You are on page 1of 5

Review questions for Chapter 5 (Filter Structure)

The following question is aimed to help you test your understanding of the slide and
lecture. There are mostly simple conceptual question. Most of the answer can be obtained
from the slide and lecture.
Slide,
topic

Question
TF true false question
S: Subjective explanation
C :Calculation and design

2: Introduction
1. (S) Give the design flow of a digital filter design and implementation
Hint:
1) Describe the specification ( mention passband and stopband cut off
frequency, max gain in stopband and minimum gain at passband).
2) ??
3) ??
4) Implement a suitable structure
5) ??
6) Choose a DSP processor
7) Write the code (C or assembly)
8) Download program into the DSP processor
2 (TF) A transfer function can have only one structure (true/false)
3) Why use flow diagram represent a filter structure?
4) What is a filter structure?
4: Intro

1.What determine an implementation complexity of a particular structure?


2.How does the order of the filter affect the complexity?
3. How the order of the filter affect the filter performance?

5: System
Function and
LCCDE

4. (TF) The effect of quantization error depend on the filter structure


implemented.
1. Given an impulse response how can the transfer function and frequency
response obtained analytically.?
2. How can an impulse response of a digital filter obtained experimentally?
3. What does an LCCDE equation describe?

4. Explain how Z transform can be used to analyze an LTI system?

6: System
Function and
LCCDE

1. Identify the filter coefficients given the transfer function H(z).


q

bz
Y z
H ( z)

X ( z)
a z
k 0
p

k 0

2. (TF) The transfer function H(z) define the system behavior in time domain
(true/false)
7.

1. Identify the filter coefficient given the LCCDE equation below. Is it a FIR or
IIR filter ? What is the order?

1
3
y[n] y[n 1] y[n 2] x[n] 2 x[n 1] x[n 2]
4
8
2. The following matlab function designs an IIR filter using the elliptic method,
order=8, the cut off freq is 300Hz, sampling frequency = 4000 Hz
[b, a] = ellip(8, .5, 50, 300/2000)
The coefficient vector is given as
b=
0.0046 -0.0249
0.0046

0.0655 -0.1096

0.1289 -0.1096

0.0655 -0.0249

a=
1.0000 -6.9350 21.5565 -39.1515 45.3884 -34.3665 16.5896 -4.6673
0.5860
Write the LCCDE equation. Draw the direct form 2 filter structure.
8: Block
diagram and
signal flow
graph

1. Name the 3 elements in the block diagram or signal flow graph.


2. Draw the block diagram for the LCCDE equation. Find the transfer
function H(z)
y[n]= 0.5x[n] 0.2 x[n-1] + 0.4 x[n-3]
y[n]= 0.5x[n] 0.2 x[n-1] +0.4 y[n-1]
y[n]= 0.5x[n] 0.2 x[n-1] +0.4 y[n-1] + y[n-2]

9.

10 Block
diagram

1. Differentiate between recursive and non-recursive LCCDE equation.


2. (TF) The signal flow graph implement the LCCDE equation

1. Given the transfer function below, draw the block diagram. What is the
order?
H ( z)

4
1

1 0.7 z 0.3 z 2

Draw the structure. Find the difference equation.


2. Derive the difference equation given the structure implementation by the
block diagram below. What is the order? Derive the transfer function. Is the
filter stable?

x[n]

y[n]

0.7

z-1
y[n-1]

-2
12

z-1
y[n-2]

What is the difference between canonic and non canonic filter structure? Which
one is preferable, why?

13

Write the LCCDE equation for both the 2 structures below and show that they
are the same.

Equivalent
structure

13 IIR
structure
15-27
Direct form 1
and 2

1. Name the 3 different structures to realize a IIR filter.


1. Show that both the structure give the same transfer function H(z)

Direct form 1 , first order

w1[n]

w2[n]

Direct form 2 , first order


Hint: Use w[n] and w2[n] to simplify the LCCDE equation.
H(z)= Y(z)/X(z) = Y(z)/W1(z) * W1(z)/X(z)
Which one is canonic and which one is non canonic structure? How many
memory elements required in each structure?
Explain the advantage of direct form 2 compared with direct form 1.
36 - 41

Learn to factorize polynomial


http://www.wtamu.edu/academic/anns/mps/math/mathlab/col_algebra/col_alg_t
ut7_factor.htm
http://www.sosmath.com/algebra/factor/fac02/fac02.html
Now try to factorize this (without looking at the slide). Draw all the different
possible realization for cascade structure.
H z

0.44 z 1 0.362 z 2 0.02 z 3


1 0.4 z 1 0.18 z 2 0.2 z 3

H(z) = 1 + z-1 + 0.75z-2 + 0.375z-3 + 0.125z-4


If you do not know how to factorize the polynomial, you will not be able to
draw the cascade structure.
34: Parallel
structure

In order to get the parallel structure you must master the partial fraction
expansion method
http://www.swarthmore.edu/NatSci/echeeve1/Ref/LPSA/PartialFraction/Partial
Fraction.html
http://www.purplemath.com/modules/partfrac3.htm
What is the advantage of parallel structure?

You might also like