Escolar Documentos
Profissional Documentos
Cultura Documentos
Finzi Neto
12/12/14
SUMRIO
a. Introduo
b. Fundamentos de Filtros Passa Baixa
I. Resposta de Butterworth.
II. Resposta de Tschebyscheff.
III. Resposta de Bessel
IV. Fator de Qualidade Q
c. Projeto de Filtros Passa Baixa
I. Primeira Ordem
II. Segunda Ordem
III. Ordens Superiores
d. Projeto de Filtros Passa Alta
I. Primeira Ordem
II. Segunda Ordem
III. Ordens Superiores
Prof. Roberto M. Finzi Neto
12/12/14
SUMRIO
e. Projeto de Filtros Passa Banda
I. Segunda Ordem
II. Quarta Ordem
f. Projeto de Filtros Rejeita Banda
I. Filtro Active Twin-T
12/12/14
BIBLIOGRAFIA
12/12/14
a. Introduo
u Def.: circuito ou dispositivo que permite a passagem de sinais
12/12/14
nal conditioning stages. System power supplies often use band-rejection filters to supjust add a linear phase
shift to each frequency component, thus contributing to a constant
press the 60-Hz line frequency and high frequency transients.
time delay. These are called all-pass filters.
In addition, there are filters that do not filter any frequencies of a complex input signal, but
At high frequencies
(> 1a MHz),
all ofshift
these
filters
usually
consist thus
of passive
components
just add
linear phase
to each
frequency
component,
contributing
to a constant
These (R),
are called
all-pass filters.
such as inductorstime
(L),delay.
resistors
and capacitors
(C). They are then called LRC filters.
At high frequencies (> 1 MHz), all of these filters usually consist of passive components
In the lower frequency
range (1 Hz to 1 MHz), however, the inductor value becomes very
such as inductors (L), resistors (R), and capacitors (C). They are then called LRC filters.
large and the inductor itself gets quite bulky, making economical production difficult.
In the lower frequency range (1 Hz to 1 MHz), however, the inductor value becomes very
active
filters
becomeitself
important.
filters economical
are circuits
that usedifficult.
an oplarge and
the inductor
gets quite Active
bulky, making
production
In these cases,
a. Introduo
erational amplifier (op amp) as the active device in combination with some resistors and
In these cases, active filters become important. Active filters are circuits that use an opA altascapacitors
frequncias
(>
1MHz)
esses
filtros
constitudos
de
to provide
anamplifier
LRC-like
performance
at low
frequencies
16-1).
erational
(opfilter
amp)
as the active
device
inso
combination
with(Figure
some resistors
and
capacitors
to provide
an LRC-like filter performance
lowindutor
frequencies (Figure
16-1).
dispositivos passivos
(RLC).
A frequncias
menoresat o
se torna
R
VIN
c
T~
R
V
c
_
VIN
V
C2
R1
VIN
II
! R! R
R1
C2
II
11
Vouz
T TVouz
u Os filtro ativos
so compostos de um ou mais amplificadores
Figure 16-1. Second-Order Passive Low-Pass and Second-Order Active Low-Pass
Figure
16-1. Second-Order
Passive Low-Pass
and Second-Order
Low-Pass
operacionais,
resistncias
e capacitores.
PodemActive
gerar
a mesma
12/12/14
16.2
u n ddesign
a m(Buttere nequations
t a l s of toL ocalw - P a s s Filters
er covers
active
filters.transfer
It introduces
the of
three
mainfollowed
filter
optimizations
arts
with the
general
function
a filter,
byFthe
hebyscheff,
and Bessel),
followed
by five Fsections
most
common
16.2
u n dchapter
a mdescribing
e n t closes
a l s theof
L aoThe
w
- Pmost
a s son
Filters
ulate
the individual
circuit
components.
The
with
section
practical
simple
low-pass filter is the passive RC low-pass
resign
applications:
low-pass,
high-pass,
band-pass,
band-rejection,
and
all-pass
filhints for single-supply filter designs.
R
The most
er than resembling just another filter book, the individual
filtersimple
sectionslow-pass
are writ- filter is the passive RC low-pass network show
R
ookbook style, thus avoiding tedious mathematical derivations. Each section
c
the
general
transfer
function
of
a
filter,
followed
by
the
design
equations
to
cale n t a l s of L o w - P a s s Filters
c
individual circuit components. The chapter closes with a section on practical
he
most
simple low-pass
filter is the passive RC low-pass network shown in Figure 16-2.
ts for
single-supply
filter designs.
VIN
VIN
VOUT
VOUT
Figure 16-2. F i r s t - O r d e r P a s s i v e R C L o w - P a s s
b. Fundamentos
Filtros
Figure de
16-2.
F i r s t - O Passa
r d e r P a sIts
s iBaixas
vtransfer
e R C function
L o w - P ais:
ss
VIN
VOUT
of L o w - P a s s Filters
c
u O filtro passa baixa
mais simples composto por uma
1 nica rede RC
VOUT
A(S) =
a--~
(01)
wheres the
+ ~complex1 frequency
+ sRC variable, s =~+(~, allows fo
RC
pure sine waves, the damping constant, (~, becomes zero
1
where the
complex
frequency variable, s =~+(~, allows for any time var
a--~
1
onde
s = j=j2f para
qualquer
sinal
senoidal.
For a normalized presentation of the transfer function, s i
r d e r P a s s i v eA(S)
R C L=o w - P a s s
pure sine waves, frequency,
the damping
(~, becomes
= j(o.
s + ~
1 + sRC
o r - 3constant,
dB frequency,
(oc, andzero
has and
theses relation
u A frequncia
de corte do filtro ocorre quando o sinal sofre uma
RC
s + ~
_2filter's
J~
VIntransfer
corner
Rcorner
Cf = j ~ of theRlow-pass
C
or a normalized presentation
ofthe
function, s is referredWith
tos the
RC
the
frequency
in Figure 16-2 b
s- sdo
~= sRC
Jrcc
funo
transferncia
filtro
se- torna:
equency,
ou
r - 3 Normalizada,
dB frequency, (oc, a
and
has these
relationships:
complex frequency variable, s =~+(~, allows for any time variable signals.
For and the transfer function A(s) results in:
waves, the dampings constant,
and sthe
= j(o.
_ J~ (~, becomes
f = j ~ zero With
corner frequency of the low-pass1 in Figure 16-2 being fc = 1/2
(05)
(06)
s- ~ of the transfer
- Jrccfunction, s issreferred
malized presentation
thethe
filter's
cornerfunction A(s) results in:
= sRC to
and
transfer
o r - 3 dB frequency, (oc, and has these relationships:
magnitude
of the gain response is:
With the corner frequency of the low-pass in Figure 16-2 being fc =The
1/2~RC,
1 s becomes
Prof. Roberto M. Finzi Neto
12/12/14
A(s) = 1 + s
s the
_ transfer
J~
ffunction
= j ~ A(s) results in:
= sRC and
s- ~
- Jrcc
A(s) = 1 + s
IAI =
s- ~
_ J~
f = j~
- Jrcc
The magnitude
gaintorna-se:
response is:
u A magnitude da resposta da funo
de ganho ofdothe
filtro
IAI =
1
(07)
,/1 + Q 2
u Quando >> 1 (f >> fc) a funo de ganho apresenta uma taxa de atenuao
For frequencies O. >> 1, the rolloff is 20 dB/decade. For a stee
de aproximadamente 20dB do dcada.
canvlidas
be connected
as shown alcanar
in Figure 16-3.
u As equaes (01) e (07) s so
se Ioutin=series
0. Podemos
essa To avoid
Fundamentals of Low-Pass Filters
operating as impedance
the do
individual filt
condio ao se incluir um Amplificador
Seguidor converters,
de Tensoseparate
na sada
circuito. Podemos ainda aumentar a ordem do filtro serializando N estgios.
R
262
VIN
c_
VOUT
u Cada estgio
aumenta a taxa de atenuao em mais 20dB/dcada de freq.
Figure 16-3. Fourth-Order Passive RC Low-Pass with Decoupling Amplifiers
The resulting transfer function is:
Prof. Roberto M. Finzi Neto
A(s) =
12/12/14
1
(1 + ~,s)(1 + ~2 s) ... (1 + ~n S)
-10
-20
II
"k
m -30
"o
._c
-40
r
o
4th Order L~
i'~
I[
~1
I -5o
_<
-60
,~
4th Or~
b. Fundamentos de Filtros Passa Baixas
-70
II
-Pass Filters
-10
-20
II
-90
"k
m -30
"o
I
._c
-40
r
o
4th Order L~
i'~
I[
~1
-180
m
1-
I -5o
_<
-e-
-60
-70
-80
0.01
Order
,~
4th Or~
0.1
-270
II
1
Frequency-
II
10
Note:
-360
0.01
100
0,1
1
Frequency
12/12/14
--
10
100
Curve1" 1St-orderpartial low-passfilter, Curve 2: 4th-orderoverall low-pass filter, Curve 3: Ideal 4th-orderlow-pass filter
u ConformeThegrficos
anteriores, o filtro real apresenta problemas tanto
gain and phase response of a low-pass filter can be optimized to satisfy one of the
na amplitude
do
ganho
following
three
criteria: de sada quanto na fase do sinal.
u Podemos otimizar
a sada
filtro para algum dos critrios abaixo:
1) A maximum
passbanddo
flatness,
I. Sada com ganho maximamente plano para f < fc.
2) An immediate passband-to-stopband transition,
II. Transio imediata e aguda entre a banda de passagem e de atenuao.
A linearde
phase
response. na fase do sinal.
III. Resposta3) linear
variao
For that essas
purpose, the
transfer function must
allow for complex
poles and needs
to bepermitir
of
u Para permitir
otimizaes,
a funo
transferncia
deve
the following type:
a existncia
de polos complexos. A funo transferncia se torna:
A(s) =
A0
A0
+ bns2)
(08)
onde A0 o ganho
DC do filtro e ai eandbiaiso
os coeficientes dos filtros em
where Ao is the passband gain at dc,
and bi are the filter coefficients.
cada estgio.
Since the denominator is a product of quadratic terms, the transfer function represents
u As sees aaseries
seguir
descrevem
trsstages,
principais
de filtros
of cascaded
second-orderas
low-pass
with ai and biotimizaes
being positive real coefficients.os
These
coefficients
define the complex
for each second-order filter
que garantem
tipos
de respostas
(I, pole
II elocations
III) citadas.
stage, thus determining the behavior of its transfer function.
12/12/14
The following three types of predetermined filter coefficients are available listed in table
10
Active filters are RC networks that include an active device, such as an operational
fier (op amp).
Section 16.3 shows that the products of the RC values and the corner frequency
yield the predetermined filter coefficients ai and b i, to generate the desired transfe
tion.
The following paragraphs introduce the most commonly used filter optimizations.
Figure 16-5 plots the gain response of different orders of Butterworth low-pass filte
0
-10
_.
-20
- 3 0 ~ - - ~ - 2 n d O r d e r / ~
I I IIIIIII
P~I IIIIIII
I I IIIIIII
I I I~IIIII
I I IL21"II
II
I I IINII
-40
-50
-6O
0.01
0.1
10
100
Frequency-
12/12/14
11
Fundamentals of Low-Pass
9
-.
:-
..
