Ch05 Higher Order Networks PDF

Chapter 5
Higher-Order Networks
Artemio P. Magabo
Professor of Electrical Engineering
Electrical and Electronics Engineering Institute

University of the Philippines - Diliman
Revised by Luis G. Sison, Jan, 2004
Revised by Michael Pedrasa, May 2012
Revised by Jhoanna Pedrasa, July 2005
Second-Order Transients
Consider the homogeneous differential equation
d2x
dx
a 2 +b
+ cx = 0
dt
dt
dx
with initial conditions x(0)=X0 and
(0)=X 0.
xt
The solution can be shown to be an exponential of
the form
x = K
st
where K and s are constants. Substitution gives

2
as K
st
+ bsK
st
+ cK
Department of Electrical and Electronics Engineering
st
=0
p2
After canceling the exponential term, we get the

characteristic equation
2
as + bs + c = 0
Using the quadratic formula, we get the two roots
2
- b b - 4ac
s1, s2 =
2a
Assuming the roots are real and distinct, the
solution will consist of two exponentials. Thus
x(t) = K1
s1t
+ K2
s2 t
dx
K1 and K2 can be evaluated using x(0) and
(0).
dt
p3
Source-Free Series RLC Network

Consider the circuit
shown. From KVL, we
get for t 0
di
1
L + Ri + idt = 0
dt
C
t=0
E
+
-
L
+
vC
Differentiating, we get
2
di
di 1
L 2 +R
+ i=0
dt C
dt
This is a homogeneous second-order differential
equation.
p4
The characteristic equation is
1
Ls + Rs + = 0
C
R
1
2
s + s+
=0
L
LC
2
or
From the quadratic formula, we get the two roots

2
R
1
R
s1 , s2 =

2L
LC
2L
Note: There are three types of root depending on
the value of the term inside the square root sign.
p5
1. Overdamped Case: The roots are real and

distinct when
2
1
R
>
LC
2L
The solution is the sum of two exponential terms
x(t) = K1
s1t
+ K2
s2 t
2. Critically Damped Case: The roots are real but

repeated when
2
1
R
=
LC
2L
p6
The solution can be shown to be
x(t) = (K1t + K2 )
st
3. Underdamped Case: The roots are complex

conjugates when
2
1
R
<
LC
2L
If the roots are
be shown to be
x(t) =
- t
s1, s2 = - jd, the solution can

(K1 cos dt + K2 sin dt)
p7
Comparison of Responses
overdamped
critically damped
underdamped
e-t
underdamped envelope response
p8
Example: The circuit has reached steady-state

when the switch is moved at t=0. Find
di +
t=0
R
1H
i(0+) and
(0 ).
dt
From KVL, we get

for t 0,
12V
+
-
vC
-
di
L
+ Ri + v c = 0
dt
Since the circuit has reached steady-state at t=0, i
(0+)=0 and vC(0+)=12V. Substitution gives
vc ( 0+ )
di +
(0 ) =
= 12 A/s
dt
L
p9
Example: The circuit

has reached steady+
state. At t=0, the
12V
switch is moved.
Find i(t) for t 0.
t=0 10
1H
+
vC
-
1
F
16
From KVL, we get for t 0,
di
1 + 10i + 16 idt = 0
dt
Differentiating the
equation, we have
di
di
+ 10 + 16i = 0
2
dt
dt
The characteristic equation is thus

2
s + 10s + 16 = 0
whose roots are s1=-2 and s2=-8.
p10
The roots are s1=-2 and s2=-8. Thus, we get

and
i(t) = K1 - 2t + K2 - 8t
di
= 2K1 - 2t 8K2 - 8t
dt
From the previous example, we ve found that

at t=0+,
di +
+
i(0 )=0 and
(0 ) = 12 A/s
dt
Substitution gives
i(0 ) = 0 = K1 + K2
di
(0+ ) = 12 = -2K1 - 8K 2
dt
+
p11
Solving simultaneously, we get K1=-2 and K2=2.

Thus,
i(t) = -2 -2t + 2 -8t Amp t 0
Note: For an over-damped case, the solution

consists of two distinct exponential terms.
p12

has reached steady+
state. At t=0, the 12V
switch is moved.
Find i(t) for t 0.
t=0
1H
+
vC
-
1
F
16
The characteristic equation can be shown to be

2
s + 8s + 16 = 0
whose roots are s1=-4 and s2=-4. Thus, we get
i(t) = K1 - 4t + K2 - 4t = K3 - 4t
A single exponential solution will not work since the
original differential equation is second-order.
p13
Assume i(t) = y(t) -4t. Differentiating twice, we get
di
- 4t
- 4t
= -4y(t) + y' (t)
dt
d2i
- 4t
- 4t
- 4t
= 16 y(t) - 8 y' (t) + y' ' (t)
2
dt
The original differential equation is
d2i
di
+8
+ 16i = 0
2
dt
dt
Substitution gives
0 = 16y(t)
-4t
- 32y(t)
- 8y' (t)
- 4t
-4t
-4t
+ y' ' (t)

-4t
+ 8y' (t)
-4t
+ 16y(t)
p14
Simplifying, we get
or
y' ' (t)
- 4t
=0
y' ' (t) = 0
Integrating twice, we get

or
y' (t) = K1
y(t) = K1t + K2
Finally, the solution is

or
i(t) = y(t)
- 4t
i(t) = K1t -4t + K2 -4t
p15
i(t) = K1t -4t + K2 -4t

Differentiating the solution, we get
di
= -4K1t - 4t + K1 - 4t - 4K2 - 4t
dt
The initial conditions are
A/sec. Substitution gives
i(0+)=0
di +
and
(0 )=-12
dt
i(0+ ) = 0 = 0 + K2
di +
(0 ) = -12 = 0 + K1 - 4K2
dt
We get K1=-12 and K2=0. Thus
i ( t ) = -12t
-4t
Amp
t0
p16
t=0

has reached steady+
state. At t=0, the
12V
switch is moved.
Find i(t) for t 0.
1H
+
1
F
16
vC
The characteristic equation can be shown to be

2
s + 6s + 16 = 0
whose roots are s1, s2=-3 j2.65. Thus, we get
or
i(t) = K1
i(t) =
(-3 + j2.65)t
-3t
(K1
+ K2
j2.65 t
(-3 - j2.65)t
+ K2
- j2.65 t
)
p17
Euler s Identities:
(1)
(2)
jx = cos x + j sin x
jx = cos x - j sin x
To prove the first identity, let y=cos x + j sin x.

dy
= -sin x + j cos x
dx
and since
as
j = 1, the equation can be re-written
dy
= j2 sin x + j cos x
dx
= j (cos x + j sin x)
p18
We get
or
dy
= jy
dx
1
dy = j dx
y
Integrating both sides, we get
ln y = j x + K
Evaluate K. When x=0, y=1.
ln 1 = j0 + K
or
Thus we get ln y = jx, or
K=0
jx = cos x + j sin x
Note: The other Euler s identity can be verified

following the same analysis.
p19
Back to the expression for the current
i(t) = -3t (K1 j2.65 t + K2 - j2.65 t )

From Euler s identities, we get
i(t) =
3t
[ K1 cos t + jK1 sin t
+ K2 cos t jK2 sin t]

where = 2.65. Combining the two cosine terms
and the two sine terms, we get
i(t) = 3t [ K3 cos t + K 4 sin t ]

where K3 = K1+K2 and K4 = j(K1-K2).
p20
Differentiate to get
di
3t
= [K3 sin t + K 4 cos t ]
dt
3 3t [ K3 cos t + K 4 sin t ]
The initial conditions are

A/sec. Substitution gives
i(0+)=0
di +
and
(0 )=-12
dt
i(0 ) = 0 = K 3
+
di +
(0 ) = 12 = K 4
dt
or K 4 = 4.54
We get
i(t) = 4.54 3t sin 2.65 t

Amp t 0
p21
Plot of the Currents

i(t)
R=10
R=8
R=6
Overdamped
Critically damped
Underdamped
time
p22
Source-Free Parallel RLC Network

Consider the circuit
shown. From KCL,
Iu(-t)
we get for t 0
dv 1
1
C
+ v + vdt = 0
dt R
L
v L
2
d v 1 dv 1
C 2 +
+ v=0
R dt L
dt
This is a homogeneous second-order differential
equation.
p23
1
1
Cs + s + = 0
R
L
1
1
2
s +
s+
=0
RC
LC
2
or
From the quadratic formula, we get the two roots

2
1
1
1
s1, s2 =

2RC
2RC LC
Note: We get three types of root depending on the
value of the term inside the square root sign.
p24
Example: Find v(t)

for t 0. Assume
u(-t)
the circuit has
reached steadystate at t=0.
First we need to find v(0+) and

