UNIT 1. CIRCUIT ANALYSIS IN FREQUENCY DOMAIN.

RESONANT CIRCUITS AND FILTERS

1. Frequency response of basic circuits

2. Transfer functions

3. Bode diagrams

4. Resonant circuits

5. Filters (passive and active)

INTRODUCTION

Although time-domain analysis is important, the key characteristics of a signal or a

communication system are easily extracted, studied or designed in the frequency domain.

Examples of signal:

• Electric supply: constant frequency (50 Hz).

• Audio signal: variable frequency 20 Hz – 20 kHz

• Telephone signal (local loop): variable frequency 300 – 3400 Hz

The signal analysis requires the knowledge of the frequency behavior of the circuits
employed in its emission / reception / processing.

This is why you have seen phasors and sinusoidal steady state (Systems and Circuits),
Fourier series and transform (Linear Systems) and, in general, complex numbers.

Telecommunication is a frequency matter, rather than a time matter.

As telecommunication engineers you have to think mostly in frequency domain.

1.1. FREQUENCY RESPONSE OF BASIC CIRCUITS

• Resistance
"! ! ("! )

𝑅 𝑍! = 𝑅 = 𝑅𝑒 $% ! !

𝑍!
! !
• Inductor
"! 𝑍! 0 = 0 (s.c.) ! ("! )
& 𝑍! ∞ → ∞ (o.c.)
$'
𝐿 𝑍" = 𝑗𝜔𝐿 = 𝜔𝐿𝑒 !
!
𝑍"
! !
• Capacitor
"! ! ("! )
𝑍" 0 → ∞ (o.c.)
𝐶 1 −𝑗 1 ($& 𝑍" ∞ =0 (s.c.) !
𝑍# = = = 𝑒 ' "
𝑗𝜔𝐶 𝜔𝐶 𝜔𝐶 !
𝑍#

! !
RLC SERIES CIRCUIT
𝐶 𝐿 1
𝑍)* = 𝑅 + 𝑗𝜔𝐿 +
𝑗𝜔𝐶
1
𝑅 When 𝜔 → 0 ⟹ 𝑍)* → ⟹ 𝑍)* depends mainly on the capacitor
𝑗𝜔𝐶
When 𝜔 → ∞ ⟹ 𝑍)* → 𝑗𝜔𝐿 ⟹ 𝑍)* depends mainly on the inductor
𝑍)*
𝑍)* expression complexity increases with the circuit. In order to reduce this complexity we
can work with 𝑠 = 𝑗𝜔: 1 𝑠 ' 𝐿𝐶 + 𝑠𝑅𝐶 + 1
𝑍)* = 𝑅 + 𝑠𝐿 + =
𝑠𝐶 𝑠𝐶
In general (any circuit), the input impedance 𝑎+ 𝑠 + + 𝑎+(, 𝑠 +(, + ⋯ + 𝑎, 𝑠 + 𝑎%
(or any other relationship) can be written as: 𝑍)* = 𝑏 𝑠 - + 𝑏 𝑠 -(, + ⋯ + 𝑏 𝑠 + 𝑏
- -(, , %
10 1.5

1
8
1

phase(Zin) (rad)
𝑅 = 1Ω
abs(Zin) (ohm)

0.5
6
𝜔= →𝑍=𝑅
𝐿𝐶
𝐿 = 1H 0

𝐶 = 1F
4
−0.5

2 −1

0 −1.5
0 1 2 3 4 5 0 1 2 3 4 5
ω ω
1.2. TRANSFER FUNCTIONS
#"
Voltage gain $# =
#!
#
Current gain $# = "
Transfer functions #!
$
Transimpedance %= "
#!
$"
Transadmittance % =
#!
𝐶
𝐿
𝑖" 𝑖!
1 1.5
"! +
–
𝑅 "! 1
0.8
0.5

phase(Gv)
0.6
abs(Gv)

0
𝑅=𝐿=𝐶=1 0.4
−0.5
0.2
' −1
#" = #!
& ! # 0 −1.5
' + C $ (% ' !
0 1
1
2
ω
3 4 5 0 1 2
ω
3 4 5

% ($ " 𝜔=
𝐿𝐶
EXAMPLE

𝑅# 𝐶

+
𝑣# 𝑖# 𝐿 𝑖$ 𝑅$ 𝑣$
–

𝑖' 𝐿𝐶𝑠 '

Check: 𝐺. = =
𝑖, 𝐿𝐶𝑠 ' + 𝑅' 𝐶𝑠 + 1

𝑣' 𝑅' 𝐿𝐶𝑠 '

𝐺/ = =
𝑣, 𝑅, + 𝑅' 𝐿𝐶𝑠 ' + 𝐿 + 𝑅, 𝑅' 𝐶 𝑠 + 𝑅,

𝑣' 𝑅' 𝐿𝐶𝑠 '

𝑍= =
𝑖, 𝐿𝐶𝑠 ' + 𝑅' 𝐶𝑠 + 1

𝑖' 𝐿𝐶𝑠 '

𝑌= =
𝑣, 𝑅, + 𝑅' 𝐿𝐶𝑠 ' + 𝐿 + 𝑅, 𝑅' 𝐶 𝑠 + 𝑅,
POLES AND ZEROS

In general, if " = !! , the transfer function will be:

#& " & + #&!"" &!" + ### + #"" + #!
$ (" ) = ai , bi real
!% " % + !% !"" % !" + ### + !"" + !!

& ($ ! %# )($ ! %" )!!!($ ! %" )

' ($ ) = zi , pi complex
($ ! ## )($ ! #" )!!!($ ! #! )

𝐾 constant

"! zeros of the transfer function # ("! ) = !

"! poles of the transfer function # ( "! ) = !