. ,
Tschebyscheff
Filters
b. Fundamentos de 16.2.2
Filtros
PassaLow-Pass
Baixas
II. Resposta de Tschebyscheff The Tschebyscheff low-pass filters provide an even higher gain rolloff above fc. How
as Figure 16-6 shows, the passband gain is not monotone, but contains ripp
u Apresentam alta taxa de atenuao
acima
ordem
inicial
constant magnitude
instead. de
For a fgiven
filter order,
the higher
the passband ripple
c, com
higher
the filter's
roiloff.
igual a 2. Entretanto, o ganho
no
constante
na faixa de frequncia
10
de passagem.
m
-10
d Order
-20
-30
-40
-50
--60
0.01
0.1
10
100
Frequency--
12/12/14
12
01
"0
I -~80
W
(=
a.
I
~
-270 ~
-360
0.01
0.1
1
Frequency-
10
100
1.2
-90
L
01
"0
I -~80
0.8
0.6
W
(=
a.
I
~
0.4
-270 ~
~
-360
0.01
0.2
0
0.1
1
Frequency-
10
0.01
100
12/12/14
0.1
1
Frequency --
10
100
13
8, -./"%0%%
8, -./"%0%%
B&%6:
B&%6:
&!6E1)
&!6E1)
&)
&)
&#
&#
G#%6
Resposta de Bessel G#%6
/
%#
Baixa taxa de atenuao
acima de c e no possui resposta plana
%#
abaixo de fc.
4 %
6
!
4 %
6
!
!
A resposta em atraso
constante
o torna a melhor escolha para filtrar
sinais de comunicao
digitais
(onda
quadrada).
8I
85?6
8,-!:
8I
85?6
8,-!:
"
"
14
12/12/14
%"%#$$!9"%$-(
%"%#$$!9"%$-(
bi
Q=
ai
(09)
Q=
f2 f1
(10)
12/12/14
15
Figure 16-10.
12/12/14
30
20
m
-o
~l~l
10
c
m
(:1
i
<C
,.,..
-10
I III
2ndStage-
........
~
I
I~! IIII1~ . X ~
I~XI
,3,rI Stage- ~ [
-20
4th Stage
5th Stage
-30
0.01
!1111
0.1
Frequency-
10
16
16.3 Low-Pass
FilterTheDesign
In addition, the ratio v ~'~ _- Q is defined
as the pole quality.
higher the Q value, the
ai
more a filter inclines to instability.
Equation 16-1 represents a cascade of second-order low-pass filters
tion of a single stage is:
Ai(s ) =
A~
(1 + ais + bis2 )
Equation 16-1 represents a cascade of second-order low-pass filters. The transfer funcu Atravs da serializao For
de a filtros
de 2, Eq. (11), e 1, Eq. (12),
first-order filter, the coefficient b is always zero (bl=0), thus yie
tion of a single stage is:
A~
(1 + ais + bis2 )
A~
A(s) = 1 + a l s (16-2) (12)
(11)
1st order o
1st order
a=l
u Na maioria dos casos, A0 =
1 para
The
andyielding:
second-order filter stages are the building blocks
For a first-order filter, the coefficient b is always
zerofirst-order
(bl=0), thus
que Av(OL) do AO no tenha de ser
ters.muito 2nd order~ J 2nd order
al , bl
alto.
A~
(16-3)
u A(s)
C a d= a1 +easl st g i o p r e c iOften
s a the
s e rfilters operate at unity gain (Ao=l) to lessen the stringent d
1st order
2nd order
o
;
a2, b2
dimensionado de maneira queamp's
o Q open-loop
dos 3rd ordergain.
al
estgios
anteriores filter
sejam
menores
do blocks for higher-order filThe first-order
and second-order
stages
are the building
2nd order
2nd orderstages up to the sixth order.
Jof filter
Figure 16-114thshows
order o the cascading
que
o
Q
do
estgio
atual.
al
a2 ' b 2
ters.
order number consists of second-order stages only, while filters with an
uOs valores de Q de diversos filtros so
1st order stage2nd
2nd order I
include
an stringent
additional
first-order
atorder
the beginning.
Often theapresentados
filters operate at em
unity tabela
gain (Ao=l)
lessen
the
on the opa2,
5th order odemands
;
no tofinal
deste
b2
al
a3' b3
I
amp's open-loop
gain.
mdulo.
,
6th order o
bI
2nd order
al , bl
2nd order
2nd order
a3, b3
Figure 16-11 shows the cascading of filter stages up to the sixth order. A filter with an even
17
Prof. Roberto
M. Finzi
Neto
order number
consists
of second-order
stages only, while12/12/14
filters with an odd order number
include an additional first-order stage at the beginning.
Figure 16-11. Cascading Filter Stages for Higher-Order Filters
a2, b2
5th order o
1 + R~ a2, b2
VIN
Design
A(s)Low-Pass
= 2nd
1 order
+Filter
cocR1ClS
The transfer functions of the circuits are:
2nd order
6th order o
R1
~,~~.,~.~
....... - ~
.,
al
..
:~ ,
Vou
a3' b3
2nd order
:~.,..~
T
al , bl
a2, b2
a3, b3
R2
R2
R1
R The
~ transfer functions of Figure
16-10 are:
demonstrated
the circuits
the
twot
er(PB)
its coefficient
Q. Therefore, to comparison
avoid the saturation between
of the individual
stages,
c. Projeto
de Filtros Passa BaixasThe
in the order ofgenerates
rising Q values. aThe
Q values
for each
filter orde
The negative sign indicates that the inverting
180
~ phase
shift
fro
R2 placedamplifier
R2
I. Primeira Ordem
Tables 1 6 - 4 through
1 + R~ order) in Section 16-9,and
A(s) 16-10.
=
R~
the filter input to the output.
R2
= 1 + filtro
cocR1ClSusando as configuraes 1 de
-!- cocR2ClS
u Podemos implementarA(s)esse
A 0 = 1 + R~
16.3.1
F i r s t - O r d e r L oPorm,
w - P a s s Filter
amplificadores
inversores between
e no
inversores.
abordaremos
aqui
The coefficient comparison
the
two
transfer
functions
and
Equation
16-3 yield
The
negative sign indicates that the inverting amplifier generates a 180 ~ phase shift fro
16-13. First-Order
Inverting
Low-Pass
Filter
apenas a topologia no
inversora.
Figures 16-12 and 16-13 show a first-order low-pass filter in the inve
the filter input to the output.
configuration.
u A funo transferncia
apresentadainverting
na
Eq.
(13) 1e Ao
os =valores
R2
and
R2 dos
a
1
=
cocR1C
The transfer functions
of
the
circuits
are:
The
coefficient
comparison
between
the
two
transfer
functions
and
Equation 16-3 yield
A 0 = 1 +so
R~definidos nas Eq. (14) e (15).
R~
a1
coeficientes
VIN
Vou
R2
1 + R~
(13)
A(s) = 1 a+ 1cocR1ClS
= cocR1C 1
VIN-~~/~/~
R2
and
R2
A 0 =and
1 + R~
VOUT
Ao =
R2
(14)
ToA(s)
dimension
the
circuit,
specify
the
R~ corner fr
=
R~
a 1 = cocR1C 1
(15)
and then
C1, and
solve for
1 -!- cocR2ClS
a 1 resistors
= e)cR2CR1
1 and R2:
and
Ra
a 1 = e)cR2C 1
The negative
sign indicatesthe
thatcircuit,
the inverting
amplifier
generates
a 180 ~ phase
al(fc), shift
To dimension
specify
the corner
frequency
the from
dc gain (Ao), and capacit
To dimension
the circuit, specify
corner
frequency
the dc gain (Ao),
=ganho
2~fcC1
projeto
do filtro feito
definindo-se
fcR1
, othe
A0 e o (fc),
capacitor
C1. and capacit
theu
filter O
input
to the output.
_
The coefficient comparison between the two transfer functions and Equation 16-3 yields:
al
R1
R2 = 2~fcC1
A 0 = 1 + R~
al
R1 = 2~fcC1
(16)
272
and
12/12/14
R 2 = R3(A 0 - 1)
and
=
al
and
R2 = R2 al
2~fcC1
(17)
R 2= = R3(A
0 - 1)
2~fcC1
Ao
R2
R~
18
and
R1 =
a2
a2
Ao
R2 =theRfirst
1)
0
However, to design
stage
of a=
third-order
unity-gain Bessel low-pass filter, assum3 (1
ing the same values for fc and C1, requires a different value for R 1. In this case, obtain
al for a third-order Bessel filter from Table 16-4 in Section 16.9 (Bessel coefficients) to
calculate RI:Com R2 = 0, R3 pode ser removido.
A simulao
al
_ usa um
0.756LF347
12/12/14
When operating at unity gain, the noninverting amplifier reduces to a voltage follower (Figure 16-14), thus inherently providing a superior gain accuracy. In the case of the inverting
19
0 2
VIN
VOUT
C1
T"
~ . . .
,~ .
0 2
R4
A0 = 1+
R3
(18)
VIN
VOUT
C1
T"
Ao
(19)
0 2
M. Finzi
Neto
ForProf.
the Roberto
unity-gain
circuit
in Figure 16-16 (Ao=I), the 12/12/14
transfer function simplifies to:
20
VOUT
VOUT
~ . . .
Low-P
0 2
Ao
VIN
VOUT
C1
T"
Ao
Ordem
A(s)
=II. Segunda
r the
unity-gain
circuit in Figure
16-16 (Ao=I), the transfer function simplifies to:
A(s)
de =Q < 3, no incluem
R316-15.
e R4 General
o AO Sallen-Key
muda sua
topologia
Low-Pass
Filter para um
Figure
1
+
~cCl(R
1
+
R2)s
+
e)c
2
R1R2C1C2s2
or the unity-gainseguidor
circuit in Figure
(Ao=I),
to:
de 16-16
tenso
(A0the
= transfer
1). A function
funosimplifies
transferncia
se torna:
0 2
A(s)
=
e coefficient
comparison
between this transfer function and Equation(20)
16-2 yields:
1 + ~cCl(R 1 + R2)s + e)c2 R1R2C1C2s2
VOUT
Ao=l
Low-Pass
Filter
Design
he coefficientacomparison
between
this
transfer
function
and
Equation
16-2
yields:
1 = mcCl(a 1 + a2)
Abo1= l= o)c2R1R2C1C2
seguida, Calculamos Figure
R1 e 16-16.
R2 usando
Eq.Sallen-Key
(21). ALow-Pass
Eq. (22)
Unity-Gain
Filter deve ser
In order to obtain real values under the square roo
a 1 = mcCl(a 1 + a2)
observada para se obter apenas valorestion"
reais na Eq. (21).
ven C1 and bC2,
the
resistor
values
for
R1
and
R2
are
calculated
through:
The
transfer
function
of the circuit in Figure 16-15 is:
1 = o)c2R1R2C1C2
A(s) =
a,C 2values
::F /a,2C2
2 - R24blC,C
2
iven C1 and C2, the resistor
for R1 and
are calculated
through:
(21)
a1,2
4=fcC 1C2
2 ::F Neto
/a,2C2
Prof. Robertoa,C
M. Finzi
a1,2
2 - 4blC,C 2
4=fcC 1C2
4b 1 Ao
02>_ 0 1 ~ 2
al
(22)
For
the unity-gain circuit in Figure 16-16 (Ao=I), the 21
transfer function si
12/12/14
Example 16-2.A(s)Second-Order
Unity-Gain Tschebyscheff
=
In order to obtain real values under the square root, C 2 must satisfy the following condition"
4b 1
02>_ 0 1 ~ 2
al
c. Projeto
de FiltrosUnity-Gain
Passa Baixas
(PB)Low-Pass Filter
Example
16-2. Second-Order
Tschebyscheff
II. Segunda Ordem
The task is to design a second-order unity-gain Tschebyscheff low-pass filter with a corner
Exemplo: projete um filtro PB de 2 ordem com fc = 1kHz e otimizao de
frequency of fc = 3 kHz and a 3-dB passband ripple.