At t=0-, the circuit
has reached steadystate.
1
H
8
1A
2F
dv + .
(0 )
dt
1
iL(0-)
vC(0-)
-
Since the inductor is shorted, vC(0-) = 0 V and iL

(0-) = 1A. Thus, v(0+) = 0 V.
p25
The equivalent circuit at t > 0 is

1
iL
1
H
8
From KCL we get,
v
-
dv
2
+ 8v + 8 vdt = 0
dt
2F
We have found that v(0+) = 0 V and iL(0+) = 1A.

dv + = -0.5 V/s.
Substitution gives
(0 )
dt
Next, we solve the differential equation describing
the circuit. Differentiating our KCL equation we
get
d2v
dv
dt
+8
dt
+ 8v = 0
p26
The characteristic equation is 2s2 + 8s + 8 = 0

whose roots are s1 = s2 = -2.
We have shown that for repeated roots, the
solution is of the form
v(t) = K1-2t + K2t-2t
and
dv
= -2K1 -2t + K2 -2t - 2K2t -2t
dt
Evaluating the solution and its derivate at t=0+ and

substituting the initial conditions v(0+)=0 V and
dv +
(0 ) = - 0.5 V/s we get K1 = 0 and K2 = -0.5.
dt
Thus,
v(t) = -0.5t-2t
t>0
p27
End
Higher-Order Transients
Consider the homogeneous differential equation
dnx
dn1x
dx
an n + an1 n1 + ... + a1
+ a0x = 0
dt
dt
dt
dx
(0)=X
with initial conditions x(0)=X0,
xt
n 1
2
x
dx
'''''''
'' , d
(
0
)
=
X
(0) = X0
0
n 1
2
dx
dt
The solution can be shown to be an exponential of

the form
st
x = K
where K and s are constants.

p29
Substitution gives
ansnK st + an 1sn 1K st + ... + a1sK st + a0K st = 0

After canceling the exponential term, we get the
characteristic equation.
ansn + an 1sn 1 + ... + a1s + a0 = 0

This is a polynomial of nth order and there will be n
roots. The type of response will depend on the
values of these roots. Assuming all the n roots are
real and distinct, the solution can be shown to be
x = K1
s1t
+ K2
s2 t
+ ... + Kn 1
sn 1t
+ K n
snt
p30
Example: Consider the differential equation
d3v
d2v
dv
+ 14 2 + 56
+ 64v = 0
3
dt
dt
dt
dv
with initial conditions v(0)=7 volts,
(0)=-24 v/s
dt
d2v
2. Find v(t).
and
(0)=112
v/s
dt 2
s3 + 14s2 + 56s + 64 = 0
The roots of the characteristic equation can be
shown to be s1=-2, s2=-4 and s3=-8.
p31
Since the roots are real and distinct, the solution is
v(t) = K1
-2t
+ K2
-4t
+ K 3
-8t
Differentiating twice, we get
dv
= 2K1-2t 4K2- 4t 8K3-8t
dt
2
dv
-2t
- 4t
-8t
= 4K1 + 16K2 + 64K3
2
dt
2
dv
d
v at t=0
Evaluate the expressions for v,
and
2
dt
dt
and use the initial conditions.
p32
We get
v(0) = 7 = K1 + K2 + K3
dv
(0) = 24 = 2K1 4K 2 8K3
dt
d2v
(0) = 112 = 4K1 + 16K2 + 64K3
2
dt
Solving simultaneously, we get K1=4, K2=2 and
K3=1. The final solution is
v(t) = 4
-2t
+ 2
-4t
+ 1
-8t
V
p33
d3i
d2i
di
+ 10 2 + 32
+ 32i = 0
3
dt
dt
dt
3
s + 10s + 32s + 32 = 0
The roots of the characteristic equation can be
shown to be s1=-2, s2=-4 and s3=-4. The solution
is
i(t) = K1-2t + K2t-4t + K3-4t
The constants K1, K2 and K3 can be evaluated if the

values of i, di/dt and d2i/dt2 are known at t=0.
p34
d5i
d4i
d3i
d2i
di
+ 6 4 + 17 3 + 28 2 + 24
+ 8i = 0
5
dt
dt
dt
dt
dt
5
s + 6s + 17s + 28s + 24s + 8 = 0

The roots of the characteristic equation are
s1 = -1, s2 = - 1, s3 = -2 and s4 , s5 = -1 j 3
The solution is of the form
i(t) = K1-t + K2t-t + K3-2t + -t (K4sin 3t + K5cos 3t )

p35
Getting the Differential Equation

1. Using nodal analysis or loop analysis, write the
KCL or KVL equations that describe the circuit.
2. Differentiate the equations, if necessary, to
eliminate any integral expressions.
3. In every equation, replace the derivatives with
operators.
4. Eliminate all variables, except one, using any
appropriate method.
5. Simplify as necessary and replace the operators
with the corresponding derivative terms.
p36
Example: Find the

differential equations
+
that describe the
v(t)
mesh currents i1 and
i2 in the network
shown.
8
i1
1H
1
F i2
16
10
First, write the mesh equations for the circuit.

Mesh 1:
Mesh 2:
8i1 + 16 (i1 i2 )dt = v(t)
t
di2
+ 10i2 + 16 (i2 i1 )dt = 0
dt
Then, differentiate the mesh equations to eliminate

the integrals.
p37
We get
di1
d
8
+ 16i1 16i2 =
v(t)
dt
dt
d2i2
di2
16i1 + 2 + 10
+ 16i2 = 0
dt
dt
(a)
(b)
d
Using operators, let D= . Substitution gives
dt
(1)
(8D + 16) i1 16i2 = Dv(t)
16i1 + (D2 + 10D + 16)i2 = 0
(2)
Next, multiply equation (1) by 16 and equation (2)

by (8D+16), then add the resulting equations. This
will eliminate the current variable i1.
p38
16
(8D+16)
(8D + 16) i1 16i2 = Dv(t)

2
16i1 + (D + 10D + 16)i2 = 0
(1)
(2)
We get
(8D3 + 96D2 + 288D) i2 = 16Dv(t)

which simplifies to
(D2 + 12D + 36) i2 = 2v(t)

The differential equation for the current i2 is
d2i2
di2
+ 12
+ 36 i2 = 2v(t)
2
dt
dt
p39
Similarly, if we multiply equation (2) by 16 and

equation (1) by (D2+10D+16), then add the
resulting equations, we will eliminate the current
variable i2. We get
3
(8D + 96D + 288D) i1 = D(D + 10D + 16)v(t)

which simplifies to
2
(D + 12D + 36) i1 = (D + 10D + 16)v(t)

1
8
The differential equation for current i1 is
d2i1
di1
d2v(t)
+ 12
+ 36 i1 = 0.125
2
dt
dt
dt 2
dv(t)
+ 1.25
+ 2v(t)
dt
p40
Alternative Procedure: First, solve for i1 from (b)

and differentiate the resulting equation. We get
and
i1 =
1
16
di1
=
dt
1
16
d2i2
+
2
dt
d3i2
+
3
dt
5
8
5
8
di2
+ i2
dt
d2i2 di2
+
2
dt
dt
Next, substitute the equations in (a). We get
d2i2
di2
+ 12
+ 36 i2 = 2v(t)
2
dt
dt
A similar procedure, applied on equation (a), will
result in the differential equation for current i1.
p41
Example: Find the

+
that describe the node
vS(t)
voltages v1 and v2 in
the network shown.
8 v1
1H
1
F
16
v2
10
REF
First, write the node equations for the circuit.
Node 1:
1 (v
1
8
vS ) +
Node 2: (v2
1
16
dv1 t
+ (v1 v2 )dt = 0
dt
1 v = 0
v1 )dt + 10
2
Then, differentiate the node equations to eliminate

the integrals.
p42
We get
or
2
d
d
v1
1
1
(v1 vS ) + 16
+ v1 v2 = 0
8
2
dt
dt
dvS
d2v1
dv1
+2
+ 16v1 16v2 = 2
dt
dt
dt
v2 v1 +
and
1
10
dv2
=0
dt
(a)
(b)
Substituting the D operator,

2
(D + 2D + 16) v1 16v2 = 2DvS (1)
v1 + (0.1D + 1)v2 = 0
(2)
p43
We can write v1 in terms of v2 from equation (2),
v1 = (0.1D + 1)v2
(3)
Substituting equation (3) into equation (1),

2
(D + 2D + 16) (0.1D + 1)v2 16v2 = 2DvS

Distributing terms and simplifying, we get
(D2 + 12D + 36)v2 = 20vS
(4)
Thus, the differential equation for v2 is

