622 Active Filters ( Edgar Sánchez-Sinencio)

• A second order filter is also known as a Biquad.

• A second – order filter consists of a two integrator loop of one lossless and one
lossy integrator
• Using ideal components all the biquad topologies have the same transfer
function.
• Biquad with real components are topology dependent .

 We will cover the following material:

- Biquad topologies
- Voltage scaling to equalize signal at all filter nodes.
- State variable Biquad derivation.
- Tow-Thomas Biquad Design Procedure
- Fully Balanced topologies using single ended Op Amps
- Non-ideal integrators and Biquads.
- Mason Rule to obtain transfer functions by inspection
1
 TWO-INTEGRATOR LOOP FAMILIES
Two-integrator loops consists of one lossless and one
lossy integrator, furthermore one is positive and the other is negative.

Why do we consider different filter topologies to obtain

the same (ideal) transfer function ?

 o 2
 KQ

1  o1
  o

Family 1a

 o 2

 o1
1
 s  KQ


2
  o 2
 KQ

1  o1
  

 o 2

1  o
 s  KQ

Family 1b

KQ o1 o 2
D(s)  1  
s s2
s 2D (s)  s 2  K Q s  o1 o2 ; S BW
K Q  1
2
All loops must be negative. 0
S 1 3
o1
 Topology using two lossy integrators

 o 2
 K Q1  KQ2

1  o1
 1
s  1
s 

• Lossy integrators are easy

to implement

 o 2

o1
1 1
 s  K Q1 s  KQ2

Two-Integrator Loop Family 2, ideal to avoid CMF in Fully Differential Structures

4
 K Q1 KQ2 o1 o 2 K Q1 K Q 2
D(s)  1    2

s s s s2

s2D(s)  s2  (K Q1  K Q2 )s  (o1 o 2  K Q1 K Q2 )

o
o2  o1 o 2  K Q1 K Q2 BW   K Q1  K Q 2
Q

Low sensitivity structure:

K Q1 K Q1 1
KQ  1
SBW KQ 
; SBW 
1 BW 1 K Q1  K Q 2 KQ2
1
K Q1

o2 o1 1
S  o 2  
o1 o1  o 2  K Q1 K Q 2 K Q1 K Q 2
1
o1 o 2
5
 Self-Loop two integrator loop

Design Equations: Given o , BW, PICK

K Q1  BW  K Q2 K Q 2 and o 2

o2  o1 o 2  (BW  K Q2 )K Q2

o1 o 2  o2  (K Q2  BW )K Q2

 o 2

1  o1
 1
s
1
s  K Qo1

s

Family 3 Self-Loop

o1 o 2 K Qo1 A self loop can be implemented by

D(s)  1  2
 adding a resistor to one (lossless)
s s
integrator.
s2D(s)  s2  K Qo1s  o1 o 2
6
 SCALING for Active-RC, MOSFET-C
and SC implementations
Zo1

ZQ
ZC1

ZK  Vo 2 
o  Vo 3
 
 Vo1 Zo 2
Z1'

Z1
An example.
Modify all the impedances connected
Vo1  k Vo1
to the output under consideration.

1
ZC1  k ZC1  k   C1  C1 / k
SC1 SC1 / k

Z1'  k Z1'  k R1'  R1'  k R1'

Note that the voltages V02 and V03 were not modified
7
 VOLTAGE SWING SCALING
A good filter design approach requires to have a proper voltage swing
at a frequency range for all internal and output nodes of the filter.

Let us consider the different types of filter implementations.

