A well-designed digital spectrometer permits to implement the optimal-filter weighting function, so optimizing the spectral quality of the measurements. Nonetheless the optimal-filter theoretical performance is difficult to attain owing to the effects related to (i) the residual aliasing in the analog-to-digital conversion, which emerges as a high-frequency perturbation of the filter weighting function, and (ii) the contribution of the quantization noise to the overall system noise. In this paper it is shown that, given a resolution of the analog-to-digital converter, a minimum for effects (i) and (ii) exists, which permits to optimize the performance of the system. Such minimum is obtained by an accurate setting of the anti-aliasing filter bandwidth, and depends also on the required weighting-function width. The design figures which permit to achieve such an optimization will be presented. A procedure of noise data fitting will be also shown, which permits to disentangle the various noise components, including the quantization noise. This latter procedure may be used as a diagnostic tool to check the proper operation of the spectrometer.
Design rules for optimization of digital spectrometers / A. Pullia, R. Grimoldi - In: 2001 IEEE Nuclear science symposium: conference record : nuclear science symposium : medical imaging conference : room-temperature semiconductor detectors : sessions on nuclear power systems : 4-10 November 2001, San Diego, California, USA. Volume 2 / [a cura di] J. A. Seibert. - Piscataway (New Jersey) : IEEE, 2002. - ISBN 0-7803-7324-3. - pp. 1004-1008 (( convegno Nuclear science symposium and medical imaging conference tenutosi a San Diego (California) nel 2001 [10.1109/NSSMIC.2001.1009723].
Design rules for optimization of digital spectrometers
A. PulliaPrimo
;
2002
Abstract
A well-designed digital spectrometer permits to implement the optimal-filter weighting function, so optimizing the spectral quality of the measurements. Nonetheless the optimal-filter theoretical performance is difficult to attain owing to the effects related to (i) the residual aliasing in the analog-to-digital conversion, which emerges as a high-frequency perturbation of the filter weighting function, and (ii) the contribution of the quantization noise to the overall system noise. In this paper it is shown that, given a resolution of the analog-to-digital converter, a minimum for effects (i) and (ii) exists, which permits to optimize the performance of the system. Such minimum is obtained by an accurate setting of the anti-aliasing filter bandwidth, and depends also on the required weighting-function width. The design figures which permit to achieve such an optimization will be presented. A procedure of noise data fitting will be also shown, which permits to disentangle the various noise components, including the quantization noise. This latter procedure may be used as a diagnostic tool to check the proper operation of the spectrometer.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.