2
d v2
dv2
+ 12
+ 36v 2 = 20v S (t)
dt
dt
p44
Re-write equation (4) to get the equation for v2,
20v S
v2 = 2
D + 12D + 36
Substituting into equation (3) and re-arranging
terms we get
(D2 + 12D + 36)v1 = 2DvS + 20vS
Thus, the differential equation for v1 is
dvS (t)
d2v1
dv1
+ 12
+ 36v1 = 2
+ 20v S (t)
2
dt
dt
dt
p45
The Characteristic Equation

sn + an-1sn-1 + a1s + a0 = 0
The roots of the characteristic equation give the
form of the solution to the non-homogenous
differential equation. This is generally
x t = K1
s1t
+ K 2
s2 t
+ ... + Kn
snt
and is known as the transient response. The

solution is a sum of n exponential terms, where
n is the order of the differential equation.
An electric circuit will have a single characteristic
equation for all current or voltage variables in the
circuit.
p46
Example: In our
previous example we
have found the
for mesh currents i1
and i2 and node
voltages v1 and v2.
8 v1
+
vS(t)
1H
1
F i2
16
i1
v2
10
REF
The differential equation for the mesh currents are
d2i1
di1
+ 12
+ 36 i1 =
2
dt
dt
2
d
vS (t)
1
+
8
2
dt
1
0.8
dvS (t)
+ 2vS (t)
dt
d2i2
di2
+ 12
+ 36 i2 = 2vS (t)
2
dt
dt
p47
The differential equation for the node voltages are
dvS (t)
d2v1
dv1
+ 12
+ 36v1 = 2
+ 20v S (t)
dt
dt
dt
d2v2
dv2
+ 12
+ 36v 2 = 20v S (t)
dt
dt
The characteristic equation for this circuit is
S2 + 12s + 36 = 0
with repeated roots s1 = s2 = -6.
Thus, the transient response of all currents and
voltages in this circuit is of the form
x(t) = K1-6t + K2t-6t
p48
Example: The switch is moved from b to a at t=0.

Find the differential equations that describe the
voltages v1 and v2 for t 0.
a
4 +v2
+
v(t)
From KCL, we get
for t 0
Node 1:
Node 2:
b t=0
+
-
+v1
6
1H
REF
v1 - v2
+ v1dt = 0
6
v2 - v(t) v2 v1 1 dv 2
+
+
=0
4
6
4 dt
1
F
4
(1)
(2)
p49
Differentiate (1) and re-write the equations. We get

the differential equations
1 dv1
1 dv2
+ v1
=0
6 dt
6 dt
1
1 dv 2
5
1
- v1 +
+
v2 = v(t)
6
4 dt
12
4
The equations can be simplified into
dv1
dv 2
+ 6v1
=0
(D + 6)v1 Dv2 = 0
dt
dt
dv 2
- 2v1 + 3
+ 5v2 = 3v(t)
dt
- 2v1 + (3D + 5)v2 = 3v(t)
p50
Using operators, we get
(D + 6)v1 Dv2 = 0
- 2v1 + (3D + 5)v2 = 3v(t)
(1)
(2)
Multiply equation (1) by (3D+5) and (2) by D, then

add the resulting equations. This will eliminate the
variable v2. We get
2
(D + 7D + 10)v1 = Dv(t)
which yields the differential equation for v1.
2
d v1
dv1
d
+7
+ 10v1 =
v(t)
2
dt
dt
dt
p51
Multiply equation (1) by 2 and (2) by (D+6), then

variable v1. We get
2
(D + 7D + 10) v2 = (D + 6)v(t)
or
d v2
dv 2
d
+7
+ 10v2 =
v(t) + 6v(t)
2
dt
dt
dt
Note: The characteristic equation for v1 and v2 is

the same.
p52
Example: At t=0, the

switch is opened. Find the
differential equations that
describe the currents i1 1
+
and i2 for t 0.
v(t)
For t 0, we get from KVL
t=0
3
i1
1
F i2
20
1H
di2
+ 20 (i2 i1 ) dt = 0
4i1 + 20 (i1 i2 ) dt = v(t) and
dt
Differentiating,
di1
dv(t)
4
+ 20i1 20i2 =
dt
dt
d2i2
+ 20i2 20i1 = 0
2
dt
p53
(4D + 20) i1 20i2 = Dv(t)

20i1 + (D2 + 20) i2 = 0
(1)
(2)
Multiply equation (1) by (D2+20) and (2) by 20,

then add the resulting equations. This will eliminate
the variable i2. We get
2
(D + 5D + 20) i1 = ( D + 5) v(t)
1
4
The differential equation for i1 is

2
d i1
di1
+5
+ 20 i1 =
2
dt
dt
1
4
d v(t)
+ 5 v(t)
2
dt
p54
Multiply equation (1) by 20 and (2) by (4D+20),

then add the resulting equations. This will eliminate
the variable i1. We get
2
(D + 5D + 20) i2 = 5v(t)
The differential equation for i2 is
2
d i2
di2
+5
+ 20 i2 = 5v(t)
2
dt
dt
p55
End
Why Get Initial Conditions?

Complete Response:
1. Steady-state Response
2. Transient Response
Transient Response: General form is exponential
x = K1 s1t + K2 s2t + ... + Kn 1 sn1t + Kn snt

where K1, K2, Kn are arbitrary constants.
Answer: The initial conditions are necessary in
the determination of the numerical values of the
arbitrary constants K1, K2, Kn.
p57
Evaluating Initial Conditions

1. Assume switching operation at t=0.
2. Evaluate the inductor currents and capacitor
voltages at t=0-.
3. Find inductor currents and capacitor voltages
at t=0+.
4. Write the KVL and KCL equations describing the
network for t 0.
5. Use the KVL and KCL equations for t 0 and
their derivatives, plus the inductor currents
and capacitor voltages at t=0+ to evaluate
the required initial conditions.
p58
Example: The circuit has reached steady-state

condition with the switch in position a. At t=0, the
switch is moved to position b. If the capacitor is
initially uncharged, find i(0+),
a
1k
2
di +
di +
(0 ) and 2 (0 ).
b t=0
dt
dt
100V
The circuit is at steady

state for t<0.
100
i L ,ss (0 ) =
= 0 .1 A
1k
v C (0 ) = 0
0.1F
1H
1k
100V
+
-
iL,ss
p59
From KVL, we get for t 0,
di
1
(1) L
+ Ri + idt = 0
dt
C
whose derivative is
0.1F
1k
i
1H
d2i
di 1
(2) L
+R
+ i=0
2
dt C
dt
At t=0+,
i(0+ ) = iL,ss = 0.1 A

v C (0 + ) = 0
p60
From (1), we get at t=0+
0
di +
+
+
L
(0 ) + Ri(0 ) + vC (0 ) = 0
dt
which gives
di +
R +
(0 ) = i(0 ) = 100 A/s
dt
L
From (2), we get
d2i +
1
di +
1
+
(
0
)
=
[
R
(
0
)
+
i
(
0
)]
2
L
dt
C
dt
= -900 kA/s2
p61
Example: The network is initially at steady-state

condition with the switch open. At t=0, the switch
is closed. Find i1(0+), i2(0+),
t=0
di1 +
di2 +
(0 ) and
(0 ) .
dt
dt
10
120V
At
t=0-,
we get
i1
20
1H
i2
20
1F
10
+
120V
20
iL(0-)
vC(0-)
-
120
iL (0 ) =
=4A
30
-
vC (0- ) = 20iL (0 ) = 80 V
p62
For t 0, we get from KVL
di1
(1)
+ 20i1 = 120
dt
6
(2) 20i2 + 10
120V
20
i1
i2dt = 120
1H
i2
20
1F
At t=0+, we get
i1(0+ ) = iL (0- ) = 4 A
v C (0 + ) = v C (0- ) = 80 V
From equation (2), we get
i2 (0+ ) =
1
20
[120 vC (0+ )] = 2 A
p63
di1
(1)
+ 20i1 = 120
dt
(2) 20i2 + 10
i2dt = 120
di1 +
(0 ) = 120 20i1(0+ ) = 40 A/s
dt
di2
To get an equation involving
, differentiate (2).
dt
We get
di2
20
+ 106 i2 = 0
dt
At t=0+, we get
di2 +
(0 ) = -100 kA/sec
dt
p64
Example: The network is initially unenergized. At

t=0, the switch is closed. Determine i1(0+), i2(0+),
di2 +
di1 +
t=0 C
R2
(0 ) and
(0 ).
dt
dt
+
-
i1
R1
i2
For t 0, we get from

