3.3 Signal processing

The charge pulses produced by the detector are subsequently processed by a complex electronic system. The processing usually involves different actions, including:

• Differentiation of the pulse to separate it from the detector bias voltage.
• Establishing the electronic noise level for discrimination of the valid pulses.
• Amplification and shaping into a signal which amplitude can be determined with the best possible accuracy.
• Pulse height analysis (sorting by amplitude).
• Recording in memory to accumulate a spectrum.

The time required to process a detected event varies across the different types of detectors and pulse processing units. When the rate of incoming events increases the pulse processing electronics needs to reject some pulses during some fraction of time. Dead time is usually referred to the time required for the spectrometer to process the events, and it is usually expressed as a fraction of the real time elapsed during a measurement as.

In EDXRS the strong proportionality between the amplitude of the initially produced pulses and subsequently processed signals should accurately be maintained. As long as this condition is met, the pulse amplitude remains an accurate measure of the energy of the detected photons. Ultimately, the result of the photon counting can be represented as an energy dispersive spectrum listing the number of photons as a function of their energy.

The following sections highlight the two major signal processing approaches used in EDXRS.