Poles and zeros of the transfer function describe the behavior of the circuit in
the frequency domain.
LINEAR REPRESENTATION IN FREQUENCY
" ( !! ) = " ( !! ) # # !" (" ( !! ))
RLC circuit: Phase
𝐶 Magnitude
𝐿
Magnitude Phase
1 1.5
"! +
–
𝑅 "! 1
0.8
0.5

phase(Gv)
0.6
abs(Gv)

0
'
#" = #! −0.5
& ! # 0.4
' + C $ (% ' ! −1
% ($ " 0.2
−1.5
𝑅=𝐿=𝐶=1 0 −2
0 20 40 60 80 100 0 20 40 60 80 100
ω ω

• For large frequency ranges, the detail at low frequencies is small.

• For large magnitude ranges, the detail at low magnitudes is small.
• Polynomial terms at numerator and denominator hinder the magnitude representation.

⇒ A different graphic representation is needed ⇒ Bode diagram

PREVIOUS CONCEPTS. LOGARITHMIC UNITS (I)

• Power attenuation or gain in linear !"

units: !!
• Power attenuation or gain in !*
logarithmic units (decibels) () B&'() !"#$
!(

Linear 0.1 1 10 100 1000

dB -10 0 10 20 30

• For signal: "$!

𝑅 "! #$ =
!

( (
#( ! ," !( !(
+) B&'+) !"#$ = +)*B&'+) = +)*B&'+)
(
(
= ()B&'+) !"#$
#+ ! ," + !+ !+
PREVIOUS CONCEPTS. LOGARITHMIC UNITS (II)

• Decade: interval whose upper edge is 10 times the lower edge.

• Octave: interval whose upper edge is double of the lower edge.

upper edge upper edge

decade # = log10 octave # = log 2
lower edge lower edge

3 dec. 3 oct.
2 dec. 2 oct.
1 dec. 1 dec. 1 oct. 1 oct.
0.5 5 50 500 5000 0.5 1 2 4 8
1.3. BODE DIAGRAM

Hendrik W. Bode, years 30 – 40, employs this type of diagrams for the
analysis and design of frequency-dependent systems.

• Magnitude graph:
• Ordinates in dBs: &" #$%!" " ( !! )
• Abscissas in logarithmic scale (decades).
• Phase graph:
#$(" ( !! ))
• Ordinates in radians: " (" ( !! )) = %&'(%)
!"(" ( !! ))
• Abscissas in logarithmic scale (decades).

Re 𝑧
cos 𝛼 =
𝑧
Note: the phase of a complex 𝑧 is the angle 𝛼 which accomplish both
Im 𝑧
Im 𝑧 sin 𝛼 =
This is not exactly the same that tan#$ 𝑧
Re 𝑧
See difference between atan and atan2 in Matlab.
ADVANTAGES

• In the magnitude, the power terms turn into linear terms, and products into sums.

!"#(! " ) = " !"#(! ) !"#("!) = !"#(" ) + !"#( ! )

!!
$ !"# !
!
!"#( ! ) $ !"# !
!
$ !"# !
!!

• It allows the representation of a wide frequency range without loosing the

detail at low frequency.

10 KHz 10 KHz
1 KHz 100 KHz 1 KHz 100 KHz
FACTORIZATION OF H(j𝜔) IN POLES AND ZEROS (I)
• Any transfer function can be expressed as:

D
' *) $ % ' *) ' *) $ $"
"

% # ( ( *) ) ( ! %%! + %
"" ( ! ! + "* &
±)
+ %% ""
( =! & ) ( # & =!
% ) & & # "#
) ω c ≡ cut-off frequency
H ( *) ) = & &

*) $ ! '% *) ' *) $ $"

"
#
' ξ c ≡ damping coefficient
! %%! +
$ =! &
"" (! ! + "* "
) $ # " =! %
+ %%
)" & )" # "
""
& #

Terms:
• Gain (independent of frequency) !!
• Poles and zeros in the origin (𝜔 = 0) ( "! )± !
• Simple (first order) poles and zeros "!
!+
& #' #
! !!
• Poles and zeros of order 𝑛 $$! + !!
% '" " !' & !' #
!
• Quadratic poles and zeros # + !(" + $$ !!
'" % '" "
FACTORIZATION OF H(j𝜔) IN POLES AND ZEROS (II)

Key idea: take advantage of the properties of the logarithm and the arctangent, such
that the final graphs are obtained as the sum of the contributions of the different
terms in the transfer function.
D
' *) $ % ' *) ' *) $
"
$
% # ( ( *) ) ( ! %%! + "" ( ! %! + "* & "" "
±)
+ %%
( =! & )( # & =! % )& & )& # "
H ( *) ) = & #
*) $ ! '% *) ' *) $ $"
"
#
'
! %%! +
$ =! &
"" (! ! + "* "
) $ # " =! %
+ %%
)" & )" # "
""
& #

Magnitude:
"
))
D
)) ' )) $
%
"# $%& H ( )) ) = "# $%& %# ± "# * $%& )) + "#! $%& ! + + "# ! $%& ! + "* & + %% "" (
( =! )( & =! )& & )& #
"
#
)) !
)) ' )) $
("#! $%& ! + ( "#! $%& ! + "* " + %% ""
$ =! )$ " =! )" & )" #
FACTORIZATION OF H(j𝜔) IN POLES AND ZEROS (III)

% ' $
"
D
' *) $ *) ' *) $
% # ( ( *) ) ( ! %%! + %
"" ( ! ! + "* & "" "
±)
+ %%
( =! & )( # & =! % )& & )& # "
H ( *) ) = & #
! ' $
"
#
' *) $ *) ' *) $
! %%! + "" (! %! + "* " + %% "" "
$ =! & ) $ # " =! % )" & )" # "
& #

Phase: ' $
% "* ) "
+ # ') $ ! % "
)" "
, (' ( &) )) = $%&'$( (! # ) ± % + ( $%&'$( %% "" + ( $%&'$( % " "!
" $ =! & )$ # " =! % ! ! '% ) $" "
% %) " "
& & "# #
' $
% "* ) "
#
') $ ! % "
)" "
( ! #$%&#' %% "" ( ! #$%&#' % " "
$ =! )
& $ # " =! % ! ( '% ) $" "
% %) " "
& & "# #
CONSTANT TERM

! !

