Dynamic Circuit Analysis via

Laplace Transform
Stefano GRIVET TALOCIA ([email protected])
Igor Simone STIEVANO ([email protected])
Dept. Electronics and Telecommunications
Politecnico di Torino, Italy
http://www.emc.polito.it/
Problem statement
𝑣𝑦 𝑡

Initial conditions known

𝑔𝑘 𝑡
𝑖𝐿𝑘 0− , 𝑣𝐶𝑘 0−

𝑖𝐿𝑘
Arbitrary sources

𝑣𝐶𝑘 𝑔𝑘 𝑡 , ∀𝑡 > 0−
(Laplace transform
𝑖𝑦 𝑡 must exist)

Linear Time-Invariant (LTI), dynamic circuit

Objective: find any voltage 𝑣𝑦 𝑡 or current 𝑖𝑦 𝑡 for 𝑡 > 0

 Working example
𝑣1
𝑖1

𝑖𝐶 𝑖𝐿
𝑅1
𝑒(𝑡) 𝐶 𝑣𝐶 𝑎(𝑡) 𝐿
 Working example
𝑣1 𝑉1 (𝑠)
𝑖1 𝐼1 (𝑠)

𝑖𝐶 𝑖𝐿 𝐼𝐶 (𝑠) 𝐴(𝑠) 𝐼𝐿 𝑠
𝑅1
𝑒(𝑡) 𝐶 𝑣𝐶 𝑎(𝑡) 𝐿 𝐸 𝑠 𝑉𝐶 (𝑠)

ℒ
KCL −𝑖1 𝑡 + 𝑖𝐶 𝑡 − 𝑎 𝑡 + 𝑖𝐿 𝑡 = 0 −𝐼1 𝑠 + 𝐼𝐶 𝑠 − 𝐴 𝑠 + 𝐼𝐿 𝑠 = 0

ℒ
KVL 𝑣𝐶 𝑡 + 𝑣1 𝑡 = 𝑒 𝑡 𝑉𝐶 𝑠 + 𝑉1 𝑠 = 𝐸 𝑠

The new circuit has the same KCL/KVL of the original circuit, but in the Laplace domain

Same circuit topology

Laplace-transforming circuit elements: sources

𝑒(𝑡) 𝑣(𝑡)
𝑖(𝑡)

𝑣 𝑡 = 𝑒 𝑡 , ∀𝑖 ∀𝑡

𝑖(𝑡)
𝑎(𝑡) 𝑣(𝑡)

𝑖 𝑡 = 𝑎 𝑡 , ∀𝑣 ∀𝑡
Laplace-transforming circuit elements: sources

𝑒(𝑡) 𝑣(𝑡) ℒ 𝐸(𝑠) 𝑉(𝑠)

𝑖(𝑡) 𝐼(𝑠)

𝑣 𝑡 = 𝑒 𝑡 , ∀𝑖 ∀𝑡
ℒ 𝑉 𝑠 = 𝐸 𝑠 , ∀𝐼 ∀𝑠

𝑖(𝑡) 𝐼(𝑠)
𝑎(𝑡)
ℒ 𝐴(𝑠)
𝑣(𝑡) 𝑉(𝑠)

𝑖 𝑡 = 𝑎 𝑡 , ∀𝑣 ∀𝑡
ℒ 𝐼 𝑠 = 𝐴 𝑠 , ∀𝑉 ∀𝑠
Laplace-transforming circuit elements: resistive elements

𝑅 𝑣(𝑡)
𝑖(𝑡)

𝑣 𝑡 = 𝑅𝑖 𝑡 , ∀𝑡

𝑖(𝑡)
𝑔𝑚 𝑣𝑥 (𝑡) 𝑣(𝑡)

𝑖 𝑡 = 𝑔𝑚 𝑣𝑥 (𝑡), ∀𝑣 ∀𝑡
Laplace-transforming circuit elements: resistive elements

𝑅 𝑣(𝑡) ℒ 𝑅 𝑉(𝑠)
𝑖(𝑡) 𝐼(𝑠)

𝑣 𝑡 = 𝑅𝑖 𝑡 , ∀𝑡
ℒ 𝑉 𝑠 = 𝑅𝐼 𝑠 , ∀𝑠

𝑖(𝑡) 𝐼(𝑠)
𝑔𝑚 𝑣𝑥 (𝑡)
ℒ 𝑔𝑚 𝑉𝑥 (𝑠)
𝑣(𝑡) 𝑉(𝑠)