R o1 RQ
R1

C1 C2
RK R1' 
  Vo 3
o o
Vi  Vo1  Vo 2 R o 
2

1

R o1 C1 1
1 1 
 R QC2
R K C1 1 R o 2 C2
o  1
s o o  1
s o
Vin Vo1 Vo 2 Vo 3
1 1 R
s 2D(s)  s 2  s   1'
R QC2 R o1 R o 2 C1C2 R1
1  1  2 1  1 
 1  s  s 
Vo1 R K C1s  R Q C 2s  R k C1  R Q C 2 
H1 (s)   
Vin D(s) 1 1
s2  s 
R Q C 2 R o1 R o 2 C1C 2 8
 Vo 2
H 2 (s)   H1(s)
Vin
1
Vo3 R K C1R o 2 C2
H3 (s)  
Vin S2  S  1
R QC2 R o1 R o 2 C1C2

Let us consider that we want

H3 ( jo )  K
1
R K C1R o 2 C2 RQ Q
H3 ( jo )   K 2
o R K C1R o 2 o o R K C1R o 2 C2
R QC2
Q
K 
0R K C1
1 1
o  
R o1 C1 R oC2

RQ 1 R1
Thus RK  ; o2  
KC1R o 2 o R o1 R o 2 C1C2 R1' 9
 1  1  R QC2 
 1 
 j    j  
R K C1  R QC2  R K C1  R QC2 
o o
H1 ( jo )  
o jo
j
R QC2

2
R QC2 2   1  RQ 1
H1 ( jo )  0     1 
R K C1o  R QC2  R K C1 Q2

NUMERICAL EXAMPLE

3
o  1 Q K2
4

10
THEN

1  1 1 1
 o  ;   o  1
R QC2 Q 3 R o1 C1 R o C2
4

C1  C2  1
5
R K   2.5 R o1  R o 2  1
2
RQ  5

5 1 1
H1( jo )  1  10
2.5  1 25

TO VERIFY

2
H3 ( jo )   10
1
5
11
We want to make
Hi ( jo )  H3 ( jo )
1
Vo1  Vo1
5
C1  5C1
1 1 R
& R11  & 1 R '
SC1 SC15 5 R1  1
'
5
Before scaling
1 Vo3
Vo1  Vi ( ) ; Vo 2  Vo1
SR K C1 SR o1 C1
After
1 1 Vo3 1
Vo1  Vi ( ) ( ) ; Vo 2  Vo1 .
SR K C1 5 SR o1 C1 5

Thus we have modified Vo1 and Vo 2 without changing Vo3

This usually can be done when enough degrees of freedom exist.

12
 STATE-VARIABLE FILTER ARCHITECTURES

• Derived from state-variable techniques to solve differential

equations. The objective is to render an expression that
can be implemented using integrator building blocks
• Example
Vo ( s)  Ks
H ( s)   (1)
Vi ( s) s 2   o s   2
0
Q
Where o is the cut-off frequency, (o/Q) is the bandwidth, Q
is the quality (selectivity) factor.

Let us rewrite the above equation by multiplying

K
 X ( s)
Vo ( s ) s
 (2)
Vi ( s )   o 1  o2 
1    X ( s)
 Q s s 
13
Both numerator and denominator by X(s)/s2. Thus (2) can be split into

 o X ( s) X ( s)
X ( s )  Vi ( s )    o2 2
(3a)
Q s s
X ( s)
Vo ( s )   K (3b)
s

Observe that 1/s is an integral operator. Thus we implement (3) by

using integral building block.

14
Remember that each integrator has its time-constant . We
can associate more than one time-constant with each integrator.

 o
  o / Q 
K

o X(s) o
S
Vi   V02 (s)
I
K1 2
Vo(s)

Also observe that V02 correspond to a low-pass, Vo to a bandpass

and X(s) to a high pass. K1 could be 1 or any other suitable value

15
This two-integrator biquad consists of

– One lossless integrator (I2)

– One Lossy integrator (I1)

– Two negative closed loops: One determines the center

frequency, the other the bandwidth (or Q).

– One of two integrators must be positive, the other negative.

–The lossy integrator must have a negative feedback.