%! "#$ !! ! (!! )

! !
POLES AND ZEROS AT THE ORIGIN

General form: ( "! ) ±!

$% " !"# !!
! !
20N dB/dec. !
Zero 0 !
!

0.1 1 10 ! 0.1 1 10 !

" $% " !"# !!

! !

!
Pole 20N dB/dec. "!
!

0
0.1 1 10 ! 0.1 1 10 !
SIMPLE ZEROS. MAGNITUDE
& "' #
General form: $$! + !!
% '! "
! ""
<< ! !+ !! (# B&' $ ! # !"
Asymptotic !! "!
approximations  "! "! "!
>> ! !+ " &' #$B !" (simple zero at the origin)
! !! !! !!
Asymptotes intersection: ω = ωi ≡ cut-off frequency $( B&' ) + ! = $( B&' $ ! # !"

! 20 dB/dec.
(6 dB/oct.)
1 oct. above:

3dB $) &'(B + ! $ ! # !"

0
1 oct. below:
$) &'( # + ! B $ ! #!"

!! ! ! ! !!! !
SIMPLE POLES. MAGNITUDE
!!
General form: '%! + "( $"
% (! "#
&
(!
! ' ") $
<< ! %%! + "" ! ! " (# B&'$ ! # !"
Asymptotic !! & )! #
!! !!
approximations  ' ") $ ' ") $ "!
>> ! %%! + "" ( %% "" " &' #$B !" (simple pole at the origin)
! & )! # & )! # !!

Asymptotes intersection: ω = ωi ≡ cut-off frequency ! $( B&' ) + ! = !$( B&' $ " !# !"

0 3dB

-20 dB/dec.
(-6 dB/oct.)

!! ! ! ! !!! !
SIMPLE ZEROS AND POLES. PHASE
&& "' #!
#
±!
& "' #
$
Phase: ( $$! + ! = ±( $$! + !!
$% '! !" ! % ' ! "
% "
ω ω
<< 1 (e.g. ω = i ) " ! % !"#$
Asymptotic ωi 10 π
φ ≈ rad. (zero)
approximations  2
>> ! (e.g. ω = 10ω i ) π
! φ ≈ − rad. (pole)
2
! Zero ! Pole
!
!
!

π π
rad./dec. − rad./dec.
4 4
! !
"
! !

!
! "
!#!"!! #"!!! !! !"!! !""!! ! !!#!"! #"!!! !! !"!! !""!! !
!
MULTIPLE ZEROS AND POLES
±!
& #' #
General form: $$! + !!
% '" "

Same graphs as simple zeros and poles, but:

Slope: 20N dB/dec.

Magnitude
Deviation of 3N dB at the cut-off frequency

!
Slope: N rad./dec.
! !
Phase Zero: from 0 up to N
!
!
Pole: from 0 up to -N
!
QUADRATIC ZEROS AND POLES
2
jω ! jω $
General form: 1+ 2ξ + # & , ξ ≡damping coefficient
ω0 " ω0 %

! "# ± ## " ! #
! + "#! + ! = $ " ! =
"
= !# ± # " ! !
"
Three cases:
If ξ > 1 → real and different roots → simple zeros and poles
If ξ = 1 → real and equal roots → multiple (double) zeros and poles
If ξ < 1 → complex conjugate roots

! ! (# B&' $ = # !"
<< !
Asymptotic !"
"
approximations  '! $ ! 40 dB/dec. (zero)
>> ! ( "! #$% %% "" = &! #$%
" & !! # !! -40 dB/dec. (pole)

Between these two points the function depends on x.

QUADRATIC ZEROS AND POLES. MAGNITUDE

Zero Pole
! 40 dB/dc.
!
! = #"!
(12 dB/oct.)
! = #"!
!" !" ! =!
! =!

! ! =! !
! =!
-40 dB/dec.
! !" ! !" ! = #"!
! = #"! (-12 dB/oct.)

!#!"!! !#"!! !! "!!! ! !#!"!! !#"!! !! "!!! !

2
" jω %
For ξ = 0 → 1+ $ ' = 1−1 = 0
# ω0 &
ω =ω 0
QUADRATIC ZEROS AND POLES. PHASE
& !)( # (" #
Phase: * = %&'(%)$$ !
! !
% $ ' (( # (" ) "

! Zero ! Pole
! !
! =!
! =!
! = #"!
! = #"!
! = #"!
! = #"!
! =!
! =!
! !
"
! !
π π
rad./dec.
2 − rad./dec.
2

! "!
!#!"!! !#"!! !! "!!! ! !! "!!! !
!#!"!! !#"!!

# & 0 for ω < ω 0

0
For ξ = 0 → arctan %% (
2( undetermined for ω = ω 0
$ 1− (ω / ω 0 ) ' ±π for ω > ω 0
EXAMPLE (I)

Draw the Bode diagrams of magnitude and phase for the following transfer function:

10 + jω
Gv ( jω ) =
1+ j1.02ω − 0.02ω 2

1. Decomposition of the transfer function. Search for roots:

10 (1+ j0.1ω )
Gv ( jω ) =
(1+ jω ) (1+ j0.02ω )
2. Term identification:

K 0 = 10 → constant
jω
1+ j0.1ω → simple zero, j0.1ω = →ω c = 10 rad./s
ωc
1+ jω → simple pole, ω c = 1rad./s
1+ j0.02ω → simple pole, ω c = 50 rad./s
EXAMPLE (II)

A) Magnitude

Constant: K 0 = 10 = 20 dB
Simple zero: 1+ j0.1ω →20 log 1+ j0.1ω : ω << 10 →0 dB
ω >> 10 →20 log j0.1ω →Slope = 20 dB/dec.
!! = '( !"#$%&

Simple pole: 1+ jω →20 log 1+ jω : ω << 1→0 dB

ω >> 1→− 20 log jω →Slope = −20 dB/dec.
!! = ' !"#$%&

Simple pole: 1+ j0.02ω : ω << 50→0 dB

ω >> 50→− 20 log j0.02ω →Slope = −20 dB/dec.
!! = '( !"#$%&
EXAMPLE (III)
Slope table (dB/dec)
Simple pole Simple zero Simple pole Total slope
w
! + !! 1+ j0.1ω 1+ j0.02ω

" <! <! 0 0 0 0

! < ! < !" -20 0 0 -20

#" < ! < !" -20 20 0 0

 > !" -20 20 -20 -20

! !"#$
!"

