High rate pulse processing algorithms for microcalorimeters

TL;DR: In this article, an offline-processed microcalorimeter pulse stream with modified HPGe filter algorithms was used to provide an initial engineering evaluation of their performance as practical filters, capable of achieving sufficiently good energy resolution for most applications.

Abstract: Microcalorimeters, cryogenic radiation detectors measuring the energy of photons by the increase of temperature in an absorber, can achieve energy resolutions more than an order of magnitude better than HPGe detectors. However, due to the thermal nature of the pulse generation, the active volume has to be small to maintain good resolution, and pulse decay times are in the order of milliseconds. Consequently, the detection efficiency is low and count rates are limited, especially for commonly used “optimum filter” algorithms that require isolated pulses to measure pulse heights. This is typically solved by building systems with multiple detector elements (arrays). Large arrays, however, require that as much pulse processing as possible be performed at the front end of the electronics to avoid transferring large amounts of waveform data to a host computer for processing. Pulse processing algorithms developed by XIA LLC for use in digital spectrometers with HPGe detectors, suitably modified for the slower time scale, meet this requirement. In the work reported here, we offline-processed microcalorimeter pulse streams with modified HPGe filter algorithms to provide an initial engineering evaluation of their performance as “practical” filters, capable of achieving sufficiently good energy resolution for most applications while being a) simple enough to be implemented in the readout electronics and b) capable of processing overlapping pulses and thus of achieving higher count rates. In the course of this work, a new filter was developed that uses only a fraction of a pulse while still achieving good energy resolution for very high count rates. The success of this work suggests that future microcalorimeter read-out systems can indeed be built with electronics on which these filters are implemented in multiplexed form, taking advantage of high speed digital signal processing elements to process many channels in parallel at a large reduction in processing cost per channel.