16
 Second-Order Filter Structures

-K V2 V3
1/s 1/s
V1 V4
o
I1 I2
-o/Q

-o

V4  K o
H ( s)  
V1 s 2   o s   2
o
Q
17
 Zero implementation by addition of outputs technique
Vo

1 z Z
Qz
-K o
S
1 1
V1
V2 s s

-o/Q

-o

Vo  ( s 2  (z / Q) s  z o ) K
H ( s)  
V1 s 2  (o / Q) s  o2 18
 Feed-Forward Zeros Implementation

z
K
Qz

KZ o
Vi S 1/s S 1/s S Vo

I1 I2

o/Qo o

Vo s 2  ( z / Qz ) s   z 0
H ( s)   K 2 Why H(s) is not correct?
V1 s  (o / Qo ) s  o2

19
Tow-Thomas Biquad

 K o1
 KQ

V1  K LP K  K o2 Vo 2
o  1/ s  
Vo1
 1/ s o

 K BPP

 K BPN Tow-Thomas Biquad

with Feedforward Zero
implementations

 K LP K K o2 / s 2  ( K o2 K BPP / s  K BPN / s)(1  K Q / s)

V o R ( s)
H 2 ( s)  
V1 ( s) K Q K K o1 K o2
1 
2
s s

(s  K Q )( K o2 K BPP  K BPN )  K LP K o2
H 2 ( s) 
s 2  K Q s  KK o1K o 2

Vo1 (s) s  (K Q  K LP K / K BPP )s  K BPN K o1K / K BPP

2
H1 (s)   (K BPP )
V1 (s) s  K Qs  KKo1K o 2
2

20
 Feedforward Tow-Thomas (TT) Biquad Circuit

R1 R3
C1 R8
2 3
C2
R4 6
 R7 4
A1  11 R2 5
+ A 
+2 A3
+
R6
1
R5

V1
Observe that the regular TT Biquad does not implement a
highpass output
21
 Description of the Parameters for the Tow-Thomas Filter
General Transfer Function  1 1 R6  R6 1
s 2     
R
Ts    8  R 1C9 R 4 C9 R 7  R 7 R 3R 5C9 C10
R6  1  R8 1
s 2  s  
 R 1C9  R 7 R 3R 2 C9C10
where
R8 R6
2p  , 2z 
R 7 R 2 R 3C9 C10 R 3R 5 R 7 C9 C10
R 8C9 R 6 C9  1 R6 
Q p  R1 , Qz    
R 2 R 3R 7 C10 R 3R 5 R 7 C10  1
R R 4 7 
R

and
R8 R 4R 7
H HP  , for R1  , R5  
R6 R6
R 1R 8
H BP  , for R5, R6  
R 4R 7
R2
H LP  , for R4, R6  
R5
For the bandstop (notch)
R8 R 4R 7 R R
H notch  , for R1  , R5  6 2
R6 R6 R8
22
 Design Equations for the Tow-Thomas Filter

Let
R3  R
R 2  a 2R 3
R7  R8  R'
C1  C 2  C
1 1 R6
p  , 
aRC C R 5 RR '
R1 R6  1 R 
Qp  , Qz    6 ' 
aR R 5 RR '  R1 R 4 R 

R' R 4R '
H HP  , for R1   R 4 H HP , R 5  
R6 R6
R1
H HP  , for R5,R6  
R4
a 2R
H LP  , for R4,R6  
R5
R' R 4R ' a 2 RR 6
H notch  , for R1  , R5 
R6 R6 R'
23
 PSPICE Input file of Tow Thomas Filter

Tow - Thomas Biquad

** Description of the passive components
r1 2 3 1596698
r2 11 5 100000
r3 6 2 100000
r4 2 1 1596698
r7 3 4 100000
r8 4 11 100000
c1 2 3 9.7491D-11
c2 5 6 9.7491D-11
* Description of Op Amps
E1 3 0 0 2 2D5
E2 11 0 0 4 2D5
E3 6 0 0 5 2D5
*
VIN 1 0 AC 1
*
.AC LIN 100 6000 20000
.PLOT AC VDB(11) VP(11)
.PROBE
.END

