Dead time corrected and charge normalised maps generated with the MicroDas fast data acquisition system

https://doi.org/10.1016/S0168-583X(01)00617-6Get rights and content

Abstract

A new data acquisition system, called MicroDas, is capable of collecting data at rates of up to 100 k counts per second per station. To correct for the large dead times that are implicitly inherent with such high data rates, dead time information is also collected for each station. Since the count rate varies with different scan regions of a sample, dead time information is collected for each pixel of each station. To generate quantitative charge normalised maps of spectral features, charge information is also collected using an ultra sensitive charge-to-frequency converter. To evaluate this new system, examples are provided that demonstrate the improvement to maps when dead time and charge information are used to correct the original raw energy data. We conclude that the most quantitative accurate maps are generated when charge triggering with beam blanking is used.

Introduction

The main motivation for the development of this new data acquisition system was to collect quantitative data using large acquisition rates and to compensate for any artifacts that are introduced into the collected data. The design adhered to the method of total quantitative scanning analysis (TQSA) [2], which describes the concept of permanently storing all information about each event when performing scanning of any form.

This paper discusses the source of the artifacts associated with fast data acquisition. Then the method to collect quantitative data is discussed, with emphasis on how to compensate for these artifacts. This is followed by a discussion on how we implemented these requirements by designing the MicroDas system. Finally, the accuracy with which the MicroDas system generates charge normalised and dead time corrected maps are shown.

Section snippets

Fast data acquisition

There are five main hardware stages involved with collecting data. The first three are the NIM electronics components called the detector–preamplifier stage, the shaping amplifier stage and the analog-to-digital converter (ADC) stage. For an in-depth discussion about these stages, refer to [1] or any NIM electronics catalogue. The fourth is the interface between the NIM electronics and the computer and the fifth is the computer.

The process of acquiring data is as follows. When a particle is

Generating charge normalised maps

To generate maps that are truly quantitative, charge information is collected and is used to generate two-dimensional maps in which the data represented at each pixel is normalised to charge. This helps to improve the accuracy and the contrast of structure in the maps. Since the beam currents associated with ion microscopy are reasonably small, special hardware, called charge digitisers, are used. They produce a TTL pulse every time one unit of charge has been collected. We use the ATC-170

The MicroDas system

The MicroDas system comprises of a MicroDas unit, used to condition the signals between the NIM electronics and a computer, and “add-on” cards in a computer to record signals from the NIM electronics and to generate signals to control scan hardware. The features required on these add-on cards are a digital input port, to read the ADC's digital output buffers, two digital-to-analog converters, to provide the analog signals to scan the beam and five counters, to provide the real time, the four

Examples of quantitative maps

Fig. 2, Fig. 3, Fig. 4 show the total and copper X-ray yield maps from PIXE data in which the data were collected using charge triggering with beam blanking, charge triggering without beam blanking and smart data triggering using charge sampling, respectively. These data were collected before a PIXE tomography experiment. The geometric efficiency was 0.018 and a 3.7 MeV proton beam adjusted to a spatial resolution of 0.5 μm was used. The “ladder” in the copper map is a section of a 2000 mesh

Conclusion

The MicroDas system is capable of collecting data at very fast acquisition rates. The only limiting factors are the natural time delays associated with the detectors and NIM electronics used.

To generate charge normalised maps, charge information is also collected. However, the inherently high dead time associated with such fast acquisition rates introduces localised artifacts into maps. By also collecting dead time information at every pixel in a scan, these localised artifacts can be removed.

References (3)

  • G.F. Knoll

    Radiation Detection and Measurement

    (1989)
There are more references available in the full text version of this article.

Cited by (0)

View full text