Tschebyscheff com ripple de 3dB.
Soluo:From
da tabela,
obtemos
os coeficientes
a1 =1,0650
e b1 = obtain
1,935.
Table 16-9
(the Tschebyscheff
coefficients
for 3-dB ripple),
the coefficients
al and
bl for aCsecond-order
filter with al = 1.0650 and bl = 1.9305.
Podemos
arbitrar
1 em um valor comercial da ordem de dezenas de nF, de
acordo com fc. Ento, faamos C1 = 22nF.
Specifying C 1 as 22 nF yields in a C 2 of:
Aplicando C1 na Eq. (22), obtemos:
02 _ c14b1~ = 2 2 . 1 0 - 9 n F . 4 . 1 . 9 3 0 5 _-.. 150,1nF 180nF
al 2
" i ~ 0 - ~ - 150 nF
2
Inserting al and
bl *180
into*10
the9 resistor
equation
2 results
in"9 *180 *10 9
1,065
*10 9 ) for
4R1,
*1,9305
* 22 *10
(1,065 *180
R1 =
1632
1 . 0 6 5 . 1 5 0 . 1 0 - 9 - , / ( 1 . 0 6 5 . 195 0 . 1 0 - 9 ) 2 - 4 9. 1 2. 9 3 0 5 . 2 2 . 1 0 - 9 . 195 0 . 1 0 -9
9
R2 =
R1 =
and
~ 4 ~ . 3 . 1 0 3 . 2 24. 1* 0 _*1000
9 . 1 5*022. 1*10
0 _9 *180
9 *10 9
12/12/14
=5431
1.26 kQ
22
R1 =
~4~.3.103.22.10_9.150.10_ 9
and
1.065.150.10 -9 + ~ ( 1 . 0 6 5 . 1 5 0 . 1 0 - 9 ) 2 - 4 . 1 . 9 3 0 5 . 2 2 . 1 0 - 9 . 1 5 0 . 1 0 R2 =
4g.3.103.22.10_9.150.10_
150n
180nF
1632
5431
VouT
Soluo (cont...)
Os valores de R1 E R2
devem ser obtidos por
meio de potencimetros
A s r e s p o s t a s d e
amplitude e fase do
filtro so apresentadas
em comparao ao filtro
de 1 ordem.
Observa-se que a taxa
de atenuao bem
maior, alm do ripple de
um pico (2 ordem) na
amplitude de sada.
Prof. Roberto M. Finzi Neto
Figure 16-17.
12/12/14
23
A(s)c.=
Projeto de Filtros
Passa Baixas (PB)
A~
a4
A o = 1 4-R--~ a 4
with
= Ordem
II. Segunda
1 A(s)
+ e)cRC(3
- Ao)s + A~
(eL)cRC)2s 2
with
A o = 1 4-R--~
1 + e)cRC(3 - Ao)s + (eL)cRC)2s 2
Low-Pass
Filter Design
u A topologia Sallen-Key apresenta um caso especial
onde
Design
possvel
Low-Pass
Filter
9
R1 =
R2 = R e C
C2 = C. 16-2 yields:
The definir
coefficient
comparison
with
1 = Equation
The coefficient comparison with Equation 16-2 yields:
u Nesta simplificao, a Eq. (19) assume a forma descrita em Eq. (23),
TheThe
general atransfer
function
changes
to" to"
changes
1 transfer
=definido
(JL)oRC(3
- Eq.
Ao) (24).
a 1function
= na
(JL)oRC(3
- Ao)
com general
ganho
9
A(s)A(s)
= =
A~ A~
b
1
=
(~cRC)
b
= +2(~cRC)
2 2 2
+ e)cRC(3
- 1 Ao)s
+ (eL)cRC)2s
1 +1 e)cRC(3
- Ao)s
(eL)cRC)2s
withwith
(23)
a 4a 4
1 4-R--~ (24)
A oA=o =1 4-R--~
Given
andfor
solving
and Ao results
in"
Given
C and
solving
R and
AoR results
in" arbitrria
u AThe
partir
daC Eq.
(23)
efor
da
definio
de C, obtemos o valor de
coefficient
comparison
with
Equation 16-2
yields:
The coefficient comparison with Equation 16-2 yields:
R e A0.
]~
a 1 = (JL)oRC(3 - Ao)
]~
al =al 3
Ao=3Ao=3-
- Ao)
R=
(25)
Ra=1 = (JL)oRC(3
2~fcC
and
2~fcC
and
b 1 = (~cRC) 2
N N
= 31
Q Q
(26)
b 1 = (~cRC) 2
Thus,
Aosolving
depends
solely
on
pole in"
quality
Q and
viceunicamente
versa; Q, and with
it the filter
type,
u Eq.
mostra
aand
resposta
filtro
ser
pelo
Given
C and
for
Aothe
results
Thus,
Ao(26)
depends
solelyque
on Rthe
pole
qualitydo
Q and
vice
versa;
Q, and with itdefinida
the filter type,
Given
C is
and
forby
R the
andgain
Ao results
in"
setting
of Ao:de qualidade do filtro.
ganho
Adetermined
,solving
o qual
determina
o fator
Ao=31 and
Q ==
1
al N= 3
1 Q
24
Prof. Roberto M. FinziRNeto
12/12/14
3 - Ao
2~fcC
and
3 solely
- A oon the pole quality Q andNvice versa; Q
Thus, Ao depends
Q, and with it the filter type,
R=
Q]~
=
2~fcC
]~
Ao=3-
al
= 3
A(s) =
A~
with
Ao =
Q, and
3 - Ao
The circuit
allowstabelas
the filter type
to be changed
Soluo: arbitraremos o valor C = 22nF
e Rin3 Figure
= 1k16-18
. Das
temos
os throug
ratios R4/R3.
seguintes parmetros:
c
Ripple
bi
0,5dB
1,3827
0,86
1,0dB
1,5515
0,96
2,0dB
1,7775
1,13
1t
VN
12/12/14
Figure 16-18.
Vout
25
1, 3827
8507
2 1kHz 22nF
A0 0,5dB = 3
1
= 1,837 R4 = R 3 (A0 1)
0,86
R4 1,0dB = 1k (1,837 1) = 837
R1dB =
1,5515
9011
2 1kHz 22nF
A0 1,0dB = 3
1
1,958 R4
0,96
R2dB =
1, 7775
9638
2 1kHz 22nF
A0 1,0dB = 3
1
2,115 R4
1,13
1,0dB
1,0dB
= 1k (1,958 1) = 958
= 1k (2,115 1) = 1,115
12/12/14
26
12/12/14
27
12/12/14
28
The task is to design a fifth-order unity-gain Butterworth low-pass filter with the corner freHigher-order low-pass filters are required to sharpen a desired filter
quency fc = 50 kHz.
Filter
Coe
that purpose, first-order and second-order filter stages are
connecte
product from
of theTable
individual
frequency
First the coefficients for a fifth-order Butterworth filter are the
obtained
16-5,
Sec- responses results in the optim
sponse of the overall filter.
tion 16.9:
Table 16-5.
Butterworth Coefficients
In order to simplify the design
ai
Filter nI
al = 1
bi
a i
bl = 0
Filter 2
= 1.61 80 Passa
b2 =Baixas
116-3. Fifth-Order
Example
c. Projeto
dea2Filtros
(PB) Filter
13
1 = 0.6180 1.0000 b3 = 1
0.0000
1.000
a3
The task is to design a fifth-order unity-gain Butterworth low-pass filter
III.OFilter
rdens
Superiores
quencyvalues
fc = 50 and
kHz. calculating the
Then dimension each
partial 1filter by specifying
capacitor
1.4142
1.000 f = 50kHz
0.71com
Filter Coef
Exemplo:2 projete
e simule
umthe
filtro
PB1.0000
de quinta
ordem,
c
required resistor values.
First the coefficients for a fifth-order Butterworth filter are obtained fro
resposta3de Butterworth.
1
1.0000
tion 0.0000
16.9:
Soluo.
2
1.0000
1.0000
Table 16-5. Butterworth Coefficients
1. Obtenha
os coeficientes atravs das tabelas
4
a 11
i2
VIN~ C1TiT
1
12
1
1.8478
a i
0.7654
1.0000
bi
1.0000
1.0000
1.0000
1.6180
0.6180
1.4142
VouT
1.000
1.272
ai
Filter I
Filter 2
Filter 3
0.719 al = 1
k1.390
i = a2 = 1.61 80
fci/fc
a3 = 0.6180
1.8478
1.0000
al 1 1
1.0000
= 0.0000
3.18 k~
R 1 = 2 ~ c C 12 2 2 ~ . 5 0 . 10.7654
01.801
3 H z9. 1 . 1 0 - 9 F 1.0000
1.0000
4
1 4k.Q.
3.16
2
b3 = 1
0.0000
1.000
0.0000
1.000
Then
dimension each partial
filter by specifying0.62
the capacitor values
1.0000
0.859
required
resistor values. 1.448
1.0000
1.62
1.0000
1.000
0.71
First-OrderUnity-GainLow-Pass
2. Dimensionar cada
filtro1.931
parcial
usando1.0000
os coeficientes0.676
obtidos
First
1
9 Filter
3
12
0.0000
1.000
2.1 Primeiro
estgio
11.0000
Ordem
1.4142
1.0000
1.000
a1
23
1.0000
1.0000
1.272
With C1 = lnF,
0.5176
1.0000
1.479
Defina C1 = 1nF e calcule R1.
0.
1.00
bi
0.54
bl = 0
Q1.31
i b2 = 1
VIN~
TiT
0.719 C1
1.000
1.390
0.745
1.2470
1.0000
1.117
12/12/14
1.0000
0.0000
1.000
Figure16-20. 1.0000
First-OrderUnity-GainLow-Pass
0.4450
1.499
1.6180
1.0000
0.859
0.52
0.71
1.00
1.93
;
0.54
1.31
0.55
0.80
29
2.25
0.62
VouT
Second Filter
Filter
Second
02
02
Second Filter
IIII
VIM
VIM
02
01
01
II
"" TT
VOUT
VOUT
VIM
c. Projeto
With
III.Ordens Superiores
With
Figure 16-21.
16-21. Second-Order
Second-Order Unity-Gain
Sallen-Key
Figure
Unity-Gain(PB)
Sallen-Key Low-Pass
Low-Pass Filter
Filter
de
Filtros
Passa Baixas
C1 =
C1 =
01
820 pF,
820 pF,
" T
VOUT
4b 2
Soluo (cont...)
C 2 >_ C 1 4b
- 2 = 8 2 0 . 1 0 - 1 2 F . ~ 4.1
= 1.26 nF
4.1
C 2 >_ C 1 -a22 = 8 2 0 . 1 0 - 1 2 F . ~1.6182
= 1.26 nF
Figure
16-21.
Second-Order
Unity-Gain Sallen-Key
Low-Pass Filter
1.6182
a22
2.2
Segundo
Estgio
2
Ordem
ss FilterDesign
The j
closest
5% value
nE
Como o segundo estgio
recebe
umis 1.5sinal
pr-filtrado, podemos definir um
The closest 5% value is 1.5 nE
With C 1 = 820 pF,
d Filter
capacitor menor, ou seja
820pF.