24
Tow - Thomas Biquad
** Description of the passive components
r1 2 3 1596698
r2 11 5 100000
r3 6 2 100000
r4 2 1 1596698
r7 3 4 100000
r8 4 11 100000
c1 2 3 9.7491D-11
c2 5 6 9.7491D-11
* Description of Op Amps
E1 3 0 0 2 2D5
E2 11 0 0 4 2D5
E3 6 0 0 5 2D5
*
VIN 1 0 AC 1
*
.AC LIN 100 6000 20000
.PLOT AC VDB(11) VP(11)
.PROBE
.END

25
 RX R03

R3 RQ C2
C1

R01
Vi
R2 R02 
-+ -+ -+ VLP
Vi +- +- +- 
VLP
R2 R02
R3 C1
RQ
C2

R03
RX

KHN Fully-Differential Version

26
 Kerwin-Huelsman-Newcomb Biquad Circuit

R6

C7
2 3
C8
R5 6
 R1 4
12 A1  11 R2 5
V1 + V2 A2 
1 + A3
R3 R4 V3 + V4
R9

27
 Simple Design Equations for the KHN Filter

Let
2
R4  R5  b R6  R
R1  R2  R '
C7  C8  C

1
p 
bR ' C
1  R / R3  R / R9
Qp
1  1/ b2 b
1  1/ b2
| H HP | 
1  R3 / R  R 3 / R9
| H BP |  R ,
R3

| H LP |  R
R5 ,

28
 KHN Biquad Design Procedure

A simple design procedure is described as follows:

1. Assume that wp, Qp, and H=K are the design specifications.

2. Select convenient values* or R’, C, and R to determine the

values of R1, R2, C7, C8, R4, and R6.

3. Calculate the following element values:

1
b
R' C p
For the HP case:
R
R3 
bQ p H HP
R9 
   
R
bQ p 1 1 / b 2  H HP 1

* A rule of thumb for choosing R’ and R is to make them proportional to 10fp , when
Qp > 10, or else make R’ and R proportional to fp. Then C should be made
proportional to 1/R’ p .
29
 For the BP case:

R
R3 
H BP
R
R9 
  
bQ p 1  1 / b 2  1  H BP 
For the LP case:

R3  Q bR
Hp LP

bR
R9 
1  b2 Q p  b  Q p H LP

30
Example Design a bandpass modified KHN filter having a gain
of HBP =3, QP = 20, and o = 2p x 103r/s.

Procedure

1. Let us choose R=10 kW, C=0/01 mF, and R’=11.254 kW; that is
R1= R2 = 11.254kW and R4 = R5 = 10kW.

2. Since the values of R1 and R2 are 11.254kW, then

b 1  1.414207
R C p
'

31
3. The expression for R6  R; then R6  5kW
b2

4. For this case

R
R3   3.33 kW
H BP

and

R
R9 
   1  H
bQ p 1  1 / b 2
BP
 260 W

Exercise.- Propose component value condition to make this structure Fully

Balance Fully Symmetric Structure
32
KERWIN-HUELSMAN-NEWCOMB Biquad Circuit with f0=1KHz, Q=20, Peak
value gain =3
** Description of the passive components
r1 4 3 11.254K
r2 11 5 11.254K
r3 1 12 3.3K
* varying r4 values and parameters with the .step statement
.param R=1
r4 1 11 {R}
r5 2 6 10K
r6 2 3 5K
r9 1 0 260
c7 4 11 0.01U
c8 5 6 0.01U
* Description of Op Amps
E1 3 0 1 2 2D5
E2 11 0 0 4 2D5
E3 6 0 0 5 2D5
*
VIN 12 0 AC 1
*
.AC LIN 100 500 2K
.step DEC param R 300 30K 10
.PLOT AC VDB(11) VP(11)
.PROBE
.END 33
34
35
622 Active Filters by Edgar Sánchez-Sinencio

Second-Order Filter Characteristics

and Mason Rule to obtain Transfer Functions
• Second-Order Transfer function characteristics.
• How can you use this information for better design and tuning
• Time domain measurements and characterization
• Mason Rule is a good tool, borrow from control theory, to easily obtain
transfer functions from a signal flow graph

Analog and Mixed-Signal Center.