𝑖 𝑡 = 𝑔𝑚 𝑣𝑥 (𝑡), ∀𝑣 ∀𝑡
ℒ 𝐼 𝑠 = 𝑔𝑚 𝑉𝑥 𝑠 , ∀𝑉 ∀𝑠

…and similarly for all other resistive elements: OpAmp, ideal transformer, etc…
Laplace-transforming circuit elements: the inductor

𝐿 𝑑𝑖𝐿 𝑡
𝑣𝐿 (𝑡) 𝑣𝐿 𝑡 = 𝐿
𝑑𝑡
𝑖𝐿 (𝑡)
Laplace-transforming circuit elements: the inductor

𝐿 𝑣𝐿 (𝑡) 𝑣𝐿 𝑡 = 𝐿
𝑑𝑖𝐿 𝑡 ℒ 𝑉𝐿 𝑠 = 𝐿 𝑠𝐼𝐿 𝑠 − 𝑖𝐿 0−
𝑑𝑡
𝑖𝐿 (𝑡)

1 1
Impedance: 𝑍𝐿 𝑠 = 𝑠𝐿 Admittance: 𝑌𝐿 𝑠 = 𝑠𝐿 = 𝑍
𝐿 𝑠

𝑖𝐿 0−
𝑉𝐿 𝑠 = 𝑍𝐿 𝑠 𝐼𝐿 𝑠 − 𝐿𝑖𝐿 0− 𝐼𝐿 𝑠 = 𝑌𝐿 𝑠 𝑉𝐿 𝑠 +
𝑠

𝐼𝐿 (𝑠) 𝐼𝐿 (𝑠)

𝑍𝐿 𝑠
𝑖𝐿 0− 𝑉𝐿 (𝑠)
𝑉𝐿 (𝑠) 𝑌𝐿 𝑠
𝑠
𝐿𝑖𝐿 0−
Laplace-transforming circuit elements: the capacitor

𝐶 𝑑𝑣𝐶 𝑡
𝑣𝐶 (𝑡) 𝑖𝐶 𝑡 = 𝐶
𝑑𝑡
𝑖𝐶 (𝑡)
Laplace-transforming circuit elements: the capacitor

𝐶
𝑣𝐶 (𝑡) 𝑖𝐶 𝑡 = 𝐶
𝑑𝑣𝐶 𝑡 ℒ 𝐼𝐶 𝑠 = 𝐶 𝑠𝑉𝐶 𝑠 − 𝑣𝐶 0−
𝑑𝑡
𝑖𝐶 (𝑡)

1 1
Impedance: 𝑍𝐶 𝑠 = 𝑠𝐶 Admittance: 𝑌𝐶 𝑠 = 𝑠𝐶 = 𝑍
𝐶 𝑠

𝑣𝐶 0−
𝑉𝐶 𝑠 = 𝑍𝐶 𝑠 𝐼𝐶 𝑠 + 𝐼𝐶 𝑠 = 𝑌𝐶 𝑠 𝑉𝐶 𝑠 − 𝐶𝑣𝐶 (0− )
𝑠

𝐼𝐶 (𝑠) 𝐼𝐶 (𝑠)

𝑍𝐶 𝑠
𝑉𝐶 (𝑠)
𝑉𝐶 (𝑠) 𝐶𝑣𝐶 (0− ) 𝑌𝐶 𝑠
𝑣𝐶 0−
𝑠
 Back to the example…
𝑣1 𝑉1 (𝑠)
𝑖1 𝐼1 (𝑠)

𝑖𝐶 𝑖𝐿 𝐼𝐶 (𝑠) 𝐴(𝑠) 𝐼𝐿 𝑠
𝑅1
𝑒(𝑡) 𝐶 𝑣𝐶 𝑎(𝑡) 𝐿 𝑉𝐶 (𝑠)
𝐸 𝑠

ℒ
KCL −𝑖1 𝑡 + 𝑖𝐶 𝑡 − 𝑎 𝑡 + 𝑖𝐿 𝑡 = 0 −𝐼1 𝑠 + 𝐼𝐶 𝑠 − 𝐴 𝑠 + 𝐼𝐿 𝑠 = 0
ℒ
KVL 𝑣𝐶 𝑡 + 𝑣1 𝑡 = 𝑒 𝑡 𝑉𝐶 𝑠 + 𝑉1 𝑠 = 𝐸 𝑠

The new circuit has the same KCL/KVL of the original circuit, but in the Laplace domain

Same circuit topology

 The symbolic (s-domain) circuit
𝑣1
𝑖1 𝑅 𝐼1 (𝑠)

