READING ASSIGNMENTS

Signal Processing First

ƒ This Lecture:
ƒ Chapter 6, Sections 6-1, 6-2, 6-3, 6-4, & 6-5

Lecture 12
ƒ Other Reading:
Frequency Response ƒ Recitation: Chapter 6
of FIR Filters ƒ FREQUENCY RESPONSE EXAMPLES
ƒ Next Lecture: Chap. 6, Sects. 6-6, 6-7 & 6-8

2/25/2005 © 2003, JH McClellan & RW Schafer 1 2/25/2005 © 2003, JH McClellan & RW Schafer 3

LECTURE OBJECTIVES DOMAINS: Time & Frequency

ƒ Time-Domain: “n” = time
ƒ SINUSOIDAL INPUT SIGNAL
ƒ x[n] discrete-time signal
ƒ DETERMINE the FIR FILTER OUTPUT
ƒ x(t) continuous-time signal

ƒ Frequency Domain (sum of sinusoids)

ƒ FREQUENCY RESPONSE of FIR MAG
ƒ Spectrum vs. f (Hz)
ƒ PLOTTING vs. Frequency
• ANALOG vs. DIGITAL
PHASE
ƒ MAGNITUDE vs. Freq ƒ Spectrum vs. omega-hat
ƒ PHASE vs. Freq jωˆ
jωˆ
H ( e ) = H ( e jωˆ ) e j∠H ( e ) ƒ Move back and forth QUICKLY

2/25/2005 © 2003, JH McClellan & RW Schafer 4 2/25/2005 © 2003, JH McClellan & RW Schafer 5
 DIGITAL “FILTERING” FILTERING EXAMPLE

x(t) x[n] y[n] y(t) 6

y7 [n ] = ∑ (17 )x[n − k ]
A-to-D FILTER D-to-A
ƒ 7-point AVERAGER
ω̂ ƒ Removes cosine k =0
ω̂
ƒ By making its amplitude (A) smaller
ƒ CONCENTRATE on the SPECTRUM
ƒ SINUSOIDAL INPUT 2
y3[n ] = ∑ (13 )x[n − k ]
ƒ 3-point AVERAGER
ƒ INPUT x[n] = SUM of SINUSOIDS ƒ Changes A slightly
k =0
ƒ Then, OUTPUT y[n] = SUM of SINUSOIDS

2/25/2005 © 2003, JH McClellan & RW Schafer 8 2/25/2005 © 2003, JH McClellan & RW Schafer 9

3-pt AVG EXAMPLE 7-pt FIR EXAMPLE (AVG)

Input : x[n ] = (1.02)n + cos(2π n / 8 + π / 4) for 0 ≤ n ≤ 40 Input : x[n ] = (1.02)n + cos(2π n / 8 + π / 4) for 0 ≤ n ≤ 40

CAUSAL: Use Previous

USE PAST VALUES

2/25/2005 © 2003, JH McClellan & RW Schafer 10 2/25/2005 © 2003, JH McClellan & RW Schafer 11
LONGER OUTPUT
SINUSOIDAL RESPONSE DCONVDEMO: MATLAB GUI

ƒ INPUT: x[n] = SINUSOID

ƒ OUTPUT: y[n] will also be a SINUSOID
ƒ Different Amplitude and Phase
ƒ SAME Frequency

ƒ AMPLITUDE & PHASE CHANGE

ƒ Called the FREQUENCY RESPONSE

2/25/2005 © 2003, JH McClellan & RW Schafer 12 2/25/2005 © 2003, JH McClellan & RW Schafer 13

COMPLEX EXPONENTIAL COMPLEX EXP OUTPUT

x[n ] = Ae jϕ e jωˆ n −∞<n<∞ ƒ Use the FIR “Difference Equation”

M M
y[n ] = ∑ bk x[n − k ] = ∑ bk Ae jϕ e jωˆ ( n −k )
x[n] is the input signal—a complex exponential

k =0 k =0
M M
y[n ] = ∑ bk x[n − k ] = ∑ h[k ] x[n − k ] ⎛M ⎞
k =0 k =0 = ⎜⎜ ∑ bk e jωˆ ( − k ) ⎟⎟ Ae jϕ e jωˆ n
FIR DIFFERENCE EQUATION ⎝ k =0 ⎠

= H (ωˆ ) Ae jϕ e jωˆ n
2/25/2005 © 2003, JH McClellan & RW Schafer 14 2/25/2005 © 2003, JH McClellan & RW Schafer 15
FREQUENCY RESPONSE EXAMPLE 6.1
ƒ At each frequency, we can DEFINE {bk } = { 1, 2, 1 }
M
M H ( e jωˆ ) = 1 + 2e − jωˆ + e − j 2ωˆ
H (He (ωˆ))== ∑
EXPLOIT
∑bbkkee
jωˆ −−jω
jω
ˆˆkk FREQUENCY
SYMMETRY
RESPONSE = e − jωˆ ( e jωˆ + 2 + e − jωˆ )
kk==00
= e − jωˆ ( 2 + 2 cos ωˆ )
ƒ Complex-valued formula
ƒ Has MAGNITUDE vs. frequency Since ( 2 + 2 cos ωˆ ) ≥ 0
ƒ And PHASE vs. frequency Magnitude is H ( e jωˆ ) = ( 2 + 2 cos ωˆ )
ƒ Notation: H ( e jωˆ ) in place of H (ωˆ ) and Phase is ∠ H ( e jωˆ ) = −ωˆ
2/25/2005 © 2003, JH McClellan & RW Schafer 16 2/25/2005 © 2003, JH McClellan & RW Schafer 17