Texas A&M University
Edgar Sánchez-Sinencio
 PROPERTIES of SECOND-ORDER SYSTEMS
AND MASON’S RULE to determine transfer functions by
inspection

• Mathematical definitions and properties of Second-Order

Systems.
• Building block second-order system architectures and
properties.
• Mason’s Rule to obtain easily transfer functions and
to facilitate the generation of new architectures.
 SECOND-ORDER FILTER TYPES
Second-order blocks are important building blocks since with a combination of them
allows the implementation of higher-order filters. The general order transfer function in
the s-plane has the form:

K1s 2  K 2s  K 3
H(s) 
 s
s 2  o  2p
Q
Particular conventional cases are:

Lowpass i.e., K1  K 2  0
Bandpass i.e., K1  K 3  0
Highpass i.e., K 2  K3  0
(Notch) Band-Elimination i.e., K2  0
o
Allpass i.e., K1  1, K 2   and K 3  o2
Q
One interesting case used for amplitude equalization is the “equalizer” sometimes referred

to as Bump (DIP) Equalizer. In this case, K1  1 K 3  o2 and K 2   k o .
Q

Specific structures have different properties. Some structures have enough degrees of
freedom to allow them to change independentlyo , Q (or BW) and a particular gain H ( p )
wherep is a particular frequency, i.e., p  0, o ,  for the LP, BP and HP cases.
Furthermore, some structures have the property to have constant Q or BW while varying
f o. We will illustrate later, by examples, some of the structures with such properties.
Properties of Second-Order Systems in the time domain
o
s2  s  o2  (s  ) 2   2  s 2  2s  o2
Q
 1
where   o ,   o 1 
2Q 4Q 2

vin N(s) vos ( t )  K1  K 2et sin (t  )


1 (s  )   2
1
t1 
t A 
How to determine the pole location from a
A
step response?
e

K1, K 2  f ( N(s)) 2pN

t 2 

N CYCLES
o t 2
 1 Sinusoidal steady-state response

N(s)
sin t M() sin[ t  ()]
t

s 2  s o  o2
Q
M() o
BW 
M max Q
M max  
Q  10   N ( j )
N ( s)
2  
M( )   

s  s o    2
2 2   
 Q
o
 s   ( o2  2 ) 2   o  
j Q 

o

Only two measurements are necessary to fix the position of the complex poles. The
measurement of the frequency of peaking determines the magnitude of the poles,o ,
and the measurement of the 3-dB bandwidth determineso / Q.
 Second-Order Low-Pass Networks
H
T(s) 

s 2  s o  o2
Q
Since H is merely a magnitude scale factor, let H  o2 .
o2 1
T(s)  
o 2
s s
2
 o    s  1
2  s 
Q        Q 1
 o  o

1. For  / o  1, T( j)  1. Therefore, low frequencies are passed.

2. For  / o  1, T( j)  (o / )2 . Therefore, high frequencies are attenuated.
 1
3. For Q  1/ 2, the magnitude peaks at   1  2
o 2Q
The peak occurs at a frequency lower than o . For Q>5, the frequency of
peaking practically equals o(within 1%).
4. For Q>5, the 3-dB bandwidth is practically equal to o / Q rad / s.
5. At  / o  1, T( jo )  Q and the phase is  p / 2.
6. At  / o  1, the phase is  p.
7. For Q>5, the phase undergoes a rapid shift of p radians about o .
 10