KVL,
(1)
(2)
1
i1dt + R1(i1 i2 ) = E
C
di2
L
+ (R 1 + R 2 )i2 R 1i1 = 0
dt
p65
(1)
(2)
1
i1dt + R1(i1 i2 ) = E
C
di2
L
+ (R 1 + R 2 )i2 R 1i1 = 0
dt
Since the circuit is initially unenergized, we know

that vC(0+)=0 and iL(0+)=0. Thus
i2(0+ ) = 0
From (1), we get
vC (0+ ) + R1i1(0+ ) R1i2 (0+ ) = E

or
E
i1(0 ) =
R1
+
p66
1
i1dt + R1(i1 i2 ) = E
(1)
C
di2
(2) L
+ (R 1 + R 2 )i2 R 1i1 = 0
dt
E
+
+
i1(0 ) =
i2(0 ) = 0
R1
From (2), we get
di2 +
L
(0 ) + (R1 + R 2 )i2 (0+ ) R1i1(0+ ) = 0
dt
which gives
di2 +
R1
E
+
(0 ) =
i1(0 ) =
dt
L
L
p67
1
i1dt + R1(i1 i2 ) = E
(1)
C
di2
(2) L
+ (R 1 + R 2 )i2 R 1i1 = 0
dt
Differentiate equation (1). We get
1
di1
di2
i1 + R1
R1
=0
C
dt
dt
At t=0+, we get
1
di1 +
di2 +
+
i1(0 ) + R1
(0 ) R1
(0 ) = 0
C
dt
dt
or
di1 +
E
E
(0 ) = 2
dt
L R1 C
p68
Example: The network

has reached steadystate condition with the
switch open. At t=0, the
switch is closed. Find
the necessary initial
conditions for mesh
currents i1, i2, and i3.
Equivalent circuit at
-
R2=1k
i2
C1
i1 1H
i3
C2
vC1(0 ) + vC2 (0 ) = 18 V
C2=2C1=2F
R2
t=0-
iL (0 ) = 9 mA = iL (0 )
-
R1=2k
E=27V
+
-
R1
iL(0-)
vC1(0-)
vC2(0-)
-
p69
For t<0, C1 and C2 are in series. Thus, iC1=iC2.

Since
t
q=
idt
then qC1(0-)=qC2(0-). This means that at t=0or
C1vC1(0- ) = C2vC2(0- )
vC1(0- ) = 2vC2 (0- )
And we know that
vC1(0- ) + vC2 (0- ) = 18 V
Solving for the voltages, we get
vC1(0- ) = 12 V = vC1(0 + )
vC2 (0- ) = 6 V = vC2(0 + )
p70
R2=1k
Equivalent circuit for t 0

E=27V
C2=2C1=2F
E
R1=2k
i2
C1
i1 1H
i3
C2
From KVL, we get

(1)
(2)
(3)
1
1
27 = R 2i1 +
i2dt +
i3dt
C1
C2
1
0 = R1(i2 i1 ) +
i2dt
C1
d
1
0=L
(i3 i1 ) +
i3dt
dt
C2
p71
(1)
27 = R 2i1 +
1
C1
i2dt +
1
C1
(2)
0 = R1(i2 i1 ) +
(3)
d
1
0 =L
(i3 i1 ) +
dt
C2
1
C2
i dt
3
i dt
2
i dt
3
At t=0+, we get from (1)
1
+
+
i1(0 ) =
[27 vC1(0 ) vC2(0 )] = 9 mA
R2
+
Since i1(0+) - i3(0+)=iL(0+), then
i3 (0+ ) = i1(0+ ) iL (0+ ) = 0

p72
(1)
27 = R 2i1 +
1
C1
i2dt +
1
C1
(2)
0 = R1(i2 i1 ) +
(3)
d
1
0 =L
(i3 i1 ) +
dt
C2
1
C2
i dt
3
i dt
2
i dt
3
For resistor R1, we get
vR1(0+ ) = R1[i1(0+ ) i2 (0+ )] = vC1(0+ )

or
+
v
(
0
)
+
+
C1
i2 (0 ) = i1(0 )
= 3 mA
R1
p73
(1)
(2)
1
27 = R 2i1 +
C1
1
i2dt + C2 i3dt
1
0 = R1(i2 i1 ) +
i2dt
C1
Differentiate equations (1) and (2). We get

(4)
di1
1
1
0 = R2
+
i2 +
i3
dt C1
C2
(5)
di2
di1
1
0 = R1
R1
+
i2
dt
dt C1
di1 +
1 1
1
+
(0 ) =
[ i2(0 ) +
i3(0+ )]
dt
R 2 C1
C2
= 3 A / s
p74
(3)
(5)
d
1
(i3 i1 ) +
i3dt
dt
C2
di
di
1
0 = R1 2 R1 1 +
i2
dt
dt
C1
0 =L
di2 +
di1 +
1
(0 ) =
(0 )
i2 (0+ )
dt
dt
R1C1
= 4.5 A / s
di3 +
di1 +
1
(0 ) =
(0 ) v C2 (0 + )
dt
dt
L
= 9 A / s
p75
(4)
(5)
di1
1
1
+
i2 +
i3
dt C1
C2
di2
di1
1
0 = R1
R1
+
i2
dt
dt
C1
0 = R2
Differentiate equations (4) and (5) and evaluate at

t=0.
d2i1
1 di2
1 di3
0 = R2
+
+
(6)
2
dt
C1 dt
C2 dt
d2i2
d2i1
1 di2 (7)
0 = R1
R1
+
2
2
C1 dt
dt
dt
From (6)
d2i1 +
2
(0
)
=
9
kA/sec
dt2
From (7)
d2i2 +
2
(0
)
=
11.25
kA/sec
dt2
p76
d
1
0 =L
(i3 i1 ) +
dt
C2
(3)
i dt
3
Differentiate equation (3) and evaluate at t=0.
d2i3
d2i1
1
0 =L
L
+
i3
2
2
C2
dt
dt
or
d2i3 +
2
(0
)
=
9
kA/sec
dt2
Note: This is a third-order network so we

needed three initial conditions for each mesh
current,
2
di
d
iX +
+
X
iX(0+),
(0 ) and
(0 )
2
dt
dt
p77
Example: The network has reached steady-state

condition with the switch closed. At t=0, the switch
is opened. Find i(0+), v(0+),
t=0
dv +
di +
(0 ) and
(0 ).
dt
dt
60V
20
1H
10
10-3F
Equivalent circuit at t=0-
60
iL (0 ) =
=3A
20
vC (0 ) = 60 V
60V
20
iL(0-)
10
vC(0-)
-
p78

(1)
di
+ 20i = v
dt
3
(2) 10i + 10
At t=0+, we get
20
i
1H
idt = v
+
v
- C
-
10
10-3F
i(0+ ) = iL (0 ) = 3 A
vC (0+ ) = vC (0 ) = 60 V
10i(0 ) vC (0 ) = v(0 )
+
or
v(0 ) = 30 V
+
p79
(1)
di
+ 20i = v
dt
(2)
10i + 10
idt = v
From (1), we get
di +
(0 ) = v(0+ ) 20i(0+ ) = 30 A/s
dt
di
dv
3
10
+ 10 i =
dt
dt
At t=0+, we get
dv +
di +
(0 ) = 10 (0 ) 103 i(0 + ) = -2,700 V/s
dt
dt
p80
Example: The switch has been in position b for a

long time. At t=0, the switch is moved to a. Find
a
dv 1 +
4 +v2
+
+
v1(0 ), v2(0 ),
(0 )
dt
b t=0
dv 2 +
+
6
and
(0 ).
1
+
+v
120V
1
dt
F
60V
4
1H
REF
iL (0 ) = 6 A = iL (0+ )
60V
vC (0 ) = 36 V = vC (0 )
4
+
-
6
iL(0-)
vC(0-)
-
p81
4 +v2

From KCL, we get
120V
v2 v1
= v1dt
(1)
6
120 v2
v2 v1
(2)
=
+
4
6
+v1
1
4
dv 2
dt
6
1H
1
F
4
REF
At t=0+,
v2 (0+ ) = vC (0+ ) = 36 V
From (1), we get
v1(0+ ) = v2 (0+ ) 6iL (0+ ) = 0

p82
(1)
v2 v1
=
6
v dt
1
120 v2
v2 v1
=
+
(2)
4
6
1
4
dv 2
dt
From (2), we get
dv 2 +
(0 ) = 60 V/s
dt
At t=0+,
dv1
dv 2
=
6v1
dt
dt
dv1 +
dv 2 +
(0 ) =
(0 ) 6v1(0 + ) = 60 V/s
dt
dt
p83
End
The Steady-State Response

dnx
dn1x
dx
an n + an1 n1 + ... + a1
+ a0x = g(t)
dt
dt
dt
The steady-state response is the solution to the
non-homogenous differential equation. It is similar
in form to the forcing function g(t) plus all its
unique derivatives.
In practical electric circuits, the driving forces are
represented by a few mathematical forms, such as
a constant or a sinusoid. The method of
undetermined coefficients can be used to evaluate
the steady-state response.
p85
Method of Undetermined Coefficients