C2, the
temos:
WithCC1 1=
= 820
pF and Calculando
C 2 = 1.5 nF, calculate
values for R1 and R2 through"
With
02
01
C2 =
//
v//
The
5% value
nE
and closest
obtain
Vamos definir C em 1.5nF
e calcular
R1 eisR1.5
2 deste estgio.
and obtain
Second-Order Unity-Gain Sallen-Key
Low-Pass
Filter
2
16-21.
With
C 1 =
820 pF,
4b 2
C 2 >_ C 1 =
a22
820 pF and
4b 2
2
2
2 - 4 bnF
C
2
>_
C
1
= 8 2 0 . 1 0 - 1 2aF2.C~2 4.1
2
H-- a22=C 21.26
2010 2
a2C 2 a22 0 2 - 4 b 2 0 1 0 2 a 2 C1.6182
2 H-- a22 C 2 - 4 b 2 0 1 0 2
a2C 2 - VOUT
a22 0 2 - 4 b 2 0 1 0 2 a22
R1 =
4~fcC1C2
and
R1 =
4~:fcC1C 2
" T
R1 =
4~fcC1C2
and
R1 =
4~:fcC1C 2
II
VIM
C1=
C 1 =
v//
820 pF and
C2 =
R1 =
v//
a22 0 2 - 4 b 2 0 1 0 2
1.618.1.5.10
-9 + V/(1.618.1.5.10-9) 2
4~fcC1C2
1.618.1.5.10
-9 + V/(1.618.1.5.10-9) 2
2
a22
0 2Roberto
- 4 b 2 M.
0 1 0Finzi
2
Prof.
4~fcC1C2
a2 a2 -
Neto
and
//
4~.50.103.820 10
9 4~.50.103.820
10
9 2
a 2 C 2 H-- a22 C 2 - 4 b 2 0 1 0 2
12/12/14
and
obtain
R1
R1=and
R2 are available 1% resistors.
2 1% resistors.
R1 and R2 are 4~:fcC1C
available
a 2 C 2 H--
//
R2 thro
2
a22 C 2 -
_ 4and
. 1 . 8 2 0 . 1 0 - 1 2 . 1 . 5=
.10-9
_ 4 . 1 . 8 2 0 . 1 0 - 1 2R1
.1.5.10-9
4~:fcC1C 2
= 4.42 kQ
= 4.42 kQ
12.1.5.10
-9
12.1.5.10
-9
30
1.618.1.5.10-9a 1 --
V//(1.618.1.5.10-9) 2 - 4 . 1 . 8 2 0 . 1 0 - 1 2 . 1 . 5 . 1 0 - 9
= 1.87
4~.50.103.820 . 1 0 - 1 2 . 1 . 5 . 1 0 - 9
1.618.1.5.10 -9 + V/(1.618.1.5.10-9) 2 _ 4 . 1 . 8 2 0 . 1 0 - 1 2 . 1 . 5 . 1 0 - 9
a2 -
4~.50.103.820 10
9
= 4.42
-9
- 12.1.5.10
4.1
= 3.46 nF
0.6182
2 >_
C 1 - -being
=
3 3 0 . k..O.
10-12F.~
9 R1 = 1.45 k.Q, with the closest C1%
value
1.47
a32
3.16k
1.5n
280
-820p
4.7n
LI
II
-]
330p
12/12/14
" I
VOUT
31
12/12/14
32
820p
-]
330p
VOUT
" I
lter Design
.
Ao
trocar
as posio de resistores e capacitores.
I
r
o . , ,
02
-10
.....
<c
-20
c~
PB
-T Voo~ >v~~tt
qt
PA
R1
>
VOUT
- I
.
10
-30
0.1
1 apresenta a mesma
10
uFigure
O ganho
filtro PA
equao
Ao
16-23.doLow-Pass
to High-Pass
Transition Through
Components Exchange
Frequency-de fc para o PB.
To plot
the gain response
a high-pass
filter,
the gain
of a low-pass filter
uO The
grfico
resposta
filtroof PA
espelha
a mirror
resposta
doresponse
-24. Developing
Gain de
Response
of a do
High-Pass
Filter
I
filtro PB. at the corner frequency, .Q=I, thus replacing .0. with 1/.Q and S with-101/S in Equation 16-1.
I
The generalu
transfer
function
of a high-pass filter is then:
<c
Funo
transferncia:
m
"o
r
o . , ,
A(s) =
]-[(
A=
ai
i 1+~+#
b,)
(27)
(16-4)de
A o ganho na faixa
passagem
12/12/14
-20
-30
0.1
1
Frequency--
33
10
-30
Figure 16-25.
--
0.1
10
R3
R2
Frequency-First-Order Noninverting
High-Pass Filter
Figure 16-25.
C1
R1filter is then:
e general transfer function of a high-pass
VIN
!1
A(s) =
d.
A=
]-[(
ai
Projeto
i 1+~+#
b,)
de
VOUT
(16-4)
Ra
C1
R1
I. Primeira Ordem
VIN
!1
Figure
16-26.
First-Order
Inverting High-Pass Filter
h Aoo being
the
passband
gain.
u Se fizermos bi = 0 na Eq. (27), obtemos a funo transferncia
de um
VO
PA
1
ordem.
ce Equation The
16-4filtro
represents
ade
cascade
ofcircuits
second-order
high-pass filters, the transfer
transfer
functions
of the
are:
ction of a single
isu stage
O circuito
no inversor que implementa o PA
apresentado abaixo em
conjunto
transferncia.
Verses inversoras podem ser
1 +Aoo
~ com sua funo
and
Ai(s
A
( s ) ) ==
= - (16-5)
FigureA(s)
16-26.
First-Order
Inverting High-Pass Filter
1
1
(
a
consultadas
no
livro
referncia.
1
+
~
+
~
oJcR1C 1 s
1 6 . 4 . 1 F(ocR1C
i r s t - O r1d esr H i g h - P a s s F i l t e r
u O ganho
na faixa de passagem definido
pela Eq.
(30) of
e os
de
Figure 16-25 and 16-26 show a first-order high-pass filter inThe
the noninverting
and
the intransfer
functions
the valores
circuits are:
negative sign
indicates
that the inverting amplifier generates a 180 ~ phase shift from
verting
configuration.
h b=0 for all The
first-order
filters,
the
transfer
function
of aeFilter
first-order
filter simplifies
to:
R1
e
C1
pelas
Eqs.
(31)
(32).
High-Pass
Filter Design
High-Pass
Design
.
.
.
.
.
. . . .
the filter input to the output.
a2
a 2 FilterDesign
High-Pass
..
a2
01
A(s)
A~ comparison
The -coefficient
(28)
(16-6)
1 16-6
+ ~ proand
VOUT A (and
between the two transfer functions
(29)
s ) Equation
=
R1
dimension
the corner
(fc),and
the
To dimension theTocircuit,
specifythe
thecircuit,
cornerspecify
frequency
(fc), thefrequency
dc gain (Aoo),
vides two different passband gain factors:
(C1),for
andR1then
solve
(C1), and then solve
and(ocR1C
R2: 1fors R1 and R2:
-R2
R2
and R3
A =
R2
_
1
(30)
_R~
1
A==I+R
3
=
The negative sign indicates that the inverting amplifi
VIN
R1 -
2~fcalC1 R 1 -
2 (31)
~fcalC1
16-25.
First-Order
Noninverting
Filter circuits:
whileFigure
the term
for the
coefficient
al is the High-Pass
same
(32)
R 2for= both
R3(Aoo - R 21) = R3(Aoo - 1) and
Prof. Roberto M. Finzi Neto
a 1 --
o)cR1C 1
. 21
C11 6 . 4R
Ra
1 A oo R 2 =
12/12/14
S e c o n d 1- O6 .r 4d.e2r
and
R 2 = -R
Second-Order
High-Pass
Filter
I. Primeira Ordem
u Exerccio: projete um Filtro PA com fc = 4kHz e ganho unitrio.
12/12/14
35
have changed.
To dimension the circuit, specify the corner frequency (fc), the dc gain (Aoo),
(C1),inand
then16-27
solve allows
for R1 for
andseparate
R2:
The general Sallen-Key topology
Figure
gain setting via
Ao = 1+R4/R3.
_
a2
R1 -
'k/k/k,
C1 [ C2
IIIR 2 -= IR3(Aoo
II
- 1)
2~fcalC1
R,I- _
and
R2 =
-R
1 A oo
o~ =
cz
A(s) =
!m
VOUT
1 +
a2
R3
'k/k/k,
C1
[
C2
The unity-gain topology in Figure 16-28 is usually applied in Iow-Q filters with high gain
V'N
III - I II
aaccuracy.
4
(34)
R3
c lc
v,.
II
II
VOUT
al
R,I- _
a2
~/k/k,
--
!m
a4
" T
12/12/14
m
36
VOUT
H
~
, ,
H
High-Pass
To simplify the circuit
design,Filter
it is Design
common to choose unity-gain (c~ =
The transfer function of the circuit in Figure 16-28 then simplifies t
High-Pass Filter Design
To simplify the circuit design, it is common to choose unity-gain (c~ =
FilterFilter
Design
Design
To simplify the circuit design, it is common to choose unity-gain
(c~ ==1), andHigh-Pass
C1 High-Pass
= 02 = C.
A(s)
The transfer function
of the circuit
in
Figure
2
1+
116-28
.I__then simplifies t
~
, ,
, ,
, ,
,~
,~
~~
~ ~
, ,
, ,
..
~~
..
~~
. .
, ,
~ ~
~~
~,
~,
,~
~ ,
. ~
~ ~
,.
. .
~~
.~
~~
~~
..
,.
,.
, .. .
..
..
The transfer function of the circuit in Figure 16-28 then simplifies to:
cocalC s
eoc2alR2C2 s 2
To simplify the circuit design, it is common to choose unity-gain (c~ = 1), and C1 = 02 = C.
d. Projeto de Filtros Passa Alta (PA)
A(s) =16-28 then2 simplifies
The transfer function of the circuit in Figure
to:1
1+
.I__
A(s)
=
II.
Segunda
Ordem
2
1
+
1
.I__
simplifythe
thecircuit
circuit design,
design, itit is
totochoose
(c~ =cocalC
1),= and
C1
=transfer
02 == 02
C.function
coefficient
comparison
between
this
s and
eoc2alR2C2
s 2 = C. and Equ
o To
simplify
is common
commonThe
chooseunity-gain
unity-gain
(c~
1),
C1
cocalC s
eoc2alR2C2 s 2
u
Podemos
simplificar
projeto
dothen
filtro
definindo
C1 = C2 = C e ganho
The
transfer
function
of
the
circuit
Figure
16-28
simplifies
to: to:
A(s)inino
=Figure
he transfer function of the circuit
16-28
2
1 + then
.I__
A o o =1simplifies
l
eoc2alR2C2 s 2
1
Aoo=l
ba 11 =-u Os
A, a1 this
e b1transfer
so obtidos
(35).
The coefficient
between
functiondiretamente
and
Equation
16-5
yields:
a 1 comparison
--coeficientes
cocR1C
O)c2R1R2
02 de
cocR1C
he coefficient comparison between this transfer function and 1Equation 16-5 yields:
1
b1 =
a 1 -bA
1 o
=o=l
cocR1C
Given
C,
the
resistor
values for
O)c2R1R2
02R 1 and R2 are calculated through:
O)c2R1R2 02
Aoo=l
u E os valores de C ser 1arbitrrio. J os valores de R e R so
b1 =
O)c2R1R2
02C, the
Given
resistor
values for R 1 and R2 are calculated through:
Given C, the calculados
resistor values pelas
for R 1 and
are calculated
through:
Eq.R2(36)
e (37).