PLOT of FREQ RESPONSE

{bk } = {1,2,1} EXAMPLE 6.2

Find y[n ] when H ( e jωˆ ) is known

ω̂ and x[n ] = 2e jπ / 4e j (π / 3) n
H ( e jωˆ ) = ( 2 + 2 cos ωˆ )e − jωˆ RESPONSE at π/3

H ( e jω̂ )
x[n] y[n]

H ( e jωˆ ) = ( 2 + 2 cos ωˆ )e − jωˆ

−π π
2/25/2005
ω̂ (radians)
© 2003, JH McClellan & RW Schafer 18 2/25/2005 © 2003, JH McClellan & RW Schafer 19
 EXAMPLE 6.2 (answer) EXAMPLE: COSINE INPUT

Find y[n ] when x[n ] = 2e jπ / 4e j (π / 3) n Find y[n ] when H ( e jωˆ ) is known

One Step - evaluate H ( e jωˆ ) at ωˆ = π / 3 and x[n ] = 2 cos( π3 n + π4 )
H ( e jωˆ ) = ( 2 + 2 cos ωˆ )e − jωˆ x[n] jω̂ y[n]
H (e )
jωˆ − jπ / 3
H ( e ) = 3e @ ωˆ = π / 3
ω̂ ω̂
( )
y[n ] = 3e − jπ / 3 × 2e jπ / 4e j (π / 3) n = 6 e − jπ / 12 j (π / 3) n
e H ( e jωˆ ) = ( 2 + 2 cos ωˆ )e − jωˆ
2/25/2005 © 2003, JH McClellan & RW Schafer 20 2/25/2005 © 2003, JH McClellan & RW Schafer 21

EX: COSINE INPUT EX: COSINE INPUT (ans-2)

Find y[n ] when x[n ] = 2 cos( π3 n + π4 ) Find y[n ] when x[n ] = 2 cos( π3 n + π4 )
2 cos( πn+ 4 =e π) j (π n / 3+π / 4 )
+e − j (π n / 3+π / 4 ) H ( e jωˆ ) = ( 2 + 2 cos ωˆ )e − jωˆ
3
⇒ x[n ] = x1[n ] + x2 [n ] y1[n ] = H ( e jπ / 3 )e j (π n / 3+π / 4 ) = 3e − j (π / 3) e j (π n / 3+π / 4 )
jπ / 3 j (π n / 3+π / 4 ) y2 [n ] = H ( e − jπ / 3 )e − j (π n / 3+π / 4 ) = 3e j (π / 3) e − j (π n / 3+π / 4 )
Use
Linearity
y1[n ] = H ( e )e
y2 [n ] = H ( e − jπ / 3 )e − j (π n / 3+π / 4 ) y[n ] = 3e j (π n / 3−π /12 ) + 3e − j (π n / 3−π /12 )
⇒ y[n ] = y1[n ] + y2 [n ] ⇒ y[n ] = 6 cos( π3 n − 12 π )
2/25/2005 © 2003, JH McClellan & RW Schafer 22 2/25/2005 © 2003, JH McClellan & RW Schafer 23
MATLAB:
FREQUENCY RESPONSE Time & Frequency Relation

ƒ HH = freqz(bb,1,ww) ƒ Get Frequency Response from h[n]

ƒ VECTOR bb contains Filter Coefficients ƒ Here is the FIR case:

ƒ SP-First: HH = freekz(bb,1,ww) M M
H ( e ) = ∑ bk e
jωˆ − jωˆ k
= ∑ h[k ] e − jωˆ k
ƒ FILTER COEFFICIENTS {bk}
k =0 k =0
M
H ( e jωˆ ) = ∑ bk e − jωˆ k
IMPULSE RESPONSE
k =0
2/25/2005 © 2003, JH McClellan & RW Schafer 24 2/25/2005 © 2003, JH McClellan & RW Schafer 26

BLOCK DIAGRAMS UNIT-DELAY SYSTEM

ƒ Equivalent Representations Find h[n ] and H ( e jωˆ ) for y[n ] = x[n − 1]

x[n]
h[n ] y[n] x[n]
δ [n − 1] y[n]

{bk } = { 0, 1 }
H ( e jω̂ )
H ( e jω̂ )
x[n] y[n]
x[n]
− jω̂ y[n]
e
ωˆ ω̂ ωˆ ωˆ
2/25/2005 © 2003, JH McClellan & RW Schafer 27 2/25/2005 © 2003, JH McClellan & RW Schafer 28
 DLTI Demo with Sinusoids
FIRST DIFFERENCE SYSTEM

Find h[n ] and H ( e jωˆ ) for the Differenc e

x[n] FILTER y[n]

Equation : y[n ] = x[n ] − x[n − 1]

x[n] y[n]
δ [n ] − δ [n − 1]

H ( e jω̂ )
− jω̂ y[n]
1− e
x[n]

2/25/2005 © 2003, JH McClellan & RW Schafer 29 2/25/2005 © 2003, JH McClellan & RW Schafer 30

CASCADE SYSTEMS CASCADE EQUIVALENT

ƒ Does the order of S1 & S2 matter?
ƒ MULTIPLY the Frequency Responses
ƒ NO, LTI SYSTEMS can be rearranged !!!
ƒ WHAT ARE THE FILTER COEFFS? {bk} x[n]
H1 ( e jωˆ ) H 2 ( e jωˆ ) y[n]
ƒ WHAT is the overall FREQUENCY
RESPONSE ?
x[n]
H ( e jω̂ ) y[n]

H ( e jωˆ ) = H1 ( e jωˆ ) H 2 ( e jωˆ )

EQUIVALENT
δ [n ] h1[n ]
S1 h1[n ] h2 [n ]
S2 h1[n ] ∗ h2 [n ] SYSTEM

2/25/2005 © 2003, JH McClellan & RW Schafer 31 2/25/2005 © 2003, JH McClellan & RW Schafer 32