! !"

0.1 ! !" !" !"" !"" !""" ! !"#$%&'(

EXAMPLE (IV)
B) Phase
Constant: φ ( K 0 ) = 0 rad.
zero: 1+ j0.1ω →arctan ( 0.1ω ) : ω << 10 →0 rad.
Simple cero:
!! = '( !"#$%& π
ω >> 10 → rad.
2
π
ω = 10→arctan (1) = rad.
4
Simple pole: 1+ jω →− arctan (ω ) : ω << 1→0 rad.
Polo simple:
!! = ' !"#$%& π
ω >> 1→− rad.
2
π
ω = 1→− rad.
4
Simple pole: 1+ j0.02ω →− arctan ( 0.02ω ) : ω << 50→0 rad.
Polo simple:
!! = '( !"#$%& π
ω >> 50→− rad.
2
π
ω = 50→− rad.
4
EXAMPLE (V)
Slope table (rad/dec)
Simple pole Simple zero Simple pole Total slope
w
! + !! 1+ j0.1ω 1+ j0.02ω
0 < ω < 0.1 0 0 0 0
0.1 < ω < 1 -p/4 0 0 -p/4
"<! < ! -p/4 p/4 0 0
# < ! < !" -p/4 p/4 -p/4 -p/4
!" < ! < !"" 0 p/4 -p/4 0
#"" < ! < !"" 0 0 -p/4 -p/4
 > !"" 0 0 0 0

! % !"#$
!

!
"
!

!
"
! 0.01 0.1 ! ! !" !"" !"" ! !"#$%&'(
ASSIGNMENT
• Draw the Bode diagrams of magnitude and phase for the following transfer function:

1+ jω
Gv ( jω ) = 2
! ω $ ! ω ω2 $
jω #1+ j & #1+ j − &
" 0.5 % " 2 4 %
20 4

0
2

phase(Gv) (rad)
abs(G ) (dB)

−20
0
v

−40

−2
−60

−80 −1 0 1
−4 −1 0 1
10 10 10 10 10 10
ω (rad/s) ω (rad/s)

• Problems 1.n), 2.a) and 2.f)

SUMMARY UP TO THIS POINT

- Resistances, inductors and capacitors define the frequency behavior of the circuit.
- Poles and zeros of the transfer function® influence the frequency behavior of the circuit.
- Graphical representation:
- Linear representation:
- Little detail for low frequencies and small magnitudes.
- Polynomial terms in numerator and denominator ® difficult drawing.
- Bode diagrams ® Magnitude: () B&' " ( !! ) !"#$ , w in logarithmic scale.
® Phase: " (" ( !! )) , w in logarithmic scale.
- Advantages: easier representation (linear approximations), higher resolution at low
frequencies and small magnitudes.
- Terms identification in the transfer function. Global behavior = sum of partial behaviors.
- Influence of each term:
- cut-off frequency
- slope
- damping coefficient.
- Example: slopes table® linear-wise plot.
1.4. RESONANT CIRCUITS

Resonance phenomenon: circuits, bridges, swing, skyscrapers… 𝑖

Resistances consume (dissipate) energy. RLC
𝑣 network
Inductors and capacitors store energy.

𝑖 𝑡 = 𝐼% cos 𝜔𝑡 0.5
1 ' 1 ' '
𝐿 𝑖 𝑊" 𝑡 = 𝐿𝑖 𝑡 = 𝐿𝐼% cos 𝜔𝑡 0.4
2 2
1 '

WL or WC (J)
𝑊" 𝑡 U = 𝐿𝐼% 0.3
012 2
𝑣3 𝑡 = 𝑉% cos 𝜔𝑡 0.2
1 1
𝐶 𝑣3 𝑊# 𝑡 = 𝐶𝑣3' 𝑡 = 𝐶𝑉%' cos' 𝜔𝑡 0.1
2 2
1 '
𝑊# 𝑡 U = 𝐶𝑉% 0
012 2 0 2 𝑇+
2
4 6 8 10
t (s)

Resonance occurs when the maximum energy stored in inductors equals the maximum
energy stored in capacitors. This can take place at a single frequency or at multiple
frequencies.
𝑊" U = 𝑊# U
012 012
RESONANCE AND QUALITY FACTOR
𝑊& WL
0.5
WC
At resonant frequency: 0.4

𝑍)* ∈ ℝ → 𝑍)* |=>=% = 𝑅)* , 𝑌)* |=>=% = 𝐺)*

(J)
0.3

𝑣)* and 𝑖)* are in phase → power factor: cos 𝜑 = 1 0.2

0.1
The stored energy in the circuit is constant
0
0 2 4 6 8
t (s)

Storing energy has a cost in terms of energy dissipation due to unavoidable resistive
components.
Quality factor (of a circuit or an individual component): is a measure of the efficiency for
storing energy when it is run on alternating current.
RLC series circuit
1
𝑊012.56789: 𝑊012 𝑊012
𝑄 ≡ 2𝜋 = 2𝜋 =𝜔 R=1
𝑊:)55);169: )* , ;98)7: 𝑃0𝑇 𝑃0
0.8 R=0.2

2𝜋 0.6

abs(Gv)
𝜔=
𝑇
0.4
It depends on 𝜔
0.2
In a resonant circuit, the bandwidth depends on the quality factor
0 −1 0 1
10 10 10
ω (rad/s)
QUALITY FACTOR
𝑄!