The method of undetermined coefficients is
applied by selecting trial functions of all possible
forms that might satisfy the differential equation.
Format of Forcing
Function g(t)
Trial Function
K (constant)
c (constant)
Ktm
cmtm + cm-1tm-1 + + c1t + c0
Kt
ct
K cos t
K sin t
c1 cos t + c2 sin t
p86

1. Find the form of the transient response by
solving for the roots of the characteristic
equation.
2. Write the trial form of the steady-state response
for each forcing function using the table. If any
term in the trial function appears in the
transient response, the trial function should be
multiplied by t. If the modified trial function still
has common terms with the transient response,
another t must be multiplied until no common
term exists.
p87

3. Differentiate the trial solution as many time as
needed and substitute into the differential
equation. By equating coefficients of like terms,
form a set of algebraic equations in the
undetermined coefficients.
4. Solve for the undetermined coefficients. These
coefficients must be in terms of circuit and
driving force parameters. There are no arbitrary
constants in the steady-state response.
p88
Example: Find the steady-state response of the

second order non-homogeneous differential
2
equation
dv
dv
+5
+ 4v = 2 sin 2t
2
dt
dt
The characteristic equation is s2 + 5s + 4 = 0 with

roots s1 = -4 and s2 = -1.
Thus, the transient response is of the form
vt(t) = K1-4t + K2-t
The steady-state response is of the same form as
the right hand side of the differential equation.
From the table, the trial solution for this forcing
function is
vss(t) = c1 cos 2t + c2 sin 2t
p89
vss(t) = c1 cos 2t + c2 sin 2t

Differentiating vss(t) twice,
dvss(t)/dt = -2c1 sin 2t + 2c2 cos 2t
d2vss(t)/dt2 = -4c1 cos 2t - 4c2 sin 2t
Our original differential equation is
d2v
dv
+5
+ 4v = 2 sin 2t
2
dt
dt
Substituting the trial solution and its derivatives
into the differential equation, we get
-4c1 cos 2t - 4c2 sin 2t + 5(-2c1 sin 2t + 2c2 cos 2t)
+ 4(c1 cos 2t + c2 sin 2t) = 2 sin 2t
p90
Distributing terms and simplifying,

10c2 cos 2t - 10c1 sin 2t = 2 sin 2t
Comparing the coefficients of the left-hand side of
the equation to the right hand side,
Coefficient of cos: 10c2 = 0
Coefficient of sin: -10c1 = 2
We get
c1 = -0.2
and
c2 = 0
The steady-state response is

vss(t) = -0.2 cos 2t
p91
Example: Consider the second order

non-homogeneous differential equation
d2i
di
2
+
3
+
2i
=
12t
dt
dt2
Find the steady-state response.
The characteristic equation is s2 + 3s + 2 = 0 with
roots s1 = -1 and s2 = -2.
The transient response is of the form
it(t) = K1-t + K2-2t
From the table, the trial solution for the steadystate response is
iss(t) = c1t2 + c2t + c3
p92
diSS
= 2c1t + c2
dt
and
d2iSS
= 2c1
2
dt

into the differential equation,
or
2c1 + 3(2c1t + c2) + 2(c1t2 + c2t + c3) = 12t2

2c1t2 + (6c1 + 2c2)t + 2c3+ 3c2+ 2c1 = 12t2
Comparing coefficients,
2c1 = 12
6c1 + 2c2 = 0
2c3+ 3c2+ 2c1 = 0
p93
We get that c1 = 6, c2 = -18 and c3 = 21.

Thus, the steady-state response is
iss(t) = 6t2 -18t + 21
Note: The complete response is
i(t) = it(t) + iss(t)
= K1-t + K2-2t + 6t2 -18t + 21
The arbitrary constants K1 and K2 can be evaluated
if the initial conditions i(0) and
di
(0) are known.
dt
p94

d2x
dx
+ 12
+ 36x = 18
2
dt
dt
s2 + 12s + 36 = 0
with roots s1 = -6 and s2 = -6.
xt(t) = K1-6t + K2t-6t
p95
From the table, the trial solution for the steadystate response is
xss(t) = c1
and
d2xss dxss
=
=0
2
dt
dt

into our differential equation we get
36c1 = 18
or c1 = 0.5
Thus, the steady-state response is
xSS = 0.5
Note: The complete solution is

x(t) = xt(t) + xss(t) = K1-6t + K2t-6t + 0.5
The arbitrary constants K1 and K2 can be evaluated
if the initial conditions are known.
p96

d2i
di
-4t
+
12
+
36i
=
18
dt
dt2
Formulate the complete response.
it(t) = K1-6t + K2t-6t
The trial solution for an exponential is
iSS(t) = c-4t
and
diss
= 4c-4t
dt
d2iss
-4t
=
16c
dt2
p97
Substituting into the differential equation,
16c
4t
-4t
+ 12(-4c
-4t
) + 36(c
-4t
) = 18
Canceling the exponential terms we get
4c = 18
or
c = 4.5
Thus the steady-state response is

iss(t) = 4.5-4t A
And the complete response is
i(t) = K1-6t + K2t-6t + 4.5-4t
where K1 and K2 can be evaluated using initial

conditions.
p98
Example: Consider the second order differential

equation
d2i
di
-4t
+
12
+
36i
=
18
+
18
dt
dt2
it(t) = K1-6t + K2t-6t
The right-hand side of the equation is the response
due to a constant source and an exponentially
decaying source. From the principle of
superposition, the trial solution is
iss(t) = K + c-4t
p99
diss
= 4c-4t
dt
and
d2iss
-4t
=
16c
dt2
Substituting into the differential equation,
16c 4t + 12(-4c-4t ) + 36(K + c-4t ) = 18 + 18-4t

Which simplifies into
36K + 4c-4t = 18 + 18-4t
Comparing coefficients, we get
36K = 18
or
K = 0.5
4c-4t = 18-4t or
c = 4.5
iss(t) = 0.5 + 4.5-4t A
p100
Example: Consider the second order differential

equation
d2i
di
-6t
+
12
+
36i
=
18
dt
dt2
Formulate the steady-state response.
it(t) = K1-6t + K2t-6t
The trial solution of the steady-state response is
iSS(t) = c-6t
This will not work as it is of the form as the first
term of the transient response.
p101
Multiplying the trial solution by t, we get

iSS(t) = ct-6t
This is still of the same form as one of the terms of
the transient response, so we multiply it by again
by t.
iSS(t) = ct2-6t
This is now the final form of the trial solution.
diss
= 2ct-6t 6ct 2-6t
dt
d2iss
-6t
-6t
2 -6t
=
2c
24ct
+
36ct
2
dt
p102
Substituting into the original differential equation,

36ct2-6t - 24ct-6t + 2c-6t +12(2ct-6t - 6ct2-6t)
+ 36ct2-6t = 18-6t
Canceling the exponential terms and simplifying,
we get
2c = 18
or
c=9
iss(t) = 9t2-6t

i(t) = K1-6t + K2t-6t + 9t2-6t
where K1 and K2 can be evaluated using initial

conditions.
p103
Example: Formulate the complete response of

the second-order differential equation
d2i
+ i = sin t
2
dt
The characteristic equation is s2 +1 = 0 with roots

s1 = +j and s2 = -j.
it(t) = K1 cos t + K2 sin t
The trial solution for the steady-state response is
and
iss(t) = c1t cos t + c2t sin t
d iss
= c1(-t cos t - 2sin t) + c2(-t sin t + 2cos t)
2
dt
p104

into the differential equation, we get
-2 c1 sin t + 2 c2 cos t = sin t
Comparing coefficients, we get c1=-0.5 and c2= 0.
The steady-state response is
iss(t) = -0.5t cos t
i(t) = K1 cos t + K2 sin t -0.5t cos t
p105
Solving the Differential Equation

Consider the nth-order differential equation
dnx
dn1x
dx
an n + an1 n1 + ... + a1
+ a0x = g(t)
dt
dt
dt
1. Find the transient response xt. This is generally
an exponential of the form
x t = K1 s1t + K2 s2t + ... + Kn snt

2. Find the steady-state response xss. This is
similar in form to the forcing function g(t) plus
all its unique derivatives.
p106
3. Evaluate the initial conditions. We need the
d2x
dn1x
dx
values of x(0),
(0),
(0),
(0).
2
n 1
xt
dt
dx
4. Find the total response. Add the steady-state

response and transient response.
x(t) = x ss + K1
s1t
+ K 2
s2 t
+ ... + Kn
snt
5. Differentiate the total response (n-1) times.