R1
=fcCal
a 1 -cocR1C
=
1
al
1
a 1 =-- Given
C,
the
resistor
values
for
R
1
and
R2
are
calculated
(36)
1
cocR1C
R12 = 4=fcCb
=fcCal 1 through: (37)
R
R1b 1 ==fcCal
O)c2R1R2
02 =
1
al
al 1
R1
=fcCal
b
1
=
Prof. Roberto
Finzi Neto1
12/12/14 R 2 = 4=fcCb 1
R 2 =M.4=fcCb
Multiple Feedback
Topology
O)c2R1R216.4.2.2
02
al are calculated through:
Given C, the resistor values for R 1 and R2
37
b 1 = e)cR1CC2
Given capacitors C and C2, and solving for resistors R 1 and R2:
_ 1 -2A=
R1 - 2=fc.C.al
al
R2 = 2~;fc.blC2(1 - 2Aoo)
The passband gain (Aoo) of a MFB high-pass filter can vary significantl
tolerances of the two capacitors C and C2. To keep the gain variation a
necessary to use capacitors with tight tolerance values.
al = Band-Pass
0.756
bl =Design
0
Filter
.
Filter 2
a2 = 0.9996
b2 = 0.4772
Para o primeiro estgio (N = 1), definimos
and compute each partial filter by specifying the capacitor values and c
C = 100nF e calculamos R1 =
1
= 2.105 kQ
Prof. Roberto M. Finzi Neto
2=fcalC 1
quired
resistor values.
2~.103Hz.0.756.100.109F
12/12/14
First
Closest 1% value is 2.1
k~.Filter
38
R1 =
2~.103Hz.0.756.100.10- 9F
= 2.105 kQ
1
= 2.105 kQ
2=fcalC
1 C2~.103Hz.0.756.100.109F
With
= 100nF,
d.
Filtros
_
1
= 3.18 kQ
R 1 - ~fcCal
~.103.100.10-9.0.756
Passa
Alta (PA)
_
1
R
1
~fcCal
Closest 1% value is 3.16 k~.
_
4=fcCb 1
~.103.100.10-9.0.756
4=-103.100.10
= 3.18 kQ
- 9.0.4772
R2 _
Closest 1% value
is 3.161 k~.
0.9996
= 1.67 kQ
4=fcCb 1Figure
4=-103.100.10
- 9.0.4772
16-30 shows
the final filter circuit.
R2 _
1
0.9996
= 1.67
kQ
O Circuito
resultante
apresentado
abaixo,
usando
valores de componentes
Closest 1% value is 1.65
k
~
4=fcCb 1
4=-103.100.10 - 9 . 0 . 4 7 7 2
comerciais:
1.65k
_
100n
Figure 16-30 shows the final filter circuit.
VIN
-~kA,
II i i_~ ~---II
-II
l~176
11~176
~- _~ 1 T
2.10k
VIN
II i i_~ ~---II
-II
l~176
11~176
~- _~ 1 T
VouT
3.16k
1.65k
-~kA,
2.10k
.. -
VouT
3.16k
Faa a simulao !!!!
Prof. Robertow M. Finzi Neto
12/12/14 1
Figure
16-30.
Third-Order
Unity-Gain
Bessel
High-Pass
2.10k
.. VouT
100n
VIN
.. -
1.65k
-~kA,
Figure 16-30
100n shows the final filter circuit.
II i i_~ ~---II
-II
l~176
11~176
~ _~ T
3.16k
39
0
-3
^ -.~
.~-0
0/
-3
-,.
/I
~..0
/
corner frequency of the low-pass filter transforms to the lower and upper-3 dB free. Projeto
de
Filtros
Passa
Banda (PBn)
.~-0
1
~
~..0
v
ncies of the band-pass, ~21 and .0-2. The difference between both frequencies is deu Representa a unio de dois filtros, um PB e outro PA, de maneira a
d as the normalized bandwidth A.Q:
<
<
<
<
analogy to the resonant circuits, the quality factor Q is defined as the ratio of the mid
(38)
quency (fm) to the bandwidth (B)"
IAI [dB]'
~
~
.
~
_
Q mid
= fmfrequency,
=
fm where
=
normalized
Q1 = 1, =is" 1
B
f 2 - fl
~'~2 - ~'~1
A~'-~
(39)
0
-3
^ -.~
/,:~<, ~
.~ . ~
. ...~.~.,~-
<
0/
(16-8)
-3
~.~
-,.
/I
Qm = 1 = Q2.~1
e simplest design of a band-pass filter is the connection of a high-pass filter and a low- /
s filter in series, which is commonly done in wide-band filter applications. Thus,~ a first-~..0
nalogy to the resonant circuits, the quality factor Q is defined.~-0as the1 ratio
of the mid
er high-pass
and aM.first-order
band-pass,
while
a
Prof. Roberto
Finzi Neto low-pass provide a second-order
12/12/14
uency
(fm)
to
the
bandwidth
(B)"
ond-order high-pass and a second-order Figure
low-pass
in a fourth-order
16-31.result
Low-Pass
to Band-Pass band-pass
Transition
v
40
o n d - O r d e r Band-Pass Filter
16.5.1 S e c oTondevelop
d - O r dtheefrequency
r Band-Pass
Filter band-pass filter, apply the transresponse of a second-order
16.5.1 S e c o n d - OA0
rder
A(s) - 1 + s
Band-Pass
Filter1)Design
Replacing
s with A--!-~(s+
Band-Pass Filter
A0
A(s) - 1 + s
To develop
the frequency
response
of a second-order
e. Projeto
de Filtros
Passa
Banda
(PBn)band-pass filter, apply the transA~
A(s) = 1 transfer
+ A ~ . s + Sfunction"
2
d-Order Band-Pass formation
Filter
in Equation 16-7 to a first-order low-pass
I. Segunda
Ordem
(16-9)
Replacing
s with filters,
A--!-~(s+
1) of interest are the gain at the mid freWhen
designing band-pass
the parameters
with A--!-~(s+
1)a second-order band-pass
oReplacing
develop thesfrequency
response of
filter,
apply
the
transquency (Am) and the
quality factor
which represents
the selectivity of a band-pass
u A partir de um filtroA0PB podemos
obter
o(Q), PBn
executando
a
filter.
rmation in Equation 16-7 to a first-order
function"
A(s)low-pass
- 1 + transfer
s
yields the
general do
transfer
function
for a second-order bandtransformao abaixo na funo Therefore,
transferncia
filtro
replace Ao with Am
and A.O.
with 1/QPB.
(Equation 16-7) and obtain:
yields the general transfer function for a second-order
band-pass
filter:
A0
Am
A(s) =
-e-.s
s with A--!-~(s+ 1)
(16-10)
A~
A(s) - 1 +Replacing
s
(40)
1 + $.s + S 2
A~
A(s)LJ = 1 + A ~ . s (16-9)
+ S2
Funo
Transf.
A(s)
=yields
1 +theA general
~ . s + Substituir
S2
Figure
16-32 shows the normalized
gain response
s for
transfer
function
a second-order
band-pass
filter:of a second-order band-pass filter for
eplacing s withFiltro
A--!-~(s+
1)
different Qs.
Funo
Transf.
PBn
PB
the parameters
of interest
pelo termo acimaWhen designing band-pass filters, Filtro
A~
(16-9) represents the
When designing band-pass filters, the parameters
of
interest
at the
mid(Q),
fre-which
quency
(Am)are
andthe
thegain
quality
factor
/
= substituir
1 + A ~band-pass
+AS02 por
elds
the general
transfer
function
for
aA(s)
second-order
rl
/
u
Em
Eq.
filter.filter:the selectivity
quency
(Am)
and
the (40),
qualityDevemos
factor
(Q),
which . srepresents
0=1/ of a band-pass
-5
-10
Am para representar o ganho na banda
/ , are the gain at\ the mid freWhen
designing
band-pass
filters,
the
parameters
of
interest
A~
(16-9)
Therefore,
replace
J Ao with Am and A.O.with 1/Q (Equation 16
-15
mdia
aplicar
Eq.
(39)
para
incluir
o
f
A(s)
= 1 +equency
A~.s
+
S
2
(Am) and the quality factor (Q), which represents
the selectivity
of a band-pass
I
/
fator deAo
qualidade.
Ento:
Am
Therefore, replace
with
Am
and
A.O.
with
1/Q
(Equation
16-7)
"~ -20 and obtain:
filter.
\
-e-.s
When designing band-pass filters, the parameters of interest are the gain~Iat-25A(s)
the mid
frek
= /
Amreplace Ao with Am and A.O.with 1/Q (Equation 16-7)
Therefore,
and
obtain:
1 + $.s + \ S 2
uency (Am) and the quality factor (Q), which represents the selectivity of a band-pass
,/
LJ
-30
A(s) =
-e-.s
(16-10)
/
Am
\
ter.
(41)
1 + $.s A(s)
+ S 2=
/
-e-.s
Figure 16-32-35shows
the normalized gain (16-10)
response of a secon
LJ
herefore, replace Ao with Am and A.O.with 1/Q 1(Equation
16-7)
and
obtain:
+ $.s + S 2different Qs.
0.1
LJ
1
lO
filter.
II1
41
Prof. Roberto
M. Finzi
12/12/14
Frequency- filter for
Figure 16-32
shows
the
normalized gain response
of a second-order band-pass
AmNeto
Figure
response
second-order
band-pass
A(s) =
-e-.s 16-32 shows the normalized
Figure 16-32.gain
Gain
Response of
of aaSecond-Order
Band-Pass
Filter filter for
(16-10)
different Qs.
%%-
n-Key Topology
%-
Figure16-33. Sallen-KeyBand-Pass
Vow
Figure16-33.
Sallen-KeyBand-Pass
Figure16-33.
Sallen-KeyBand-Pass
Figure16-33. Sallen-KeyBand-Pass
%-
TheSallen-Key
Sallen-Key band-pass
circuit
in in
Figure
16-33
has has
the following
The
band-pass
circuit
Figure
16-33
The
Sallen-Key
band-pass
circuit
inG.RC~m.S
Figure
16-33
hasthe
thefollow
follot
A(s)
=
G ) s + R2C2(.Om2.S2
A(s)
G.RC~m.S
A(s) == 1 + RCe0m(3G.RC~m.S
A(s) = 1 + RCe0m(3- G.RC~m.S
s )+s R2C2(.Om2.S
2
1 + RCe0m(3-G ) G
+ R2C2(.Om2.S
2
I. Segunda Ordem
1 + RCe0m(3G ) s 16-10,
+deR2C2(.Om2.S
u Podemos usar a topologia de Sallen-Key
paracomparison
implementar
o PBn
2obtain the2f
Through coefficient
with Equation
. Sallen-KeyBand-Pass
Through
coefficient comparison with Equation 16-10, obtain the follo
ordem
usando response
redes RCofem
PAThrough
e PB. coefficient
ph shows that
the frequency
second-order
band-pass
filters gets
comparison
with Equation
16-10, obtain the
1
Through
coefficient
comparison
with
Equation
16-10, obtain th
he Sallen-Key
band-pass
circuit
in
Figure
16-33
has
the
following
transfer
function:
mid-frequency:
fm
=
2~RC
se torna a Eq. (41), onde:
1
with risingu
Q,Sua
thusfuno
making transferncia
the filter more selective.
mid-frequency: fm = 2~RC1
mid-frequency: fm = 2~RC
A(s) =
G.RC~m.S
1
mid-frequency:
fm
=
2~RC
(42)
a2
1 + RCe0m(3- G ) s + R2C2(.Om2.S2
(41)
Topology
inner gain:
R 16-10,
hrough coefficient comparison with Equation
1
mid-frequency: fm = 2~RC
a2
inner gain:
G = 1 + R~
%-
gain at fm"
Am = 3 - G
filter quality:
len-KeyBand-Pass
Q =
3-G
inner gain:
G = 1 +a 2
R~
G = 1 + R~
a2
G = 1 + R~
a 2 (43)
inner gain:
G = 1 G+ R~
gain at fm"
Am = 3G - G
(44)
gain at fm"
Am = 3 - G
G
Vow
gain at fm"
Am = 3 - G G
filter
quality:
gain quality:
at fm"
= 3 - G (45)
filter
QQAm
== 3
-G
3-G
filter quality:
Q =
3-G
The
Sallen-Key
circuit
has
the
advantage
that
quality
factor
The Sallen-Key
circuit
has the advantage
thethe
quality
factor
(Q)(Q
ca
filter
quality:
Q = that
3frequency
-G
inner
without modifying
modifyingthe
themid
midfrequency
(fm).