Inductor: 𝑖 = 𝐼% cos 𝜔𝑡 1 '

𝐿 𝑊012 𝐿𝐼% 𝜔𝐿
𝑄" = 𝜔 =𝜔 2 =
𝑃0 1 ' 𝑅
𝑅𝐼% !
𝑅 2
𝑄!'
𝑉% 1 𝑉%'
𝑖= sin 𝜔𝑡
𝜔𝐿 @ 2 𝜔 ' 𝐿 𝑅?
𝑣 = 𝑉% cos 𝜔𝑡 𝑅? 𝐿 𝑄" = 𝜔 =
𝑉 𝑉( )*+
1 𝑉%' 𝜔𝐿
𝐼⃗ = = 𝑒 $ 2 𝑅? !
𝑗𝜔𝐿 𝜔𝐿

Capacitor:
1 '
2 𝐶𝑉%
𝑣 = 𝑉% cos 𝜔𝑡 𝑅? 𝐶 𝑄# = 𝜔 = 𝜔𝐶𝑅?
1 𝑉%'
2 𝑅?
𝑉%
𝑖= cos 𝜔𝑡 − 𝜑 1 𝑉%' 1
𝐶 𝑅' +
1 2𝐶 1 𝜔'𝐶 '
𝜔'𝐶 ' 𝑅' + 1
𝑣 = 𝑉% cos 𝜔𝑡 𝑄#@ =𝜔 𝜔'𝐶 ' =
𝑅 1 𝑉%' 𝜔𝑅𝐶
𝑅 1
2
𝑅' + ' '
𝜔 𝐶
ASSIGNMENT

Check that
𝐿
1 1 1 1
𝑅? = + 1+ '
𝑄 𝑄" 𝑄"@ 𝑄"
𝑅

𝐶 1 1 1 1
𝑅? = @ + 1 + @'
𝑅 𝑄 𝑄# 𝑄# 𝑄#
RESONANCE AT RLC SERIES CIRCUIT
a 𝑖 = 𝐼' cos 𝜔𝑡 − 𝜑 𝜔% ?

𝐿 1 ' ' 1 '

𝑊" = 𝐿𝐼% cos 𝜔𝑡 − 𝜑 → 𝑊" U = 𝐿𝐼%
2 012 2
𝑣%& = 𝑉' cos 𝜔𝑡 𝑅 1 𝐼%' 𝜋 1 𝐼%
'
𝑊# = 𝐶 ' ' cos' 𝜔𝑡 − 𝜑 − → 𝑊# U =
𝐶 2 𝐶 𝜔 2 012 2 𝐶𝜔 '
$
1 $ 1 1 )*+
b 𝐶𝑣 = 𝐶 𝐼(𝑒 )*-
: 𝑒 $
2 , 2 𝜔𝐶
𝑍1A
1 ' 1 𝐼%' 1 1
At resonance: 𝑊" U = 𝑊# U → 𝐿𝐼% = → 𝜔% = → 𝑓% =
012 012 2 2 𝐶𝜔%' 𝐿𝐶 2𝜋 𝐿𝐶
𝑄?
𝑊B = 𝑊" + 𝑊# sin$𝛼 + cos$𝛼 = 1
1 ' ' 1 𝐼%' '
1 '
𝑊B U = 𝐿𝐼% cos 𝜔% 𝑡 − 𝜑 + sin 𝜔% 𝑡 − 𝜑 = 𝐿𝐼% = 𝑊" U = 𝑊# U
=% 2 2 𝐶𝜔%' 2 012 012

1 '
𝑊012 2 𝐿𝐼% 𝜔% 𝐿 1
𝑄 = 𝜔% = 𝜔% = = = 𝑄" = 𝑄#@
𝑃0 1 ' 𝑅 𝜔% 𝑅𝐶
𝑅𝐼
2 %
IMPEDANCE AROUND RESONANCE

Looking for a relationship between the input impedance and the resonance parameters (𝜔% , 𝑄)

1 𝑗 1 𝑗𝐿𝜔+ 𝜔 1 𝜔 𝜔+
𝑍)* = 𝑅 + 𝑗 𝜔𝐿 − =𝑅 1+ 𝜔𝐿 − =𝑅 1+ − = 𝑅 1 + 𝑗𝑄 −
𝜔𝐶 𝑅 𝜔𝐶 𝑅 𝜔+ 𝜔𝐿𝐶𝜔+ 𝜔+ 𝜔
1 𝐿𝜔( 1
𝜔( = ,𝑄 = =
𝐿𝐶 𝑅 𝐶𝜔(𝑅
𝜔 − 𝜔%
Relative frequency deviation: 𝛿 ≡
𝜔%

1 2𝛿 + 𝛿 ' 𝛿+2
𝑍1A = 𝑅 1 + 𝑗𝑄 1 + 𝛿 − = 𝑅 1 + 𝑗𝑄 = 𝑅 1 + 𝑗𝑄𝛿
1+𝛿 1+𝛿 𝛿+1

Close to resonance (𝜔 → 𝜔% ): 𝛿 ≪ 1 → 𝑍1A ≈ 𝑅 1 + 𝑗2𝑄𝛿 = 𝑅 + 𝑗𝑋, 𝑋 = 2𝑄𝛿𝑅

When 𝜔 > 𝜔% → 𝛿 > 0 → 𝑋 > 0 → inductive behavior

When 𝜔 < 𝜔% → 𝛿 < 0 → 𝑋 < 0 → capacitive behavior
 CONNECTION TO A REAL VOLTAGE GENERATOR
1 𝜔' 𝐿 + 𝐿
𝜔' = 𝑄= 𝐸C –
𝐿𝐶 𝑅 + 𝑅(
𝐸( 𝐸( 𝐼 𝑅
𝐼= = 𝑅C
1 𝐿𝜔' 𝜔 1
𝑅 + 𝑅( + 𝑗 𝜔𝐿 − 𝜔𝐶 𝑅 + 𝑅( 1 + 𝑗 𝑅 + 𝑅 − 𝐶
( 𝜔' 𝜔𝐿𝐶𝜔'
𝑄
1
𝐸( 𝐼 1 Q=50
= Q=100
𝑅 + 𝑅( 𝐼)*+ 𝐼)*+ 𝐼)*+ 1 + 4𝑄, 𝛿 ,
0.8 Q=200
𝐼= = ≈
𝛿+2 𝛿 + 2 1 + 𝑗2𝑄𝛿 𝐼 0.6
1 + 𝑗𝑄𝛿 1 + 𝑗𝑄𝛿 𝜑 = − tan#$ 2𝑄𝛿