6. Using the expressions for x(t) and its (n-1)
derivatives in step 5, and the initial conditions in
step 3, find the arbitrary constants K1, K2, Kn.
p107

is initially unenergized.
At t=0, the switch is
closed. Find currents i1
(t) and i2(t) for t 0.
t=0 4 1H
+
-
24V
i1
4
i2
2H
First, get the differential equations that describe

currents i1(t) and i2 for t 0. The mesh equations
are
di1
+ 12i1 8i2 = 24
dt
di2
2
+ 12i2 8i1 = 0
dt
(1)
(2)
p108
Using operators we get

(1)
di1
+ 12i1 8i2 = 24
dt
(D+12)i1 8i2 = 24 (a)
(2)
di2
2
+ 12i2 8i1 = 0
dt
-8i1 + (2D+12)i2 = 0 (b)
To eliminate i2, multiply (a) by (2D+12) and (b)

by 8 and add the resulting equations. We get
(D2 + 18D + 40)i1 = (D+12)12
or
d2i1
di1
+ 18
+ 40i1 = 144
2
dt
dt
p109
Similarly, we eliminate i1 by multiplying (a) by 8

and (b) by (D+12) and adding the equations. We
get
(D2 + 18D + 40)i2 = 96
or
d i2
di2
+ 18
+ 40i2 = 96
2
dt
dt
Alternatively, we can solve for i2 in equation (1)

and differentiate the resulting equation. We get
and
di1 3
i2 =
+ 2 i1 3
dt
di2 1 d2i1 3 di1
= 8 2 +2
dt
dt
dt
1
8
(3)
(4)
p110
Substitute (3) and (4) in equation (2). We get
d2i1
di1
+ 18
+ 40i1 = 144
2
dt
dt
This is the required differential equation for i1.
Solve for i1 in equation (2) and differentiate the
resulting equation. We get
i1 =
1
4
di2 3
di1
(5)
+ 2 i2
and
=
dt
dt
1
4
d2i2
+
2
dt
3
2
di2
dt
(6)
d2i2
di2
+ 18
+ 40i2 = 96
2
dt
dt
This is the required differential equation for i2.

p111
Next, we find the transient response.

Setting the right-hand side of the differential
equations to zero, we get the characteristic
equation
s2 + 18s + 40 = 0
Solving for the roots, we get s1=-2.6 and

s2=-15.4.
Thus, we get the transient response to be of the
form
i1t = K1-2.6t + K2-15.4t

i2t = K3
-2.6t
+ K 4
-15.4t
p112
Next, we determine the steady-state response

i1,ss and i2,ss which are both constant. The
differential equation for i1 is
Since the
d2i1ss
dt2
or
di1ss
+ 18
+ 40i1ss = 144
dt
i1,ss
144
=
= 3.6 A
40
From the differential equation for i2

2
d i2
di2
+ 18
+ 40i2 = 96
2
dt
dt
we get
96
i2,ss =
= 2.4 A
40
forcing function
is a constant
(24V) the
steady-state
response of any
current or
voltage should
also be a
constant. Thus,
i1,ss = A
di1,ss/dt = 0
d2i1,ss/dt2 = 0
p113
Alternatively, We can also draw the equivalent

circuit at steady state. We get
4
4
12i1,ss 8i2,ss = 24
8i1,ss + 12i2,ss = 0
24V
i1,ss 8
i2,ss
which gives i1,ss= 3.6 Amps

and i2,ss= 2.4 Amps.
Next, we find the initial conditions. We need
di1 +
di2 +
i1(0+),
+
i2(0 ),
(0 ) and
(0 ).
dt
dt
p114
Since the circuit was initially unenergized, we get
i1(0+ ) = i2 (0+ ) = 0
We ve previously formulated the mesh equations
di1
+ 12i1 8i2 = 24
dt
(1)
di2
(2)
2
+ 12i2 8i1 = 0
dt
di1 +
From (1), we get
(0 ) = 24 A/s
dt
di2 +
From (2), we get
(0 ) = 0
dt
p115
Next, formulate the complete response. We get
i1( t ) = 3.6 + K 1 - 2.6t + K 2 -15.4t

i2 ( t ) = 2.4 + K 3 -2.6t + K 4 -15.4t
whose derivatives are
di1
-2.6t
-15.4t
= !2.6K1"
! 15.4K 2"
dt
di2
-2.6t
-15.4t
= !2.6K 3"
! 15.4K 4"
dt
p116
Evaluate the constants K1 and K2. At t=0+, we get
i1(0 ) = 0 = 3.6 + K1 + K2
+
di1 +
(0 ) = 24 = 2.6K1 15.4K2
dt
Solving simultaneously, we get K1=-2.46 and

K2=-1.14. The final expression for current i1 is
i1(t) = 3.6 2.46
-2.6t
1.14
-15.4t
Evaluate the constants K3 and K4. We get K3=-2.89

and K4=0.49. Thus
-2.6t
i2(t) = 2.4 2.89
-15.4t
+ 0.49
A
p117
Plot of the Currents
I2
I1
t=0
+
-
24
4
i1
4
8
i2
p118
End

v1(t) and v2(t) for t 0.
a
4 +v2
+
120V
-
b t=0
+
-
60V
+v1
6
1H
1
F
4
REF
In a previous example, we got the differential

equations that describe the voltages v1 and v2
(slide 49). In another example, we derived the
initial conditions (slide 81).
p120
Example: The switch is moved from b to a at t=0.

Find the differential equations that describe the
voltages v1 and v2 for t 0.
a
4 +v2
+
v(t)
From KCL, we get
for t 0
Node 1:
Node 2:
b t=0
+
-
+v1
6
1H
REF
v1 - v2
+ v1dt = 0
6
v2 - v(t) v2 v1 1 dv 2
+
+
=0
4
6
4 dt
1
F
4
(1)
(2)
p121
Differentiate (1) and re-write the equations. We get

the differential equations
1 dv1
1 dv2
+ v1
=0
6 dt
6 dt
1
1 dv 2
5
1
- v1 +
+
v2 = v(t)
6
4 dt
12
4
The equations can be simplified into
dv1
dv 2
+ 6v1
=0
(D + 6)v1 Dv2 = 0
dt
dt
dv 2
- 2v1 + 3
+ 5v2 = 3v(t)
dt
- 2v1 + (3D + 5)v2 = 3v(t)
p122
(D + 6)v1 Dv2 = 0
- 2v1 + (3D + 5)v2 = 3v(t)
(1)
(2)
Multiply equation (1) by (3D+5) and (2) by D, then

variable v2. We get
2
(D + 7D + 10)v1 = Dv(t)
which yields the differential equation for v1.
2
d v1
dv1
d
+7
+ 10v1 =
v(t)
2
dt
dt
dt
p123
Multiply equation (1) by 2 and (2) by (D+6), then

variable v1. We get
2
(D + 7D + 10) v2 = (D + 6)v(t)
or
d v2
dv 2
d
+7
+ 10v2 =
v(t) + 6v(t)
2
dt
dt
dt
Note: The characteristic equation for v1 and v2 is

the same.
p124

a
dv 1 +
4 +v2
+
+
v1(0 ), v2(0 ),
(0 )
dt
b t=0
dv 2 +
+
6
and
(0 ).
1
+
+v
120V
1
dt
F
60V
4
1H
REF
iL (0 ) = 6 A = iL (0+ )
60V
vC (0 ) = 36 V = vC (0 )
4
+
-
6
iL(0-)
vC(0-)
-
p125
4 +v2

From KCL, we get
120V
v2 v1
= v1dt
(1)
6
120 v2
v2 v1
(2)
=
+
4
6
+v1
1
4
dv 2
dt
6
1H
1
F
4
REF
At t=0+,
v2 (0+ ) = vC (0+ ) = 36 V
From (1), we get
v1(0+ ) = v2 (0+ ) 6iL (0+ ) = 0

p126
(1)
v2 v1
=
6
v dt
1
120 v2
v2 v1
=
+
(2)
4
6
1
4
dv 2
dt
From (2), we get
dv 2 +
(0 ) = 60 V/s
dt
At t=0+,
dv1
dv 2
=
6v1
dt
dt
dv1 +
dv 2 +
(0 ) =
(0 ) 6v1(0 + ) = 60 V/s
dt
dt
p127
The differential equations are
d2v1
dv1
+7
+ 10v1 = 0
2
dt
dt
2
d v2
dv 2
+7
+ 10v2 = 720
2
dt
dt
(1)
(2)
with initial conditions
v1(0+ ) = 0
v2(0+ ) = 36 V
dv1 +
dv 2 +
(0 ) =
(0 ) = 60 V/s
dt
dt
p128
Transient response: The characteristic equation