A drawba
inner gain
gain (G) without
(fm).
A drawback
The
circuit
has the
advantage that the quality factor (
and
Am
be
adjusted
independently.
andSallen-Key
Am cannot be
adjusted
independently.
42
12/12/14
inner
gain (G) without
the mid frequency
(fm). Afactor
draw
The
Sallen-Key
circuitmodifying
has the advantage
that the
quality
Care
must
when
GGapproaches
the
of of
3, 3,
because
Care
must
be taken
taken
when
approaches
thevalue
value
because
theA
and
Am
cannot
be adjusted
independently.
inner
gain
(G)
without
modifying
the mid
frequency
(fm).
Athen
draw
filter quality:
Q =
3-G
The Sallen-Key circuit has the advantage that the quality factor (Q) can be
inner gain (G) without modifying the mid frequency (fm). A drawback is, h
and Am cannot be adjusted independently.
e. Projeto de Filtros
2Am- 1
(47.b)
(47.a)
R2 = 2 Q - 1
a 2 - 1 +Am
Q
Use as Eq. (43), (44) e (45) para definir o ganho do filtro.
or to design for a specified
Q: Multiple Feedback Topology
16.5.1.2
Exerccio: projete um Filtro PBn com frequncia central cde 1kHz, largura
= 2Q- 1
de banda deR2
200Hz.
Q
12/12/14
VIN
a1
43
VOUT
Resoluo.
Com base em fm = 1kHz, podemos definir C = 100nF e calcular R.
1
R=
1592
3
9
2 10 100 10
Como foi especificada a largura de banda (200Hz), estamos fazendo um
projeto voltado ao fator de qualidade do filtro. Usando Eq. (39), temos Q.
Podemos calcular G e Am usando Eq. (43) e (44).
O Alto valor de Am implicar num valor limitado de Vin devido saturao do
amplificador operacional.
Definindo R = 1k e usando
Q=
fm 1000
=
=5
B
200
3Q 1 3 5 1
G=
=
= 2,8
Q
5
Am =
G
2,8
=
= 14
3 G 3 2,8
12/12/14
R2 = R1 ( G 1)
R2 = 1000 ( 2,8 1)
R2 = 1,8k
44
12/12/14
45
Q _
10
= 31 8 k Q
R2 - ~frnC
=. 1 kHz. 100 nF
"
R2
R1 = - 2Am
_- 31.8 kQ = 7.96 k ~
4
- AmR1
R3 = 2Q 2 + A m
2.7.96 kQ = 80.4
200-2
igure 16-32 shows that the frequency response of second-order band-pass filters gets
eeper with rising Q. However, there are band-pass applications that require a flat gain
Quarta Ordem
esponse close to the mid frequency as well as a sharp passband-to-stopband transition.
Filtros
PBnbyde
4 ordem
permitem
a especificao
hese tasks can be
accomplished
higher-order
band-pass
filters.
II.
u
do tipo da resposta.
Band-Pass Filter Des
A isserializao
delow-pass
doisto filtros
leva onto
ao filtro de 4 ordem com
a
f particular interest
the application of the
band-pass PBn
transformation
second-order low-pass
filter,transferncia
since it leads to a fourth-order
funo
abaixo,band-pass
onde: filter.
9 Ami
is the
gain at functhe mid
frequency,
fmi, of each
partial filter
eplacing the S term
Equation
with
Equation 16-7 gives
the
general
transfer
inAm
eQ
o ganho,
frequncia
mdia
e fator
de qualidade
do
i, fmi16-2
i representa
9
^.
..
..
on of a fourth-order band-pass:
ensimo
estgio, respectivamente.
9 Qi is the pole quality of each filter
s2.A0(AQ)
2
e 1/ so os
fatores
que9definem
o quo
distante
esto
fm1 e fm
doindividual
fm
o~and 1/0~are
the factors
by which the
mid frequencies
of2the
filters,
and fm2,com
derive(16-11)
from the
mid frequency,
fm, of
the
overall
alto
valor
de Q tm
fm
fm2bandpass.
muito
A(s) = de projeto do total
b, do filtro. Filtros
1 e
prximos
band-pass filter with high Q, the mid frequencies of the two partial filt
1 + ~a,
/ k Q . S + de[ 2f+m. (Aa)2]-S2
b~ _1 4- ~a,AInQ a' Sfourth-order
3 + s4
only(49)
slightly usando
from the overall
frequency.de
Thisaproximaes
method is called staggered tuni
u O fator calculado peladiffer
Eq.
o mid
mtodo
sucessivas,
com a1 transfer
e b1 sendo
os
coeficiente
do filtro PB de 2 ordem.
imilar to the low-pass
filters, the fourth-order
functionFactor
is split
two second-or~into
needs
to be determined through successive approximation, using equat
A(s) =
[os
1 +
][,
1 +
Ami s
(48)
(16-12)
b1
= 0
(49)
46
(16-1
12/12/14
Equation 16-12 represents the connection of two second-order
filtersthe
in sewith band-pass
al and bl being
second-order low-pass coefficients of the desired filter type
ies, where
= 0 of filter 1 is:
The mid bfrequency
1
Band-Pass FilterDesign
(16-13)
fm
fml = -E
with al and bl being the second-order low-pass coefficients of the desired filter type.
e. Filter
Projeto
Band-Pass
Designde Filtros Passa Banda (PBn)
the
mid frequency
of
filter filters
2 is:
o~has
been16-2
determined,
all quantities
of theand
partial
simplify the filter After
design,
Table
lists those
coefficients,
providescan
thebe
o~calculated
values u
II. QuartaToOrdem
following
equations:
for three adifferent
Q =a1, tabela
Q = 10, and
Q = 100.
fm2
= fm'~
u Para facilitar
tarefaquality
de factors,
projeto,
abaixo
sumariza
as trs
sign
principais respostas The
commid os
respectivos
frequency
of filter 1 is: valores de , ai, bi e
Qsmid frequency of the overall forth
Table
16-2. Values of o~ For Different Filter Types and
withDifferent
fm being the
Q=1;10;100.
al
bl
al
1.4142
al
1.0000
bl
bl
The mid0.6180
frequency of filter
1 is:
1.0650
Qi = Q" (11.9305
+ (~2)b 1
a-a
1
o~has been
partial
filters
can
be
calculated
using
the
1
O
100
10
1
10 all quantities
1
Q of the
100 fm2
10
100
O determined,
= fm'~
A~
0.1
1
1
0.1
1
0.01
0.1
wing equations:0.01
fmA~
with Q being
the0.01
quality factor
of the
overall (16-1
filter.
mid
1.0032
1.0324
fml
= -E
1.438
1.4426 of theo~overall
1.0033
1.0338band-pass
1.39 filter.
1.0035the mid
1.036frequency
with fm being
forth-order
The individual gain (Ami) at the partial mid frequen
frequency
of
filter
1
is:
u Atravs
de frequency
calculamos:
The individual
pole quality, Qi,filters:
is the same for both filters:
the mid
of filter
2 is:
(X
fm
fml = -E (50) fm2 = fm'~ (51)
Prof. Roberto M. Finzi Neto
12/12/14
(16-14)
ll///Am
(53)
47
(16-1
with Q being of
thethe
quality
factorwith
of the
filter.
with fm being the mid frequency
overall
forth-order
band-pass
Am overall
being the
gain at mid filter.
frequency, fm, of the
================================
...............................................................
::::: ........ ~
................
:.
In accordance
with Equations 16-14 and 16-15, the mid
e. Projeto de Filtros Passa Banda
(PBn)
are:
II. Quarta Ordem
u Exerccio. Projete um Filtro PBn com freq. central de 10kHz, largura de
10 kHz maximamente
= 9.653 kHz planaand
banda de 1kHz, ganho unitriofmie -resposta
na faixa fro2 =
1.036
Band-Pass
Filter
Design
Band-Pass
Filter
Design
de passagem e resposta de ButterWorth.
Band-Pass
Filter Design
Resoluo.
The overall Q is defined as Q = frn/B, and for this exam
Da
tabela, obtemos
os 16-14
coeficientes:
b1 =
=1.036,
porque
In accordance
with Equations
and 16-15,athe
mid frequencies
for1thee partial
filters
1 =1.4141,
In accordance
with Equations
16-14 and 16-15,
the
frequencies
for
the partial for
filters
In accordance
andmid
16-15,
the mid
frequencies
the partial
Equation
16-16,
the Qi of both
filtersfilters
is:
Q are:
=10kHz/1kHz
=with
10,Equations
Eq. (39).16-14 Using
==========================
...............................................................
.........................................................
::::: ........ ~
================================
are:
::::: ........ ~
................
:.
................
...............................................................
:.
::::: ........ ~
................
:.
are:
Usando
as Eq.
(50)kHz
e (51) calculamos
as
freq. mdias de cada estgio.
fmi
-- 10 kHz
= 9.653
and
fro2 = 10 kHz.1.036 = 10.36 kHz
1.036
fmi -- 10 kHz
= 9.653 kHz
and
fmi
1.036 -- 10 kHz = 9.653 kHz
1.036
Qi = 10"(1 + 1"0362)'1
fro2 = 10 kHz.1.036
10.36
kHz= 10.36
and
1.6-36:1~1~,2
= kHz
14.15
fro2 = 10 =kHz.1.036
The overall Q is defined as Q = frn/B, and for this example results in Q = 10.
fator
dedefined
qualidade
o ganho
dois
estgio
sempre
iguais.
Seus
The O
overall
Q is
as Q = efrn/B,
and for dos
this
example
results
inso
Q =the
10.passband
With
Equation
16-17,
gain
of
the partial fil
The
overall
Q
is
defined
as
Q
=
frn/B,
and
for
this
example
results
in
Q
=
10.