8
Q  10
H1( w )

H2( w )
6
MagnitudeH3( w )

H4( w ) 5
H5( w )
4
H6( w )

H7( w )
3
2
2
1

0
0 0.5 1 1.5 2
o

w

0
1 o
0.5
2
3
arg( H1( w ) ) 1
5
arg( H2( w ) )
p
arg( H3( w ) ) 1.5

Phase
arg( H4( w ) ) 2
arg( H5( w ) )
2
arg( H6( w ) )
Magnitude and phase of
arg( H7( w ) )

o2
2.5


p s 2  s o  o2
3

3.5
Q
0 0.5 1 1.5 2
w
 The Second-Order Band-Pass Function

The second-order band-pass function has one zero at the origin and another
at infinity: T(s) 
Hs
o
s2  s  o2
Q
To normalize the peak value of the magnitude function to unity, let H  (o / Q) :
o 1 s
s
Q Q o
T(s)  
 2
j.
w
s 2  s o  o2  s   s  1
       Q 1
Q
H7( w ) Q
 o  o
w
j.
2
w 1
Q

Q  0.5
1
2

Q  10 o
M 1.5
1
a
0.8
2
1 P 1
3
g
H1( w ) arg( H1( w ) )

h Q  0.5 5
H2( w ) arg( H2( w ) )

n
H3( w )
0.6

2
arg( H3( w ) )
0.5

Magnitude
i
H4( w )
a
arg( H4( w ) ) 0

t
H5( w )

H6( w )
0.4 3 s
arg( H5( w ) )

arg( H6( w ) ) 0.5

and phase of
u
H7( w )
5 e
arg( H7( w ) ) o
d 1 s
0.2
Q
e Q  10 1.5

 2
s 2  s o  o2
0
0 0.5 1
w
1.5 2
o
0 0.5 1
w
1.5 2
Q
 Pole locations and properties

2p
1 Im N
 and t 2
t1 Constant BW

Im Im
 Re
j Constant Q
o

 o Re Re
 
2Q
Im
Constant o

o
Re

POLE-ZERO LOCI
In practical implementations besides the general specifications
(o , Q, H(p ) ) other particular specifications are imposed which
are application dependent. Among them are silicon area, dynamic
range, power supply rejection ratio, power consumption, tolerance,
accuracy, and sensitivity. This last parameter is often used as a
figure of merit. i.e.,
p x
p
Sx  
x p
The above definition is usually referred as normalized sensitivity (1)
due to the (x/p) factor. x and p are the variable of the network
(i.e. R, C, ) and the variable under consideration g m

(i.e., o , Q, H(p ) )
 The General Input-Output Gain Mason Formula
We can reduce complicated block diagrams to canonical form, from which the control ratio is easily written:
Vout G

Vin 1  GH
It is possible to simplify signal flow graphs in a manner similar to that of block diagram reduction. But it is also possible,
and much less time-consuming, to write down the input-output relationship by inspection from the original signal flow graph.
This can be accomplished using the formula presented below. This formula can also be applied directly to block diagrams,
but the signal flow graph representation is easier to read - especially when the block diagram is very complicated.
Signal Flow Graphs
Let us denote the ratio of the input variable to the output variable by T. For linear feedback control systems, T=Vout/Vin.
For the general signal flow graph presented in preceding paragraphs Vout is the output and Vin is the input.
The general formula for and signal flow graph is
 Pi i
T i
where Pi = the ith forward path gain 
Pjk =
jth possible product of k non-touching loop gains
  1  (1) k 1   Pjk
k j
 1   Pj1   Pj2   Pj3    
j j j
= 1 - (sum of all loop gains) + (sum of all gain-products of 2 non-touching loops) - (sum of all gain-
products of 3 non-touching loops + …
i =  evaluated with all loops touching Pieliminated.
Two loops, paths, or a loop and a path are said to be non-touching if they have no nodes in common.
 is called the signal flow graph determinant or characteristic function, since  is0the system characteristic
equation.
Examples
Let us determine the control ratio Vout/Vin and the canonical block diagram of the feedback
control system shown below:
G3
Vin    
Vout
G1 G4   G2
 