2
is
s + 7s + 10 = 0
whose roots are s1=-2 and s2=-5. We get
v1t = K1
-2t
v2t = K 3
-2t
+ K2
-5t
+ K 4
-5t
Steady-state response:
We can draw the equivalent circuit at steady state.
4
v1,ss = 0
v2,ss
6
=
(120) = 72 V
10
+
-
120V
v1,ss
+
-
v2,ss
p129
Alternatively, from the differential equations
d2v1
dv1
+7
+ 10v1 = 0
2
dt
dt
d2v2
dv 2
+7
+ 10v2 = 720
2
dt
dt
We get
v1,ss = 0
v2,ss
720
=
= 72 V
10
(1)
(2)
v1,ss = A
v2,ss = B
dv1,ss
=0
dt
dv2,ss
=0
dt
d2v1,ss
=0
2
dt
d2v1,ss
=0
2
dt
p130
Complete response: We get
v1(t) = K1-2t + K2-5t

v2 (t) = 72 + K3-2t + K 4-5t
The derivatives are
dv1
= 2K1-2t 5K 2-5t
dt
dv 2
= 2K3-2t 5K 4-5t
dt
At t=0+, we get
v1(0 ) = 0 = K1 + K2
+
dv1 +
(0 ) = 60 = 2K1 5K 2
dt
p131
Solving simultaneously, we get K1=20 and K2=-20.

Also at t=0+, we get
v2 (0 + ) = 36 = 72 + K3 + K 4
dv 2 +
(0 ) = 60 = 2K3 5K4
dt
Solving simultaneously, we get K3=-40 and
K4=4. The final expressions are
v1(t) = 20-2t 20-5t

v2(t) = 72 40
-2t
+ 4
V
-5t
t0
V
t 0
p132
Plot of the Voltages
V2
4
+
120
-
+
-
t=0
60V
+v2
+v1
6
1
1
F
4
REF
V1
p133

is initially unenergized.
At t=0, the switch is
closed. Find current i2(t)
for t 0. Assume v(t)
=20 cos 4t volts.
t=0 8 2H
+
-
v(t)
i1
i2
1H
For t 0, the mesh equations are
di1
2
+ 12i1 4i2 = v(t)
dt
(1)
di2
+ 4i2 4i1 = 0
dt
(2)
p134
Solve for i1 in equation (2) and differentiate the

resulting equation. We get
i1 =
and
di1
=
dt
1
4
di2
+ i2
dt
(3)
1
4
d2i2 di2
+
2
dt
dt
(4)
d2i2
di2
+ 10
+ 16i2 = 2v(t)
2
dt
dt
(5)
where v(t)=20 cos 4t volts.

p135
d2i2
di2
+ 10
+ 16i2 = 40 cos 4t
2
dt
dt
(5)
Transient response: The characteristic equation is
s2 + 10s + 16 = 0
The roots are s1=-2 and s2=-8. Thus
i2,t = K1 2t + K2 8t
Forced Response: Since the source is sinusoidal,
i2,ss = A cos 4t + B sin 4t

di2,ss
= 4A sin 4t + 4B cos 4t
dt
d2i2,ss
= 16A cos 4t - 16B sin 4t
2
dt
p136
d2i2
di2
+ 10
+ 16i2 = 40 cos 4t
2
dt
dt
(5)
Substitute in the differential equation (5). We get
16A cos 4t - 16B sin 4t - 40A sin 4t

+ 40B cos 4t + 16A cos 4t + 16B sin 4t = 40 cos 4t
Simplifying and comparing coefficients, we get
40 = 16A + 40B + 16A

0 = 16B 40A + 16B
Solving simultaneously, we get A=0 and B=1. Thus
i2,ss = sin 4t
p137
Initial Conditions:
The circuit is initially
unenergized so
i1(0+)
i2(0+)
=0
t=0 8 2H
+
-
v(t)
i1
i2
1H
For t 0, we ve formulated the mesh equations as
di1
2
+ 12i1 4i2 = 20 cos 4t
dt
di2
+ 4i2 4i1 = 0
dt
From (2), we get
(1)
(2)
di2
(0+) = 0
dt
p138
Complete response:
i2(t) = sin 4t + K1 2t + K2 8t
di2
= 4 cos 4t - 2K1 2t 8K2 8t
dt
Evaluating these equations at t=0 and substituting

initial conditions,
0 = K1 + K2
0 = 4 - 2K1 8K2
Solving simultaneously, we get K1=-K2=-2/3. Thus
i2 (t) = sin 4t -
2
3
2 t
2
3
8 t
A t0
p139
Plot of the Current & Voltage

V(t) = 20 cos (4t)
i2
p140
Example: Find the

complete response
vc(t) if vc(0) = 1V and
iL(0) = -62.5 mA.
10
+
-
vs(t)
2H
iL
8
+
vc
-
1
F
8
vs(t) = t2+3t+1
From KCL, we get
From KVL, we get
1 dv c 1
+ v c ! iL = 0
8 dt 8
diL
2
+ 10iL + vC = t2 + 3t + 1
dt
(1)
(2)
We can write iL in terms of vC in equation (1)
1 dvc 1
iL =
+ vc
8 dt 8
p141
Differentiating the equation for iL
diL 1 d2vc 1 dvc

=
+
2
dt 8 dt
8 dt
Substituting iL and its derivative in (2) and
simplifying, we get the differential equation for vc
d2vc
dvc
2
+
6
+
9v
=
4t
+ 12t + 4
c
2
dt
dt
Transient response: The characteristic equation

for this circuit is s2 + 6s + 9 = 0 with repeated
roots s1 = s2 = -3.
This is the critically damped case and the transient
response is of the form
vc,t(t) = (K1 + K2t)-3t
p142
Steady-state response: The differential

equation for vc is
d2vc
dvc
2
+
6
+
9v
=
4t
+ 12t + 4
c
2
dt
dt
(3)
From the table, the forced response is of the form

vc,ss(t) = c1t2 + c2t + c3
Differentiating vc,ss(t) twice,
dv c,ss
= 2c1t + c2
dt
and
d2vc,ss
dt
= 2c1
Next, we substitute the trial solution vc,ss(t) and

its derivatives into the differential equation.
p143
We get
9c1t2 + (12c1+ 9c2)t + 2c1+ 6c2+9c3 = 4t2+12t+4
Comparing coefficients,
9c1 = 4
12c1+ 9c2 = 12
2c1+ 6c2+9c3 = 4
c1 = 0.4444
c2 = 0.7407
c3 = -0.1481
Initial conditions: We are given that vc(0) = 1V

and iL(0) = -62.5 mA. To find dvc(0)/dt, evaluate
equation (1) at t=0.
dvc
1
(0) = 8iL (0) vc (0) = 1.5V/sec

dt
8
p144
Complete response:
vc(t) =0.4444t2 + 0.7407t 0.1481 + (K1 + K2t)-3t
dvc
= 0.8888t + 0.7407 + (-3K1 + K2 -3K2t)-3t
dt
Evaluating the complete response at t=0 and
substituting initial conditions
vc(0) = 1 = 0.1481 + K1
dv c
(0) = -1.5 = 0.7407-3K1 + K2
dt
K1 = 1.1481
K2 = 1.2036
Finally, we have the complete solution

vc(t) =0.4444t2 + 0.7407t 0.1481
+ (1.1481 + 1.2036 t)-3t
p145
Numerical Methods of Solving

Differential Equations
Artemio P. Magabo
Professor of Electrical Engineering

University of the Philippines - Diliman
Revised by Luis G. Sison, Jan 23, 2004
Revised by Jhoanna Pedrasa, July 2005
Numerical Methods
Solving differential equations is a fundamental
problem in science and engineering. Sometimes,
we can find closed-form solutions using calculus.
In general, however, there is no analytic solution
and the differential equation have to be solved
numerically.
Two methods for numerically approximating the
solution to ODEs are
w Euler Method
w Runge-Kutta Method
p147
Comparison of Analytical and

Numerical Solutions of ODEs
Analytic solution method
Numerical solution method

y
t2, y2
t3, y3
t1, y1
y(0)=b
t0, y0
t
t
Solve the ODE to find a

family of solutions.
Choose the solution satisfying
the correct initial conditions.
Find an analytic formula for y
(t)
Start with the initial conditions