|I/Ires|
𝛿+1 𝛿+1 𝐼)*+ 0.4
𝛿≪1
0.2
Universal resonance curve
0
−0.04 −0.02 0 0.02 0.04
1 1.5 δ
2
1 1 Q=50
"#
0.8 − tan 2𝑄𝛿 Q=100
1 + 4𝑄! 𝛿 ! 0.5 1 Q=200
0.6
φ(I/Ires)
|I/Ires|

φ(I/Ires)
0.4 0
−0.5
0.2 −1 −1

0 −1.5
−4 −2 0 2 4 −4 −2 0 2 4
Qδ Qδ −2
−0.04 −0.02 0 0.02 0.04
δ
 BANDWIDTH (dB) 1 𝑃/01 dB

X-dB bandwidth of a circuit: the frequency range around the 0.8

𝑃/01 − 𝑋 dB
resonance frequency in which the dissipated power, expressed in dB, 0.6
is greater than or equal to the power in resonance minus X dB. 𝐵.
0.4

𝐵- = 2 𝜔- − 𝜔' such that 𝑃 dB ≥ 𝑃)*+ (dB) − 𝑋 dB 𝜔

0.2
−0.04 −0.02 𝜔0 ( 𝜔. 0.02 0.04

1 𝑃 𝑃 #-
Linearly: 𝑃 = 𝑅 𝐼 , 𝑃 𝐼, By definition: 10 log $' ≥ −𝑋 dB → ≥ 10 $'
2 𝑃)*+ 𝑃)*+
K =
1 , 𝑃)*+ 𝐼)*+ , #. 1
𝑃)*+ = 𝑅 𝐼)*+
2 Usually 𝑋 = 3 dB → 10 $' ≈
2

Looking for a relationship between 𝐵D and the resonance parameters (𝜔% , 𝑄)

𝐼 #- 𝐼
Working with signals: ≥ 10 ,' 𝐼/01 1 1
𝐼)*+
𝐼 1 #- , 1 1 - 0.8

= → 10 ,' = → 𝛿- = ± 10 − 1
$'
1
𝐼)*+ 1 + 4𝑄, 𝛿 , 1 + 4𝑄, 𝛿 , 2𝑄 𝛿. = ± 0.6
2𝑄
0.4

𝜔- − 𝜔' 𝜔' - 𝛿
𝛿- = → 𝜔- − 𝜔' = 𝛿- 𝜔' → 𝐵- = 2𝛿- 𝜔' → 𝐵- = 10$' − 1 𝜔' 0.2
−0.04 −0.02−𝛿. 00 𝛿. 0.02 0.04
𝜔' 𝑄 𝐵. =
𝑄
RESONANCE AT RLC PARALLEL CIRCUIT
a

𝑣%& = 𝑉' cos 𝜔𝑡 𝑅? 𝐿 𝐶 𝐼C ↑ 𝑅C 𝑉 𝑅? 𝐿 𝐶

b 𝑍1A
Assignment: Following the same steps that series RLC, check in the parallel RLC that:
1 1
𝜔' = 𝜔' =
𝐿𝐶 𝐿𝐶
𝑅/ 𝑅/ T𝑅(
𝑄 = 𝜔' 𝐶𝑅/ = = 𝑄0 = 𝑄12 𝑄 = 𝜔' 𝐶 𝑅/ T𝑅( =
𝜔' 𝐿 𝜔' 𝐿
𝑅/ 𝑅/ 𝑉)*+ 𝑉)*+ 𝜔' 3
𝑍%& = 𝑉= ≈ 𝐵- = 10$' − 1
𝜔 𝜔' = 𝛿+2 𝛿 + 2 1 + 𝑗2𝑄𝛿 𝑄
1 + 𝑗𝑄 𝜔 − 𝜔 1 + 𝑗𝑄𝛿 1 + 𝑗𝑄𝛿
' 𝛿+1 𝛿+1
𝐸(
𝑉1 U = 𝑗𝜔' 𝐿𝐼)*+ = 𝑗𝜔' 𝐿 = 𝑗𝑄𝐸(
)*+ 𝑅 + 𝑅(
RLC series: 𝑄 ≡ over−voltage factor
𝐼)*+ 1 𝐸(
𝑉0 U = −𝑗 = −𝑗 = −𝑗𝑄𝐸(
)*+ 𝜔' 𝐶 𝜔' 𝐶 𝑅 + 𝑅(
RLC parallel: 𝐼0 U = 𝑗𝑄𝐼( , 𝐼1 U = −𝑗𝑄𝐼( → 𝑄 ≡ over−current factor
)*+ )*+
ASSIGNMENT

• Obtain R, L and C in the circuit for obtaining a resonance (𝑣E ) at 1000 rad/s
with a bandwidth of 100 rad/s.

𝐶
𝐿

"! +
–
𝑅 "!

• Problems 6, 9 and 11.

1.5. FILTERS

The frequency response of electric circuits allows us to obtain selective properties.

Electric filter: circuit designed for obtaining a gain and/or loss of magnitude in a given range
of frequencies.

Ideal filters:

low-pass high-pass band-pass band-reject

1 1 1 1

𝜔, 𝜔 𝜔, 𝜔 𝜔# 𝜔$ 𝜔 𝜔# 𝜔$ 𝜔

It is impossible to design/manufacture ideal filters.

