller pileup peaks. Note also the improved efficiency below 1keV due to the new trigger algorithm.

In most pulse processing systems, and for various reasons (eg. pole-zero and baseline correction errors), energy peaks tend to shift with count rate. In this regard the DXP architecture stands alone. The absence of pole-zero circuitry and the careful handling of baseline data yield a measure of energy with virtually no dependence upon the incident rate. This maintains data integrity during incident rate fluctuations, and faster data acquisition in general at virtually no cost.

The data in figure 3. was collected using an HPGe detector from Canberra and the DXP Saturn digital x-ray spectrometer, again running both the old (revision G) and new (revision J) filtering algorithms. The peak energy from an Fe55 source is plotted as a function of rate, which is varied to more than four times the maximum throughput. While most competing systems would simply fail at this intensity, the DXP running revision G code continues to collect data at a shift of slightly over 0.10%. With the new release the effect is further reduced to an astounding 0.02%!