Using Equation 16-16, the Qi of both filters is:
clculos so feitos abaixo:
Am i = 14.1510
= 1.415
Qi = 10"(1 + 1"0362)'1
Qi = 10"(1
Prof. Roberto M. Finzi Neto1.6-36:1~1~,2
+= 1"0362)'1
48
14.1512/12/14
1.6-36:1~1~,2
= 14.15
With Equation 16-17, the passband
gain of the
partial
filters at fml
and and
fm2 calculates
to: that Qi and
The
Equations
16-16
16-17 show
Ami of
1.036between Q and Am, there are two options to solve for R2: eise of the dependency
Qi = 10"(1 + 1"0362)'1
Using
Equation
16-16, the Qi of both filters is:
set the gain
atThe
mid
frequency:
1.6-36:1~1~,2
overall Q is defined as Q = frn/B, and for this example results in Q = 10.
ns 16-14 and 16-15, the mid frequencies for the partial filters
10"(1the+ Qi1"0362)'1
2Using
A m Equation
- Qi
1 =16-16,
of both filters is:
1.6-36:1~1~,2
= 14.15
a 2 - 1 +Am
and
fro2 = 10 kHz.1.036 = 10.36 kHz
Qi = 10"(1 + 1"0362)'1
= 14.1
Am i = 14.1510
With Equation 16-17, the 1.6-36:1~1~,2
passband gain=of14.15
the partial filters at fml and fm2 calculates to:
= 1.415
Q = frn/B,
for this example
esign
forand
a specified
Q: results in Q = 10.
The fm2
Equations
16-16
With Equation
the passband
gain ofBanda
the partial filters
at fml and
calculates
to: and 16-17 show that Qi and
e. Projeto
de16-17,
Filtros
Passa
(PBn)
= 1.415
Am i = 14.1510
II. =Quarta
R2
2 Q - Ordem
1
ter in Paragraph 16.5.1.2.
= 1.415
Am i = 14.1510
Q
+ 1"0362)'1u A especificao de Am e Q para um mesmo filtro PBn leva ao problema de no ser
i
i
6-36:1~1~,2 = 14.15
The Equations 16-16 and 16-17 show that Qi and Ami of the partial
filters need
to be indeTo design
the individual
second-order band-pass
possvel usar a topologia de Sallen-Key. Teremos de usar
a topologia
Multi
Feedpreviously
determined
for the partial f
pendently
adjusted.
Theand
only
circuit
that
accomplishes
thispartial
task filters
is
theneed
MFBto
band-pass
filThe Equations
16-16
16-17
show
that
Qi and Ami of the
be inde- quantities
Back
(MFB).
Definindo
C=10nF,
as
Eq.
e
clculos
so
apresentados
abaixo.
eedback
Topology
assband gain of the pendently
filters
at fml and
calculates
to: accomplishes this task is the
MFB
band-pass
filter.
adjusted.
Thefm2
only
circuit that
MFB
band-pass
fil- The resistor values for bot
ter inpartial
Paragraph
16.5.1.2.
ter in Paragraph
c 16.5.1.2.
F i l t e r 1"
Filter 2:
= 1.415To design the individual second-order band-pass filters, specify C = 10 nF, and insert the
0
To design the individual second-order band-pass filters, specify Qi
C = 10 nF, and14.15
insert the = 46.7 kQ
R22 Q
previously determined quantities for the partial filtersR21
into
the
resistor
equations
of the
:TL.fmlC
=.9.653
kHz.
10
nF
=fm
previously determined quantities for the partial filters into the resistor equations of the
6-17 show that Qi
and band-pass
Ami of the partial
filters need
to be indeMFB
filter.
forboth
bothpartial
partial
filters
calculated
MFB band-pass
filter.The
Theresistor
resistor values
values for
filters
are are
calculated
below.below.
a1
VIN
F i l t e r F i l1"
t e r 1"
FFi il lt teer r 2 :2 :
VOUT
al I --
R21
-
2Ami
46.7 k~
- 2.-
1.415
1 6 . 5 kf~
R12
Qi Qi filters, 14.15
14.15C = 10 =nF,
=46.7
46.7insert
kQ theR22
14.15
kQ
R22- - Q~Q~ - 14.15 = 43.5
ond-order band-pass
specify
and
= kQ
43.5 kQ
- :TL.fmlC
=.9.653
kHz.
10
nF
=fm2C
=.
10.36
kHz.
10 nF
R3
R21
- R21
:TL.fmlC
AmiRll
1.415.16.5 kQ
=.9.653
kHz.
10
nF
=fm2C
=.
10.36
kHz.
10
nF
ntities for the partial filters into the resistor equations of the
R31 2Qi 2 + Am~ = 2"14"152 + 1.415 = 58.1 Q R32 = 2Q
esistor values for both partial filters are calculated below.
R21
46.7 k~
R22
43.5 k~
46.7
R22 - 2 . - 43.5
a l I FR21
- i- l t e-r 2Ami
- 2 . -k~ 1 . 4 1 5
1 6 . 5 kf~
R12
- 2Ami
1 . Figure
4 1k~
5
1 5 . 4 kf~
2:
16-35
compares the gain response of a fo
a l I - - - 2Ami
- 2.1.415
1 6 . 5 kf~
R12
- 2Ami
- 2.- 1.415
1 5 . 4 kf~
with Q = 1 and its partial filters to the fourth-orde
6.7 kQ
R22 Q~ 14.15
= 43.5 kQ
- AmiRll
1.415.16.5
- AmiR12
1.415.15.4 k~
=fm2C =. 10.36
kHz. 10kQ
nF
= 54.2 Q
R31- AmiRll
2Qi
2
+
Am~
=
2"14"152
+
1.415
=
58.1
Q
R32
=
=
1.415.16.5
kQ
AmiR12
1.415.15.4
2Qi- 2 +
Am~ 2"14"152
+ 1.415 k~
49
Prof. Roberto M. Finzi
Neto
12/12/14
FBBand-Pass
R31band-pass
2Qi 2 + Am~
= 2"14"152
+ 1.415
= 58.1
R32following
= 2Qi 2 + transfer
FB
circuit
in43.5
Figure
16-34
hasQ the
function:
Am~ = 2"14"152
+ 1.415
R22
k~
= 54.2 Q
Q _
10
= 31 8 k Q
R2 - ~frnC
=. 1 kHz. 100 nF
"
R2
R1 = - 2Am
_- 31.8 kQ = 7.96 k ~
4
- AmR1
R3 = 2Q 2 + A m
2.7.96 kQ = 80.4
200-2
e.
h-Order Band-Pass
Filter (Staggered
Tuning)Passa
Projeto
de Filtros
II.
Banda (PBn)
A(s) =
[os
1 +
+
Prof. Roberto M. Finzi Neto
][,
1 +
Ami s
(48)
(16-12)
12/12/14
Equation 16-12 represents the connection of two second-order band-pass filters in se-
50
IAI [dB]
16.6
Band-Rejection
F i l t e 0r
L esign
D
0-
IAI [dB]
-3
-3
\ ra
A
A
A band-rejection filter r is used
to suppress a certain frequency
0
0i I
quencies.
-3
-3
\
Two of the most popular band-rejection filters are the active twin
i I
Robinson circuit,
both of which are second-order filters.
Banda
(RBn)
L_
r
normalized
bandwidth of
A.O.:
The
corner frequency
the low-pass transforms to the low
cies of the band-rejection
~1
s2)and -0.2. The difference be
As = ~ mA~
a xfilter
- + ~"2min
(54)
A(s) = 1 A.O.:
+ Af~.s + s 2
normalized bandwidth
Identical to the selectivity of a band-pass filter, the quality of
(55)
As = ~ m a x ~"2min
as"
IAI [dB]
0
-3
0-
-3
~1-0
L_
r
296
<1-0
Af~
Therefore, replacing
16-19 with 1/Q yields"
Q
fmA.O. in Equation
1
B
Af~
51
A(s) = A~ + s2)
12/12/14
Therefore, replacing A.O.
1 +in~Equation
. s + S 2 16-19 with 1/Q yields
width A.O.:
Band-Rejection
Filter Design
Therefore, replacing A.O. in Equation 16-19 with 1/Q
yields"
= ~max~"2min
A(s) = A~ + s2)
electivity of a band-pass filter, the1 quality
+ ~ . s +of Sthe
2 filter rejection is defined
(16-20)
C
C
F~2
16.6.1
fm Active
1 Twin-T Filter
V,N
B
Af~The original twin-T filter, shown in Figure 16-37, is a passive RC-network with a quality
factor of Q = 0.25. To increase Q, the passive filter is implemented into the feedback loop
Projeto
cing A.O. in Equation
16-19de
withFiltros
1/Q yields"Rejeita Banda (RBn)
of an amplifier, thus turning into an active band-rejection filter, shown in Figure 16-38.
s) =
f.
--[- 2c
1 + ~ . s + S2
R1
V,N
k(1 + s 2)
1 +2(2-k).s+s
2
VOUT
--[- 2c
vO T
12c
A(s) =
R1
mid-frequency"
fm =
win-T Filter
52
12/12/14
1
2~RC
inner gain:
a2
G = 1 + R---~
Design
Comparing the variables of Equation 16-21 with Equation 16-20 provides the equations
VOUT
--[- 2c
passband gain"
A o =theGfilter parameters:
that determine
C
VOUT
--[- 2cR
1 +2(2-k).s+s
R1
mid-frequency"
fm =
1
1
2~RC
Figure
16-38.
Active
TwinT Filter
C
C
rejection quality: Q = 2(2 - G)
F~2
F~2
R1
a2
inner gain:
G = 1 + R---~
The
transfer
function
ofcan
thebe
active
twin-T
filter
is:
The twin-T circuit
that
the
quality
factor
(Q)
varied
via
the
inner
V,NR has the advantage
passband gain" A o = G
Figure
16-38.
Active
T Filter
gain (G)
without
modifying
the midTwinfrequency
(fm). However, Q and Am cannot be adjusted
RVOUT
Figure 16-38.
Active
Twin-Rejeita
T Filter
--[- 2c
f.
Projeto
de
Filtros
Banda
(RBn)
1
k(1 + s 2)
rejection quality: Q A(s)
= 2(2 - =
G)
independently.
a. Topologia
Active
twin-T
VOUT
1 filter
+filter
2 (is:2 is:
-k).s+s
2
--[2c
The transfer
function
of
the
active
twin-T
The
transfer
function
of
the
active
twin-T
The
twin-T
circuit
has
the
advantage
that
the
quality
factor
(Q)
can
be
varied
via
the
inner
band-pass,desta
To set the
frequencytransferncia
of the
specifytopologia
fm and C, and
then solve for
R:Eq. (54) e est
umid
A funo
Qderivada
gain (G) without modifying
the mid frequency (fm). However,
and Am cannot beda
adjusted
R1
independently.
descritas
(59)
e++ (60).
A(s)=(58),
=
k(1variables
ss2)2) of Equation 16-21 with Equation
(16-21)
Comparing
the
16-20
pr
R=
1 nas Eq. (57),
A(s)
k(1
(16-21
Band-Rejection
set the mid frequency
of the+ band-pass,
specify
fm and
and then solve for R:
2 C,C,
R1
2~fmC
u O projeto
do Tofiltro
envolve
e
1 1especificar
+ 22((22the
-- kk))..sfilter
s++s sfmparameters:
2 para ento resolver em R
that
determine
R=
1
usando Eq.Comparing
(58). the
2~fmC
Equation
16-21
with Equation
provides
the equations
erBecause of the dependency between Qvariables
and Am,ofthere
are two
options
to solve16-20
for R2:
eithe
variables
of
Equation
16-21
with
Equation
the equation
u DevidoComparing
dependncia
QQ and
e Am,
A
podemos
calcular
das duas
mid-frequency"
= for R2:16-20
1ei- R2 provides
determine
theentre
filter
parameters:
or tobetween
design
for
am
specific
Because
of the dependency
there
are twoQ:
optionsfm
to solve
ther to set the gain at midthat
frequency:
2~RC
ther to
set the
gain filter
at mid
frequency:
formas
abaixo.
