Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Abstract: This paper deals with using some MATLAB functions and tools for designing the
first-order analogue Chebyshev filters and IIR Chebyshev filters. The first part of this paper is
focused on a design of an analogue filter via Chebyshev approximation approach, including
features and mathematical background of this iso-extremal approximation, approximation
of a normalized low-pass (NLP) filter, and some mathematical formulas for calculating its
fundamental parameters, such as constructing a tolerance scheme, an order of the Chebyshev
approximation, poles of the NLP’s transfer function, characteristic equation, and group delay.
Due to some frequency transformation formulas implemented in MATLAB, un-normalized forms
of frequency functions for low-pass (LP), high-pass (HP), band-pass (BP), and band-stop (BS)
analogue filters are also available. The second part of this paper describes a design of the first-
order IIR Chebyshev filter via Filter Visualization Tool (FVT), and functions, implemented in
Signal Processing Toolbox, whereas a conversion from the analogue form into the digital form
is done and discussed for bilinear transformation only.
For example, let us consider these values of un-normalized For the BP filter at ω0 = 2000π rad · s−1 :
radial frequencies (in radians per second):
For the LP and HP filters: ωmd = 1500π, ωmh = 2500π
. (31)
ωpd = 1200π, ωph = 2800π
Table 1. Summarization of Conversion Formu-
las and Corresponding Callings of MATLAB Where Bm = |ωmh − ωmd | / (2π) = 500,
Functions; Doleček J. (2009),THE MATH-
WORKS Inc. (2011) Bp = |ωph − ωpd | / (2π) = 800,
/ (ωmd Bm ) ∼
Filter Conversion Formulas MATLAB Function
Ωp = ω02 − ωmd
2
= 1.17,
Type
/ (ωpd Bm ) ∼
2 2
Ωs = ω0 − ωpd = 2.13.
Ωp = 1
NLP che1abp(n,Rp)
Ωs
ω p = ω 0 Ωp = ω 0 For designed BP filter, the radial frequency-magnitude
LP lp2lp(b,a,Wo) characteristic and corresponding MATLAB code are as
ω s = ω 0 Ωs
ωp = ω0 /Ωp = ω0 follows, thus:
HP lp2hp(b,a,Wo)
ωs = ω 0 /Ωs [NumHPPs, DenHPPs] = lp2bp(NumHs, DenHs, omega0, Bomega);
Ωs = ω02 − ωpd
2 / ω B
pd m HPPs = tf(NumHPPs, DenHPPs)
Bm =|ωmh − ωmd | / (2π)
BP Bp = ωph − ωpd / (2π) lp2bp(b,a,Wo,Bw) [HPP, omegaPP] = freqs(NumHPPs, DenHPPs, omega);
ωmh < ωph ωmd > ωpd plot(omegaPP, abs(HPP))
Bm < Bp
Ωs = ωpd Bm / ω02 − ωpd 2
Finally, for the BS filter at ω0 = 2000π rad · s−1 :
Bm =|ωmh − ωmd | / (2π)
BS Bp = ωph − ωpd / (2π) lp2bs(b,a,Wo,Bw)
ωmh > ωph ωmd < ωpd ωpd = 1500π, ωph = 2500π
. (32)
Bm > Bp ωmd = 1200π, ωmh = 2800π
−1
−1
2 · π 1 − z
Ĥ z = Ĥ s → −1
. (35)
tg fps π 1 + z
f
ω0 2000π 1
pω = = = = 0.25. (36)
ωs 8000π 4
ACKNOWLEDGEMENTS
Fig. 8. Un-normalized IIR low-pass filter: “Normalized”
radial frequency-magnitude characteristics with the This work is supported by project SP2012/111, named
prewarped mode (light curve) and without this mode “Data Acquisition and Processing from Large Distributed
(dark curve); the Cartesian coordinates. Systems II” of Student Grant Agency (VSB - Technical
In comparison with the non-prewarped mode, value of ω0 University of Ostrava).
equals to 0.42. So, result of (36) is not considered.
REFERENCES
4. CONCLUSIONS Davı́dek V., Laipert M., and Vlček M. (2006). Ana-
This paper deals with using some mathematical back- logové a čı́slicové filtry. ČVUT.
ground - the Chebyshev approximation, Chebyshev poly- Doleček J. (2009). Modernı́ učebnice elektroniky 6. dı́l
nomials, selected MATLAB functions from Signal Process- Kmitočtové filtry, generátory signálů a převodnı́ky dat.
ing Toolbox, and other tool (e. g. Filter Visualization Tool) BEN - technická literatura.
to design the first-order (also kind I) analogue Chebyshev Nečas J. et al. (1977). Aplikovaná matematika I (A až
filters and IIR Chebyshev filters. Because this paper is L). SNTL, Prague.
mainly centred on the mathematic viewpoint, there are Rektorys K. et al. (2000a). Přehled užité matematiky I.
lots of fundamental (maybe essential) formulas (de Moivre, Prometheus.
Euler, also Jacobi, Gegenbauer, or Rodrigues). Features Rektorys K. et al. (2000b). Přehled užité matematiky II.
of the Chebyshev polynomials are also mentioned and Prometheus.
discussed in forms of definition, representations, roots, and THE MATHWORKS Inc. (2011). Signal processing
extrema. toolboxTM : User’s guide. Version 6.16 (Release 2011b),
offline.
The second part of this paper is focused on designing
the first-order Chebyshev filters in p-plane (the NLP fil-