Solve one small time step at a
time
Solve approximately at each
time step
Find pairs of points (t0,y0),
(t1,y1),
p148
Euler Method
Consider the first-order differential equation
dx = f(x,t)
dt
with initial condition x(t0) = X0. Integration gives
x(t) =
=
or
t0
f(x, t) dt
f(x, t) dt +
x(t) = X0 +
t0
t0
f(x, t) dt
f(x, t) dt
p149
If the interval from t0 to t is made very small, then
t 0 + t
f ( x , t )dt
t0
f ( x , t ) t0 ( t )
where t = t t0. Thus, we get
x(t0 + t) x(t0 ) + f(x, t) t t

0
Note: This simple method of numerical integration

is referred to as the Euler method. Unfortunately,
even with a small step size t, the method is not
very accurate.
p150
First-Order Differential Equation

Example: In the circuit shown, the switch is closed
at t=0. The source is described by v(t) = 20t volts.
Find i for t 0 using the Euler method. Use a step
size t=0.001 sec.
10
For t 0, we get from
KVL
v(t)
t=0
di
0.1
+ 10i = 20t
dt
di
or
= 200t 100i = f(i, t)
dt
0.1H
with initial condition i(0+)=0.

p151
Applying the Euler method successively, we get
i(0.001) = i(0+ ) + f(i, t) t =0+ (t)
= 0 + [200(0) 100(0)](0.001) = 0
i(0.002) = i(0.001) + f(i, t) t =0.001 (t)
= 0 + [200(0.001) 100(0)](0.001)
= 0.0002
i(0.003) = i(0.002) + f(i, t) t =0.002 (t)
= .0002 + [200(.002) 0.02](.001)

= 0.00058
p152
i(0.004) = i(0.003) + f(i, t) t =0.003 (t)
= .00058 + [200(.003) 0.058](.001)

= 0.00181
Continue!
Exact solution: The transient response is
it = K
R
t
L
= K
-100t
The steady-state response: Let iss=K1t+K2
diss
= K1
dt
p153
Substitution gives
0.1K1 + 10(K1t + K2 ) = 20t

Comparing coefficients, we get
10K1 = 20
and
0.1K1 + 10K2 = 0
or K1=2 and K2=-0.02. The steady-state response

is
iss = 2t 0.02 A
t0
Since i(0+)=0, we get
i(t) = 2t 0.02 + 0.02-100t

t0
p154
Comparison of Results:
Time
Exact
Euler
Error
0.001
0.0000968
0.0000968
0.002
0.0003746
0.0002
0.0001746
0.003
0.0008164
0.00058
0.0002364
0.004
0.0014064
0.001122
0.0002844
0.005
0.0021306
0.0018098
0.0003208
Note: For better accuracy, use a smaller step size

t. Better yet, use a more accurate method.
p155
Solution of the State Equation

The Euler method for integrating a first-order
differential equation is of the form
x(t0 + t) x(t0 ) + f(x, t) t t

0
.
where x=f(x,t). The method can be extended to

the case when x is a vector. Consider the state
equation
x = f(x, t) = A x + Bu
Substitution gives
x(t0 + t) x(t0 ) + [Ax(t0 ) + Bu(t0 )]t

(I + tA) x(t0 ) + tB u(t0 )
Identity matrix
Step-size
p156
Example: The network is initially unenergized. At

t=0, the switch is closed. t=0
8 2H
Use the Euler method
with t=0.02 sec to find +
iL1 and iL2 for t 0. Let v
iL2
v(t) iL1 4
1H
(t)=20 cos 4t V.
For t 0, we get from KVL,
di1
2
+ 12i1 4i2 = v(t)
dt
di2
+ 4i2 4i1 = 0
dt
p157
In matrix form, we get
i L1
iL 2
iL1
iL 2
0 .5
0
v(t)
with initial conditions iL1(0+) = iL2(0+) = 0.

Recall the Euler method
x(t0 + t) (I + tA) x(t0 ) + tB u(t0 )

I+tA =
1 0
0 1
+ 0.02
0.88 0.04
0.08 0.92
p158
x(t0 + t) (I + tA) x(t0 ) + tB u(t0 )

where
I+tA =
0.88 0.04
0.08 0.92
and
t=0.02
We get
iL1(t+t)
iL2(t+t)
0.88 0.04
iL1(t)
0.08 0.92
iL2(t)
0.88 0.04
iL1(0+)
0.08 0.92
iL2(0+)
0.2cos 4t
0
At t=0,
iL1(0.02)
iL2(0.02)
0.2
0
0.2
0
p159
iL1(t+t)
iL2(t+t)
0.88 0.04
iL1(t)
= 0.08 0.92
iL2(t)
0.2cos 4t
We ve found
iL1(0+)
iL2(0+)
0
0
At t=0.02 second:
iL1(0.04)
0.88 0.04
=
iL2(0.04)
0.08 0.92
iL1(0.02)
and
0.2
0
iL2(0.02)
0.1994
0
0.2
0
0.3754
0.016
At t=0.04 second:
iL1(0.06)
iL2(0.06)
0.88 0.04
0.3754
0.08 0.92
0.016
0.1974
0
0.5284
0.0448
p160
Comparison of Results
From a previous example, we got
i2 (t) = sin 4t - 23 2t + 23 8t A t 0
Comparing the actual value with the estimate,
Time
Actual
Euler
Error
0.02
0.00748
0.0
0.00748
0.04
0.02801
0.016
0.01201
0.06
0.05894
0.04475
0.01419
0.08
0.09800
0.08344
0.01456
p161
Plot of the Current
p162
Runge-Kutta Method
Euler s method is rarely used in practice because
truncation error per step size is relatively large. A
more popular method is the fourth-order RungeKutta method.
k1 = t f(xn, yn )
k 2 = t f(xn + 12 t, yn + 12 k1 )
k 3 = t f(xn + 12 t, yn + 12 k 2 )
and
k 4 = t f(xn + t, yn + k 3 )
yn+1 = yn + 16 k1 + 13 k 2 + 13 k 3 + 16 k 4
p163

Ch05 Higher Order Networks PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Ch05 Higher Order Networks PDF

Transféré par

Droits d'auteur :

Formats disponibles

Chapter 5

Electrical and Electronics Engineering Institute

Revised by Jhoanna Pedrasa, July 2005

where K and s are constants. Substitution gives

Department of Electrical and Electronics Engineering

After canceling the exponential term, we get the

Source-Free Series RLC Network

The characteristic equation is

From the quadratic formula, we get the two roots

1. Overdamped Case: The roots are real and

2. Critically Damped Case: The roots are real but

The solution can be shown to be

3. Underdamped Case: The roots are complex

s1, s2 = - jd, the solution can

Department of Electrical and Electronics Engineering

underdamped envelope response

Department of Electrical and Electronics Engineering

Example: The circuit has reached steady-state

From KVL, we get

Example: The circuit

From KVL, we get for t 0,

The characteristic equation is thus

The roots are s1=-2 and s2=-8. Thus, we get

From the previous example, we ve found that

Department of Electrical and Electronics Engineering

Solving simultaneously, we get K1=-2 and K2=2.

i(t) = -2 -2t + 2 -8t Amp t 0

Note: For an over-damped case, the solution

Department of Electrical and Electronics Engineering

Example: The circuit

The characteristic equation can be shown to be

Assume i(t) = y(t) -4t. Differentiating twice, we get

The original differential equation is

+ y' ' (t)

Department of Electrical and Electronics Engineering

y' ' (t)

y' ' (t) = 0

Integrating twice, we get

Finally, the solution is

i(t) = K1t -4t + K2 -4t

Department of Electrical and Electronics Engineering

i(t) = K1t -4t + K2 -4t

Department of Electrical and Electronics Engineering

Example: The circuit

The characteristic equation can be shown to be

Department of Electrical and Electronics Engineering

To prove the first identity, let y=cos x + j sin x.

j = 1, the equation can be re-written

Department of Electrical and Electronics Engineering

Integrating both sides, we get

Thus we get ln y = jx, or

Note: The other Euler s identity can be verified

Back to the expression for the current

i(t) = -3t (K1 j2.65 t + K2 - j2.65 t )

[ K1 cos t + jK1 sin t

+ K2 cos t jK2 sin t]

i(t) = 3t [ K3 cos t + K 4 sin t ]

The initial conditions are

i(t) = 4.54 3t sin 2.65 t

Plot of the Currents

Source-Free Parallel RLC Network

The characteristic equation is

From the quadratic formula, we get the two roots