𝑖𝐶 𝑖𝐿 𝐼𝐶 (𝑠) 𝐼𝐿 (𝑠)
𝑅1 𝑉𝐶 (𝑠)
𝑒(𝑡) 𝐶 𝑣𝐶 𝑎(𝑡) 𝐿 𝑍𝐶 𝑠 𝐴(𝑠) 𝑍𝐿 𝑠

𝐸 𝑠 𝑣𝐶 0−
Time-domain circuit 𝑠 𝐿𝑖𝐿 0−
ℒ
Symbolic circuit

• Same topology of time-domain (original) circuit (from KVL and KCL)

• Circuit variables are the Laplace transforms of time-domain circuit variables
• Elements are defined by Laplace-transformed characteristic equations
• All symbolic circuit equations are the Laplace transforms of time-domain equations
• Symbolic circuit equations are ALGEBRAIC, not differential
• The solution of the symbolic circuit is the ℒ-transform of the time-domain solution
Units
+∞
Voltage 𝑣 𝑡 ; units: Volts 𝑉 𝑠 = 𝑣 𝑡 𝑒 −𝑠𝑡 𝑑𝑡 units: Volts×seconds
0−
ℒ +∞
Current 𝑖 𝑡 ; units: Amps 𝐼 𝑠 = 𝑖 𝑡 𝑒 −𝑠𝑡 𝑑𝑡 units: Amps×seconds
0−

𝑅 𝑣(𝑡) 𝑉 𝑠 =𝑍 𝑠 𝐼 𝑠
𝑍𝑅 𝑠 = 𝑅
𝑖(𝑡)

𝐼(𝑠) 𝑉 𝑠
𝐶 1 𝑍 𝑠 =
𝑣(𝑡) 𝑍𝐶 𝑠 = 𝐼 𝑠
𝑠𝐶 𝑍 𝑠 𝑉(𝑠)
𝑖(𝑡)

Units: Ohms
𝐿 𝑣(𝑡) 𝑍𝐿 𝑠 = 𝑠𝐿 …assuming no
𝑖(𝑡) initial conditions
Symbolic circuit analysis
Remarks

 The equivalent circuit in the s-domain is composed of circuit elements

defined by linear algebraic constitutive relations (Z(s) or Y(s))
 The symbolic circuit shares the same properties and methods of resistive
networks (e.g., superposition, Thévenin and Norton Theorems,…)

Procedure

1) Determine the initial conditions

2) Generate the equivalent (symbolic) circuit in the s-domain
3) Solve the circuit (any method for resistive circuits can be used)
4) Determine the transient response via inverse Laplace transformation
Example
R1
Determine the transient response e(t )  3  12u(t ) V
e(t )
of the voltage v(t), for t  0. 
C R2 v(t ) R1  3 


R2  6 
C 2F
Example
R1
Determine the transient response e(t )  3  12u(t ) V
e(t )
of the voltage v(t), for t0s 
C R2 v(t ) R1  3 


1) Initial conditions R2  6 
C 2F
3
3V 6

vC (0 ) 6 vC (0 )  3 2V

3 6

2) Equivalent circuit in the s-domain

3
Z C  1 2 s ; YC  2 s
E (s) 6
 CvC (0 )  2vC (0 )  4 A
 V (s)
1 2s 15
E ( s)  !

2vC (0 ) s
 3
Example
E (s) 6

 V (s)
1 2s
3) Symbolic circuit solution in s

E ( s) 2vC (0 )  4 A
 2vC (0 )
E ( s) 1 1
V ( s)  3   vC (0 ) 
1 1 6 s 1 1
  2s s
3 6 4 4

4) Conversion to t-domain via Laplace inversion

15 1 2 10 8
V ( s)    
6  1 1 s s 1
s s   s 
 4 4 4 ℒ −1   t
1
v (t )  10  8e 4 u(t ) V
 
General properties of LTI circuits – Preview
Structure of the solution

1 16
V ( s )  vC (0 )  E ( s)
s 1 4 s 1 4
zero-input zero-state
(natural) (forced or
response input-output)
response

V ( s)  VZI ( s)  VZS ( s)
Mode (natural frequency =1/4)
1
 t
(pole)
 1 
VZI ( s)  vC (0 )  vZI (t )  vC (0 )e u(t )
4
s 1 4

Order of complexity N = number of natural frequencies ≡ poles (N  nC+nL)

General properties of LTI circuits – Preview

V ( s)  VZI ( s)  VZS ( s)

16 Network VZS ( s )
VZS ( s )  E ( s ) H ( s) 
s 1 4 function E ( s)

In general,
 for each source, a Network function is defined as: output/input (with all the
other inputs, i.e., sources, and initial conditions = 0)
 VZS(s) and VZI(s) share the same poles