1.5.1. PASSIVE FILTERS (I)

Without active components (amplifiers) ⇒ no gain, just attenuation.

𝑅
Example: single low-pass filter
+
1 1 𝑣2
–
𝐶 𝑣3
𝐺/ 𝑗𝜔 = → 𝜔3 =
1 + 𝑗𝜔𝑅𝐶 𝑅𝐶

1 𝐺4
𝐺/ 𝑗𝜔 = 11
𝜔 '
1+ 𝜔 10.8
3 0.6
2
𝜔
= − tan(,
0.4

𝜑 𝐺/ 𝑗𝜔
𝜔3 0.2

0 −2
10
𝜔,
0
10
2
10 𝜔

Bode diagram: 𝐺4 (dB) 𝜑 𝐺4 (rad)

00 00
−10
Single pole 𝜋 −0.5

−20 −
4−1
−30
𝜋
−40
𝜔, 𝜔 − −1.5 𝜔, 𝜔
2
−50 −2 0 2 −2 −2 0 2
10 10 10 10 10 10
1.5.1. PASSIVE FILTERS (II)
𝐶
Example: single high-pass filter
𝑣2 +
𝑅 𝑣3
–
𝑗𝜔𝑅𝐶 1
𝐺/ 𝑗𝜔 = → 𝜔3 =
1 + 𝑗𝜔𝑅𝐶 𝑅𝐶
𝐺4
𝜔 11
𝜔3 10.8
𝐺/ 𝑗𝜔 =
𝜔 ' 0.6
2
1+ 𝜔 0.4
3 0.2

𝜋 𝜔
0 −2
10
𝜔,
0
10
2
10
𝜔
𝜑 𝐺/ 𝑗𝜔 = − tan(,
2 𝜔3

Bode diagram: 𝐺4 (dB) 𝜑 𝐺4 (rad)

2

𝜋
00 1.5
2
−10
𝜋1
Single zero at the origin −20
40.5
and single pole at 𝜔3 −30

−40
𝜔, 𝜔 0 100 −2
𝜔,
0
10
2
10
𝜔
−50 −2 0 2
10 10 10
 𝐿 𝐶
1.5.1. PASSIVE FILTERS (III)
𝑣2 +
𝑅 𝑣3
–
Example: single band-pass filter zero at the origin

𝑅 𝑗𝜔𝑅𝐶 𝐺4
𝐺4 𝑗𝜔 = =
1 1 + 𝑗𝜔𝑅𝐶 − 𝜔 , 𝐿𝐶 11
𝑅 + 𝑗 𝜔𝐿 − 𝜔𝐶 10.8
$
quadratic pole at 𝜔' = 10 2
0.6

𝑅𝐶𝜔 0.4

𝐺4 𝑗𝜔 = 0.2

𝑅𝐶𝜔 , + 1 − 𝜔 , 𝐿𝐶 , 0 −2
𝜔( 𝜔
𝜔) 10 𝜔5
0 2
10 10

𝜋 𝜔
𝜑 𝐺4 𝑗𝜔 ≈ − 2 tan#$
2 𝜔' 𝑅 𝑅 , ,
−𝐿 + 𝐿 + 4𝜔 '
𝑅𝐶𝜔 1 𝜔# = 𝑅
𝐵𝑊 → = 2 𝐵𝑊 = 𝜔5 − 𝜔# =
𝑅𝐶𝜔 , + 1 − 𝜔 , 𝐿𝐶 , 2 𝑅 𝑅 , 𝐿
,
𝐿+ 𝐿 + 4𝜔'
𝜔5 =
2
𝐺4 (dB) 𝜑 𝐺4 (rad)
Bode diagram:
𝜋2
0 0

Single zero at the origin 21

20 dB/dec
−10 −20 dB/dec

and quadratic pole at 𝜔% −20 00

−30 𝜋−1
𝜔 − 𝜔
−40 −2
10
𝜔(
10
0 2
10
2−2 −2
𝜔( 0 2
10 10 10
1.5.1. PASSIVE FILTERS (IV) 𝐿

Example: single band-reject filter +

𝑣2 𝐶 𝑅 𝑣3
–
#
quadratic zero at 𝜔( = !"
(𝜉 = 0)
,
𝑅 1 − 𝜔 𝐿𝐶 𝐺4
𝐺4 𝑗𝜔 = =
𝐿 𝐿 11
1 + 𝑗𝜔 − 𝜔 , 𝐿𝐶
𝑅+ 𝐶 𝑅 0.8
1
𝑗𝜔𝐿 + 𝑗𝜔𝐶 quadratic pole at 𝜔( =
# 0.6

!" 0.4
!6$
(𝜉 = $7
) 0.2

0 100 𝜔
−2
𝜔0
10
(
2
10

Bode diagram:
𝐺4 (dB)
100
zero
pole 40 dB/dec 𝜑 𝐺4 (rad)
50 𝜋2
Gv
Quadratic zero at 𝜔% 21
00
and quadratic pole at 𝜔%
00
−40 dB/dec
−50
𝜋
−1
− 𝜔
2 𝜔(
𝜔( 𝜔 −2 −2 0 2
−100 −2 0 2
10 10 10
10 10 10
ASSIGNMENT

In an RLC series circuit obtain 3 different filters taking 𝑣! , 𝑣" and 𝑣# as output voltage.

+ 𝐿 𝑣"
𝑣C –

𝑅 𝑣!

𝐶 𝑣#
1.5.2. ACTIVE FILTERS
Drawbacks of passive filters:
• Gain < 1 (< 0 dB). They do not add energy to the signal.
• Inductors are needed. They are expensive, not accurate, bulky and difficult to assemble
in printed circuits.
It is possible to generate all the types of primary filters just with operational amplifiers,
resistances and capacitors.

