We wish to use impulse invariance or the bilinear transformation to design a discrete-time filter that meets specifications of the following form: 

For historical reasons, most of the design formulas, tables, or charts for continuous-time filters are normally specified with a peak gain of unity in the passband; i.e., 

Useful design charts for continuous-time filters specified in this form were given by Rabiner, Kaiser, Herrmann, and Dolan (1974).

(a) To use such tables and charts to design discrete-time systems with a peak gain of (1 + δ₁), it is necessary to convert the discrete-time specifications into specifications of the form of Eq.(P7.3-2). This can be done by dividing the discrete-time specifications by (1 + δ₁). Use this approach to obtain an expression for δ₁ and δ₂ in terms of δ₁ and δ_2. 

(b) In Example 7.2, we designed a discrete-time filter with and maximum passband gain of unity. This filter can be converted to a filter satisfying a set of specifications such as those in Eq. (P7.3-1) by multiplying by a constant of the form (1 + δ₁). Find the required value of δ₁ and the corresponding value of δ₂ for this example, and use Eq. (7.19) to determine the coefficients of the system function of the new filter.

(c) Repeat Part (b) for the filter in Example 7.3.

Transcribed Image Text:

1 -- đ1 < |H(el")]I s 1 + 81. 0 s \wi < wp. |H(ei®)l s d2. 1 – ổ1 s|H(jN)l < 1. 0 < [2]