For the example at hand,

1
15 1 6 A B  t
VZS ( s)     vZS (t )  Au (t )  Be 4
u (t )
s s 1 4 s s 1 4
General properties of LTI circuits – Preview

A B 1
 t For exponentially stable
VZS ( s)    vZS (t )  Au (t )  Be 4 u(t ) circuits 0 as t 
s s 1 4

Part associated to the

network only

Part associated to the source (excitation) only

The steady-state of an LTI circuit with DC sources is constant (DC)

Comparing approaches for LTI (dynamic) circuit analysis
𝑔𝑘 𝑡
Write state equations
𝑑𝒙 𝑡
= 𝑨 𝒙 𝑡 + 𝑩𝒈 𝑡
𝑑𝑡
𝑖𝐿𝑘

𝑣𝐶𝑘

Solve state equations for

𝒙(𝑡) (diagonalization)

Find initial conditions

𝑣𝐶𝑘 0− , 𝑖𝐿𝑘 0−

Substitution Theorem

Solution 𝑦(𝑡)
Comparing approaches for LTI (dynamic) circuit analysis
𝑔𝑘 𝑡
Write state equations
Derive high-order ODE in 𝑦(𝑡)
𝑑𝒙 𝑡
= 𝑨 𝒙 𝑡 + 𝑩𝒈 𝑡 + required initial conditions
𝑑𝑡
𝑖𝐿𝑘

𝑣𝐶𝑘

Solve state equations for

𝒙(𝑡) (diagonalization)

Solve high-order ODE

Find initial conditions
𝑣𝐶𝑘 0− , 𝑖𝐿𝑘 0−

Substitution Theorem

Solution 𝑦(𝑡)
Comparing approaches for LTI (dynamic) circuit analysis
𝑔𝑘 𝑡
Write state equations
Derive high-order ODE in 𝑦(𝑡)
𝑑𝒙 𝑡
= 𝑨 𝒙 𝑡 + 𝑩𝒈 𝑡 + required initial conditions
𝑑𝑡
𝑖𝐿𝑘

𝑣𝐶𝑘

Solve state equations for

𝒙(𝑡) (diagonalization)

ℒ
Derive symbolic circuit Solve high-order ODE
Find initial conditions
𝑣𝐶𝑘 0− , 𝑖𝐿𝑘 0−
𝐺𝑘 𝑠
Substitution Theorem
𝑠𝐿𝑘

1 𝑥𝑘 0−
𝑠𝐶𝑘 Solve for 𝑌(𝑠) as a
resistive circuit
ℒ −1
Solution 𝑦(𝑡)
Comparing approaches for LTI (dynamic) circuit analysis
𝑔𝑘 𝑡

𝑖𝐿𝑘

𝑣𝐶𝑘

ℒ
Derive symbolic circuit
Find initial conditions
𝑣𝐶𝑘 0− , 𝑖𝐿𝑘 0−
𝐺𝑘 𝑠

𝑠𝐿𝑘

1 𝑥𝑘 0−
𝑠𝐶𝑘 Solve for 𝑌(𝑠) as a
resistive circuit
ℒ −1
Solution 𝑦(𝑡)
General solution of LTI (dynamic) circuits
𝑔𝑘 𝑡
By superposition (symbolic circuit analysis):

𝑖𝐿𝑘

𝑣𝐶𝑘 𝑌 𝑠 = 𝐻𝑘′ 𝑠 𝑥𝑘 0− + 𝐻𝑘 𝑠 𝐺𝑘 𝑠
𝑘 𝑘
zero-input zero-state
(natural) (forced or
ℒ response input-output)
response
Derive symbolic circuit
Find initial conditions
𝑣𝐶𝑘 0− , 𝑖𝐿𝑘 0− All coefficients 𝐻𝑘′ 𝑠 and 𝐻𝑘 𝑠
𝐺𝑘 𝑠 are rational functions of 𝑠 with
𝑠𝐿𝑘
the same poles 𝑝𝑛 .

1 𝑥𝑘 0−
𝑠𝐶𝑘 Solve for 𝑌(𝑠) as a
resistive circuit
ℒ −1
Solution 𝑦(𝑡)
The zero-input response

No independent sources: By superposition (symbolic circuit analysis):

set 𝐺𝑘 𝑠 = 0

𝑌 𝑠 = 𝐻𝑘′ 𝑠 𝑥𝑘 0− + 𝐻𝑘 𝑠 𝐺𝑘 𝑠
𝑘 𝑘
𝑌𝑍𝐼 𝑠 = 𝑌 𝑠
𝐺𝑘 𝑠 =0 zero-input
𝑌𝑍𝐼 (𝑠): (natural)
response