H1

H2

The signal flow graph is

G3

Vin 1 G1G 4 G2 1 1 Vout

     

H1

 H2

There are two forward paths: P1  G1G 2G 4 , P2  G1G3G 4

There are three feedback loops:
P11  G1G 4H1, P21  G1G 2G 4H2 , P31  G1G3G 4H2
There are no non-touching loops, and all loops touch both forward paths; then
1  1, 2  1
Therefore the control ratio is
C P11  P2 2 G1G 2G 4  G1G3G 4
T  
R  1  G1G 4H1  G1G 2G 4H 2  G1G3G 4H 2
G1G 4 (G 2  G 3 )

1  G1G 4H1  G1G 2G 4H 2  G1G 3G 4H 2
From Equations (8.3) and (8.4), we have
G  G1G 4 (G 2  G3 ) and GH  G1G 4 (G3H2  G 2H2  H1)
GH (G 2  G 3 )H 2  H1
Therefore H 
G G 2  G3
The canonical block diagram is there fore given by
G
R  C
G1G 4 (G 2  G3 ) 

H
(G 2  G 3 )H 2  H1
G 2  G3
The negative summing point sign for the feedback loop is a result of using a positive sign in the GH
formula above.
Example
Draw a signal flow graph for the following resistance network in which v2 (0)  v3 (0)  0.
v 2 is the voltage across C1.
R1 R2
o
2  o
 
v1
i1 i2 v3
C1 C2
o  
o
The five variables are v1, v2 , v3 , i1, and i 2 ; and v1 is the input. The four independent
equations derived from Kirchoff’s voltage and current laws are
 1   1  1 t 1 t
i1    v1    v 2 , v 2   i1dt   i 2dt ,
 R1   R1  C1 0 C1 0
 1   1  1 t
i 2    v 2  
 R2 
 v3 ,
 R2 
v3 
C2 0
i 2dt
The signal flow graph can be drawn directly from these equations:
t
 1 / R1 ( 1 / C1 )  dt  1/ R 2
0
t
(1 / C1 )  dt
t
1 / R1 1/ R 2 (1 / C2 )  dt 1
 
0   0  
v1 i1 v2 i2 v3 v3
In Laplace transform notation, the signal flow graph is given by
 1 / R1  1 / sC1  1/ R 2

1 / R1 1 / sC1 1/ R 2 1 / sC2 1
     
V1 I1 V2 I2 V3 V3
ECEN 622 (ESS)
Example of use of Mason Rule in a 3rd Orders State-Variable (observable) Filter

Vi
bo b1 b2 b3

S 1/s S 1/s S 1/s S Vo

-a2
-a1

-ao

bo b1 b2
 2  b3
3 b3s3  b2s2  b1s  bo
H1 s   s s s  3
a 2 a1 a o
1  2  3 s  a 2 s 2
 a1s  a o
s s s

What is wrong with this application of Mason’s Rule?

51
Vi
bo b1 b2 b3

B1 B2
S 1/s S 1/s S 1/s S Vo

-a2
-a1

-ao

bo B1B2 b1B1B2 b2 B2  a 2   a 2 B1a1 

   1   3 1   2 
 b
3 2 s  s   s 
H 2 s   s s s
a aB BBa
1  2  1 2 1  1 32 o
s s s
b3s3  b32 B2  b3a 2  b1B1B2 s2  b2 B2a 2  b3B1a1 s  B1B2 bo
H2 
s3  a 2s2  a1B1s  B1B2a o
52
 A possible implementation of H2 (s) using Active-RC follows

C/b3

Rb3

Rb1
C C C

Rbo
Vin RB1
RB2
Vo
Ra2

Ra1

Rao -1

53