The Operational Amplifier

• Very versatile circuit, with many applications: sum, subtraction, differentiation, integration,
I-V conversion…
• It is commercially available as an integrated circuit (resistances, capacitors and
transistors).
saturation
µA741: 10V

−𝑣#
𝑣# 𝑣2

saturation 10V
OPERATIONAL AMPLIFIER 𝑅9
𝑣5

+ 𝑅8 +
𝑅. ↑↑ 10F − 10,' Ω
Equivalent circuit: 𝑣5 𝐴 𝑣5 − 𝑣) 𝑣3 𝑅G ↓↓ 1 − 50Ω
≡ 𝑣)
–

–
𝑣) 𝐴 ↑↑ 10F − 10H Ω

Ideal model: 𝑣5
𝑅. → ∞ 𝐴 → ∞ ⟹ 𝑣E → ∞
𝑅G → 0
+
𝐴 𝑣5 − 𝑣) In order to avoid this: 𝑣I = 𝑣(
–
𝑣)
𝐴→∞
𝑅. → ∞ ⟹ 𝑖I = 𝑖( = 0

Assignment: compare the real and the ideal models for this inverter circuit.
𝑅$

𝑅$ 𝑅# 𝑅3
𝑣) 𝑣E
𝑅#
?
𝑣J
– 𝑅8 +
𝐴 𝑣5 − 𝑣)
𝑣2 – 𝑣3
𝑣2 + 𝑣3 𝑣E 𝑅'
=−
𝑣5 𝑣J 𝑅,
INVERTER AND NON-INVERTER CIRCUITS

Generalizing for sinusoidal steady state:

Inverter circuit
𝑍$
𝑍# 𝑍# – 𝑍$
– ≡ 𝑣# 𝑣
+
𝑍# #
+

Non-inverter circuit

–
+ 𝑍$
≡ 𝑣# 1+ 𝑣
𝑍# #
–
𝑍$
𝑍#
 𝐶#
ACTIVE LOW-PASS FILTER
𝑅$
𝑅#
1st order (lab. session 2): –
𝑣2 + 𝑣3
𝑅:
2nd order: 𝐶$
𝐾%
𝑅# 𝑅$ In general: 𝐻 𝜔 =
𝜔' 𝜔
– − ' + 𝑗2𝜉 𝜔 + 1
𝐶#
𝜔3 3
𝑣2 + 𝑣3
𝑅6
𝐾+ = −
𝑅7
𝑅6 1
− 𝜔9 =
𝑅7
𝐺5 𝑗𝜔 = 𝑅6 𝑅8 𝐶7 𝐶8
1 1 1 8
1 + 𝑗𝜔 𝑅6 𝑅8 𝐶7 𝐶8 𝑅7 𝐶7 𝑅8 𝐶7 𝑅6 𝐶7 − 𝜔 𝑅6 𝑅8 𝐶7 𝐶8
+ +
1 𝐶8 1 1 1
𝜉= + + 𝑅6 𝑅8
2 𝐶7 𝑅7 𝑅8 𝑅6
E.g.: 𝑅$ = 𝑅, = 𝑅. = 5 kΩ
e
X
𝜔6 = 2000 r/s → 𝑓6 = 318𝐻𝑧
𝐾' = −1
X
𝐶$ = 𝐶, = 0.1 µF 764 r/s
𝜉 = 1.5 → 2 single poles with different 𝜔6 → 𝜔3 = |
5236 r/s
Problem: Design with the circuit above a LPF with 𝜔3 = 2000 r/s and 𝜉 = 1, 0.75, 0.5, 0.25
varying 𝐶, and 𝐶' .
ACTIVE HIGH-PASS FILTER
+ 1 𝑅
1st order: 𝑍, = 𝑍' =
– 𝑗𝜔𝐶, 1 + 𝑗𝜔𝑅𝐶'
𝑅
𝑣2 𝐶$ 𝑣3
1 + 𝑗𝜔𝑅 𝐶, + 𝐶'
𝐺/ =
𝐶# 1 + 𝑗𝜔𝑅𝐶'

𝐺4
1
Single zero at: 𝜔3, =
𝑅 𝐶, + 𝐶'
1
Single pole at 𝜔3' =
𝑅𝐶'
𝜔,# 𝜔,$ 𝜔

1st order filters: just one zero and/or one pole.

For improving the selectivity: 2nd order filters.
For improving even more: higher order (chapter 3) or cascade of low order filters.
ACTIVE BAND-PASS FILTER
𝑗𝜔𝐾%
2nd order (lab session 3): In general: 𝐺/ = 𝜔
−𝜔 ' + 𝑗𝜔 𝑄% + 𝜔%'
𝑗𝜔% 𝐾% 𝐾% 𝑄
𝐶# 𝑅$ Highest gain at 𝜔 = 𝜔% : 𝐺/ U
= 𝜔% 𝜔% =
𝑅# =% '
−𝜔% + 𝑗 '
+ 𝜔% 𝜔%
–
𝜔% 𝑄
𝐶$
𝑣2 𝑅: + 𝑣3 Check that 𝐵 = 𝜔I − 𝜔( =
𝑄

1
−𝑗 𝜔
𝑅, 𝐶,
Solving this circuit: 𝐺/ =
𝑅,
1 1 1+ €𝑅
K
−𝜔 ' + 𝑗𝜔 + +
𝑅' 𝐶, 𝑅' 𝐶' 𝑅, 𝑅' 𝐶, 𝐶'
Identifying terms:
𝑅,
𝑄 𝑅' 𝐶, 𝐶' 𝑅' 𝐶, 𝐶' 1+ €𝑅 𝜔% 𝐶, + 𝐶'
K
𝑅, = →𝑄= 𝐵= =
1+ €𝑅 𝜔% 𝐶, + 𝐶' 𝐶, + 𝐶' 𝑅, 𝑅' 𝐶, 𝐶' 𝑄 𝑅' 𝐶, 𝐶'
K
𝜔% =
𝑅, 𝑅' 𝐶, 𝐶'
𝐾% 𝑄 𝑅' 𝐶'
𝐺/ U = =−
=% 𝜔% 𝑅, 𝐶, + 𝐶'
ASSIGNMENT

• Problems 13.b) and 15.