The coefficients 𝐻𝑘′ (𝑠) are (strictly proper) rational functions

𝐺𝑘 𝑠 = 0
𝜇𝑛 𝜇
𝑠𝐿𝑘
𝑅𝑛 Partial fraction
𝑌𝑍𝐼 𝑠 = 𝜇
𝑠 − 𝑝𝑛 decomposition
𝑛 𝜇=1
−1
1 𝑥𝑘 0−
ℒ 𝜇𝑛
𝑠𝐶𝑘 𝜇
𝑅𝑛
𝑦𝑍𝐼 𝑡 = 𝑡𝜇−1 𝑒 𝑝𝑛 𝑡 𝑢(𝑡)
𝜇−1 !
𝑛 𝜇=1
 𝑗𝜔
Stability

The circuit is (asymptotically) stable iff

all poles have strictly negative real part
𝜎

ℜ 𝑝𝑛 < 0, ∀𝑛 poles
𝑡→+∞
𝑦𝑍𝐼 𝑡 0

The coefficients 𝐻𝑘′ (𝑠) are (strictly proper) rational functions

𝐺𝑘 𝑠 = 0
𝜇𝑛 𝜇
𝑠𝐿𝑘
𝑅𝑛 Partial fraction
𝑌𝑍𝐼 𝑠 = 𝜇
𝑠 − 𝑝𝑛 decomposition
𝑛 𝜇=1
−1
1 𝑥𝑘 0−
ℒ 𝜇𝑛
𝑠𝐶𝑘 𝜇
𝑅𝑛
𝑦𝑍𝐼 𝑡 = 𝑡𝜇−1 𝑒 𝑝𝑛 𝑡 𝑢(𝑡)
𝜇−1 !
𝑛 𝜇=1
The zero-state (input-output) response

No initial conditions: By superposition (symbolic circuit analysis):

set 𝑥𝑘 0− = 0

𝑌 𝑠 = 𝐻𝑘′ 𝑠 𝑥𝑘 0− + 𝐻𝑘 𝑠 𝐺𝑘 𝑠
𝑘 𝑘
𝑌𝑍𝑆 𝑠 = 𝑌 𝑠
𝑥𝑘 0− =0 zero-state
(forced or
𝑌𝑍𝑆 𝑠 : input-output)
response

The coefficients 𝐻𝑘 (𝑠) are rational functions

𝐺𝑘 𝑠

𝑠𝐿𝑘
𝑌 𝑠
𝐻𝑘 𝑠 = 𝑥𝑘 0− =0
𝐺𝑘 𝑠
𝐺𝑘′ 𝑠 =0, ∀𝑘 ′ ≠𝑘
1 𝑥𝑘 0−
𝑠𝐶𝑘
Transfer function between 𝐺𝑘 𝑠 and 𝑌(𝑠)
The impulse response

No initial conditions: By superposition (symbolic circuit analysis):

set 𝑥𝑘 0− = 0
𝑌 𝑠 = 𝐻𝑘′ 𝑠 𝑥𝑘 0− + 𝐻𝑘 ′ 𝑠 𝐺𝑘 ′ 𝑠
Set 𝐺𝑘 ′ 𝑠 = 0, ∀𝑘 ′ ≠ 𝑘 𝑘 𝑘 ′ ≠𝑘

+ 𝐻𝑘 𝑠 𝐺𝑘 𝑠
Set 𝐺𝑘 𝑠 = 1 ⇔ 𝑔𝑘 𝑡 = 𝛿(𝑡)

𝑌 𝑠
𝑌 𝑠 = 𝐻𝑘 𝑠 𝐻𝑘 𝑠 = 𝑥𝑘 0− =0
𝐺𝑘 𝑠 = 1 only! 𝐺𝑘 𝑠
𝐺𝑘′ 𝑠 =0, ∀𝑘 ′ ≠𝑘
ℒ −1
𝑠𝐿𝑘
𝑦 𝑡 = ℎ𝑘 𝑡 Impulse response
1 𝑥𝑘 0−
𝑠𝐶𝑘
between input 𝐺𝑘 𝑠 and output 𝑌(𝑠)

ℎ𝑘 𝑡 = ℒ −1 𝐻𝑘 𝑠
Example
𝑉 𝑠
Find the transfer function 𝐻 𝑠 = 𝑉𝑜 and the corresponding impulse response ℎ 𝑡
𝑖 𝑠

𝐿 = 2𝐻, 𝑅 = 4Ω, 𝐶 = 0.5𝐹

Example
𝑉 𝑠
Find the transfer function 𝐻 𝑠 = 𝑉𝑜 and the corresponding impulse response ℎ 𝑡
2 𝑠