Can a second order bandpass sigma delta modulator achieve high signal-to-noise ratio for lowpass inputs

Ling, Bingo Wing-Kuen and Ho, Charlotte Yuk-Fan (2008) Can a second order bandpass sigma delta modulator achieve high signal-to-noise ratio for lowpass inputs. Chaos, Solitons and Fractals, 37 (3). pp. 928-930. ISSN 0976-0779

Full content URL: http://dx.doi.org/10.1016/j.chaos.2006.09.082

Documents
letter_chao_SDM_second_order.pdf
[img]
[Download]
[img]
Preview
PDF
letter_chao_SDM_second_order.pdf - Whole Document

60kB
Item Type:Article
Item Status:Live Archive

Abstract

Institutively, second order SDMs usually achieve lower SNR than high order ones because high order loop filters can achieve better noise shaping characteristics. Moreover, the signal transfer function should be designed to have large values and the noise transfer function should be designed to have small values at the passband of loop filters in order to achieve good noise shaping characteristics, so SNR should be high if input signal bands match passbands of loop filters and low otherwise. Based on this argument, one may expect that SNR will be low when input signals have lowpass characteristics while loop filters have bandpass characteristics.

However, since the above argument is based on the noise shaping theory which is formulated using a linear model, while quantizers in SDMs are nonlinear components, the linear model may not explain nonlinear system behaviors. In this letter, a counterexample is given to illustrate that a second order bandpass interpolative SDM may also give a very high SNR for lowpass inputs.

Additional Information:Institutively, second order SDMs usually achieve lower SNR than high order ones because high order loop filters can achieve better noise shaping characteristics. Moreover, the signal transfer function should be designed to have large values and the noise transfer function should be designed to have small values at the passband of loop filters in order to achieve good noise shaping characteristics, so SNR should be high if input signal bands match passbands of loop filters and low otherwise. Based on this argument, one may expect that SNR will be low when input signals have lowpass characteristics while loop filters have bandpass characteristics. However, since the above argument is based on the noise shaping theory which is formulated using a linear model, while quantizers in SDMs are nonlinear components, the linear model may not explain nonlinear system behaviors. In this letter, a counterexample is given to illustrate that a second order bandpass interpolative SDM may also give a very high SNR for lowpass inputs.
Keywords:chaotic filter bank, cryptographical system, Cryptography using chaos
Subjects:H Engineering > H620 Electrical Engineering
H Engineering > H310 Dynamics
Divisions:College of Science > School of Engineering
ID Code:2678
Deposited On:13 Jun 2010 11:10

Repository Staff Only: item control page