Eq.
(61)
e
(62)
that
determine
the
parameters:
the active twin-T filter is:
Active
Twin-T Filter
mid-frequency"
fm =
1
(62)
= (Ao- 1)R,
~. ~
= (Ao- 1)R,
9 .
.-
(61)
2~RC
a2
k(1 + s 2)
(16-21)
mid-frequency"inner
fm gain:
=
1
G
=
1
+
R---~
2~RC
298 active twin-T filter is:
ransfer
+ 2 ( 2 - kfunction
) . s + s of 2the
a2
gain:
G = 1 + Filter
R---~
16.6.2 inner
Active
Wien-Robinson
s of Equation
16-20 provides
the
equations
ain2 Figure
A(s)16-21
= with Equation
k(1 + s 2)
The
Wien-Robinson
bridge
16-39
passive band-rejection filter w
passband
gain"
A o(16-21)
= G is a(59)
(57)
inner
gain:
G
=
1
+
R---~
passband
gain"
A
o
=
G
parameters:
1 +2(2-k).s+s
2 tial output. The output voltage is the difference between the potential of a con
determineProf.
theRoberto
filter M.
parameters:
Finzi Neto
n:
a2
G = 1 + R---~
12/12/14
VIN
53
1
The twin-T circuit has the advantage that the quality
factor (Q) can be varied via the inner
R~
~ 2 R 1 factor (Q)
rejection
quality:
Q
=
2(2
G)
The twin-T
the advantage
the
gain (G) without
modifyingcircuit
the midhas
frequency
(fm). However,that
Q and
Amquality
cannot be
adjusted
can b
R=
1
1
=
1592
9
2 fm C 2 10000 10 10
1
1
R2 = R1 1
= 1000 1
= 950
2 10
2 Q
A simulao resulta apresentada a seguir.
Prof. Roberto M. Finzi Neto
12/12/14
54
12/12/14
55
g. Exerccios
1. O computador de bordo de uma aeronave militar usa o Sistema de Navegao
12/12/14
56
Table16-4.
COEFICIENTES PARA
FILTRO DE BESSEL
n ordem do filtro
i nmero de filtros
parciais.
ai , bi coeficientes do
estgio do filtro.
fci frequncia de corte
do estgio de filtro.
ki razo entre o fci e fc do
estgio.
Qi Fator de qualidade do
estgio.
Os dados dos filtros de
ordens 9 e 10 so
encontrados na pgina 316
do livro.
Prof. Roberto M. Finzi Neto
BesselCoefficients
a i
1.0000
0.0000
1.000
1.3617
0.6180
1.000
0.58
1
2
0.7560
0.9996
0.0000
0.4772
1.323
1.414
0.69
bi
ki=
fci I fc
Qi
1.3397
0.7743
0.4889
0.3890
0.978
1.797
0.52
0.81
1
2
3
0.6656
1.1402
0.6216
0.0000
0.4128
0.3245
1.502
1.184
2.138
0.56
0.92
1
2
3
1.2217
0.9686
0.5131
0.3887
0.3505
0.2756
1.063
1.431
2.447
0.51
0.61
1.02
1
2
3
4
0.5937
1.0944
0.8304
0.4332
0.0000
0.3395
0.3011
0.2381
1.648
1.207
1.695
2.731
w
0.53
0.66
1.13
1
2
3
4
1.1112
0.9754
0.7202
0.3728
0.3162
0.2979
0.2621
0.2087
1.164
1.381
1.963
2.992
0.51
0.56
0.71
1.23
0.5386
1.0244
0.0000
0.2834
1.857
1.277
12/12/14
1
2
57
n
0.52
Filter Co
Table 16-5.
COEFICIENTES PARA
FILTRO DE BUTTERWORTH
n ordem do filtro
i nmero de filtros
parciais.
ai , bi coeficientes do
estgio do filtro.
fci frequncia de corte
do estgio de filtro.
ki razo entre o fci e fc do
estgio.
Qi Fator de qualidade do
estgio.
Os dados dos filtros de
ordens 9 e 10 so
encontrados na pgina 317
do livro.
Prof. Roberto M. Finzi Neto
Butterworth Coefficients
ai
1.0000
0.0000
1.000
1.4142
1.0000
1.000
0.71
1.0000
1.0000
0.0000
1.0000
1.000
1.272
1.00
1
2
1.8478
0.7654
1.0000
1.0000
0.719
1.390
0.54
1.31
1
2
3
1.0000
1.6180
0.6180
0.0000
1.0000
1.0000
1.000
0.859
1.448
0.62
1.62
1
2
3
1.931 9
1.4142
0.5176
1.0000
1.0000
1.0000
0.676
1.000
1.479
0.52
0.71
1.93
1
2
3
4
1.0000
1.801 9
1.2470
0.4450
0.0000
1.0000
1.0000
1.0000
1.000
0.745
1.117
1.499
0.55
0.80
2.25
1.9616
1.6629
1.1111
0.3902
1.0000
1.0000
1.0000
1.0000
0.661
0.829
1.206
1.512
0.51
0.60
0.90
2.56
1.0000
1.8794
0.0000
1.0000
1.000
0.703
1
2
3
4
"
12/12/14
1
2
bi
ki=
fci/fc
Qi
58
0.53
Table 16-6.
COEFICIENTES PARA
FILTRO DE Tschebyschef
COM 0,5dB de Ripple
n ordem do filtro
i nmero de filtros
parciais.
ai , bi coeficientes do
estgio do filtro.
fci frequncia de corte
do estgio de filtro.
ki razo entre o fci e fc do
estgio.
Qi Fator de qualidade do
estgio.
Os dados dos filtros de
ordens 9 e 10 so
encontrados na pgina 318
do livro.
1.0000
0.0000
1.000
1.3614
1.3827
1.000
0.86
1
2
1.8636
0.0640
0.0000
1.1931
0.537
1.335
1.71
1
2
2.6282
0.3648
3.4341
1.1509
0.538
1.419
0.71
2.94
1
2
3
2.9235
1.3025
0.2290
0.0000
2.3534
1.0833
0.342
0.881
1.480
1.18
4.54
1
2
3
3.8645
0.7528
0.1589
6.9797
1.8573
1.0711
().366
1.078
1.495
0.68
1.81
6.51
1
2
3
4
4.0211
1.8729
0.4861
0.1156
0.0000
4.1795
1.5676
1.0443
0.249
0.645
1.208
1.517
1.09
2.58
8.84
5.1117
0.276
0.68
2
3
4
1.0639
0.3439
0.0885
11.960
7
2.9365
1.4206
1.0407
0.844
1.284
1.521
1.61
3.47
11.53
1
2
5.1318
2.4283
0.0000
6.6307
0.195
0.506
12/12/14
ai
bi
ki =
fci I fc
Qi
59
m
1.06
Filter Co
COEFICIENTES PARA
FILTRO DE Tschebyschef
COM 1dB de Ripple
n ordem do filtro
i i-simo filtro parcial.
ai , bi coeficientes do
estgio do filtro.
fci frequncia de corte
do estgio de filtro.
ki razo entre o fci e fc do
estgio.
Qi Fator de qualidade do
estgio.
Os dados dos filtros de
ordens 9 e 10 so
encontrados na pgina 319
do livro.
Prof. Roberto M. Finzi Neto
ai
1.0000
0.0000
1.000
1.3022
1.5515
1.000
0.96
1
2
2.2156
0.5442
0.0000
1.2057
0.451
1.353
2.02
1
2
2.5904
0.3039
4.1301
1.1697
0.540
1.417
0.78
3.56
1
2
3
3.5711
1.1280
0.1872
0.0000
2.4896
1.0814
0.280
0.894
1.486
1.40
5.56
1
2
3
3.8437
0.6292
0.1296
8.5529
1.9124
1.0766
0.366
1.082
1.493
0.76
2.20
8.00
1
2
3
4
4.9520
1.6338
0.3987
0.0937
0.0000
4.4899
1.5834
1.0432
0.202
0.655
1.21 3
1.520
1.30
3.16
10.90
5.1019
0.276
0.75
2
3
4
0.8916
0.2806
0.0717
14.760
8
3.0426
1.4334
1.0432
0.849
1.285
1.520
1.96
4.27
14.24
1
2
6.3415
2.1252
0.0000
7.1711
0.158
0.514
12/12/14
bi
ki=
fci/fc
Qi
60
m
1.26
COEFICIENTES PARA
FILTRO DE Tschebyschef
COM 2dB de Ripple
n ordem do filtro
i i-simo filtro parcial.
ai , bi coeficientes do
estgio do filtro.
fci frequncia de corte
do estgio de filtro.
ki razo entre o fci e fc do
estgio.
Qi Fator de qualidade do
estgio.
Os dados dos filtros de
ordens 9 e 10 so
encontrados na pgina 320
do livro.
Prof. Roberto M. Finzi Neto
1.0000
0.0000
1.000
1.1813
1.7775
1.000
1.13
1
2
2.7994
0.4300
0.0000
1.2036
0.357
1.378
2.55
1
2
2.4025
0.2374
4.9862
1.1896
0.550
1.413
0.93
4.59
1
2
3
4.6345
0.9090
0.1434
0.0000
2.6036
1.0750
0.216
0.908
1.493
1.78
7.23
3.5880
0.373
0.90
2
3
0.4925
0.0995
10.464
8
1.9622
1.0826
1.085
1.491
2.84
10.46
1
2
3
4
6.4760
1.3258
0.3067
0.0714
0.0000
4.7649
1.5927
1.0384
O.154
0.665
1.218
1.523
1.65
4.12
14.28
4.7743
0.282
0.89
2
3
4
0.6991
0.2153
0.0547
18.151
0
3.1353
1.4449
1.0461
0.853
1.285
1.518
2.53
5.58
18.39
8.3198
1.7299
0.0000
7.6580
O.120
0.522
12/12/14
2
ai
bi
ki =
fci/fc
Qi
61
1.60
Filter Co
COEFICIENTES PARA
FILTRO DE Tschebyschef
COM 3dB de Ripple
n ordem do filtro
i i-simo filtro parcial.
ai , bi coeficientes do
estgio do filtro.
fci frequncia de corte
do estgio de filtro.
ki razo entre o fci e fc do
estgio.
Qi Fator de qualidade do
estgio.
Os dados dos filtros de
ordens 9 e 10 so
encontrados na pgina 321
do livro.
Prof. Roberto M. Finzi Neto
1.0000
0.0000
1.000
1.0650
1.9305
1.000
1.30
1
2
3.3496
0.3559
0.0000
1.1923
0.299
1.396
3.07
1
2
2.1853
o. 1964
5.5339
1.2009
0.557
1.410
1.08
5.58
1
2
3
5.6334
0.7620
0.1172
0.0000
2.6530
1.0686
o. 178
0.917
1.500
2.14
8.82
3.2721
0.379
1.04
2
3
0.4077
0.0815
11.677
3
1.9873
1.0861
1.086
1.489
3.46
12.78
1
2
3
4
7.9064
1.1159
0.2515
0.0582
0.0000
4.8963
1.5944
1.0348
0.126
0.670
1.222
1.527
1.98
5.02
17.46
4.3583
0.286
1.03
2
3
4
0.5791
o. 1765
0.0448
20.294
8
3.1808
1.4507
1.0478
0.855
1.285
1.517
3.08
6.83
22.87
0.0000
0.098
ai
bi
ki =
fci/fc
12/12/14
1
10.175
9
Qi
62
12/12/14
63