Methods of Formulating

Chapter 3 Dynamic Equations

Mathematical Models of 1. Kirchoff’s Laws and Newton’s Second

Electromechanical Systems Law of Motion (or D’Alemberts
Principle)

Artemio P. Magabo 2. Lagrange Equations

Professor of Electrical Engineering

Note: Emphasis will be on method 2.

Department of Electrical and Electronics Engineering
University of the Philippines - Diliman
Department of Electrical and Electronics Engineering

Electromechanical Transducer Dynamic Equations of the

The electrical system consists of several mutually-
Electrical System
coupled windings. The mechanical system contains For the kth coil, we get
only one moving part. d
i1 R1 vk = R kik + λk k = 1,2 ,... n
+
dt
v1 L11 where n

-
x
λk = ∑L
j =1
kj ( x) i j
i2 R2 L12
K
+
L22 fmag Substitution gives
v2
L1n n di j n dL kj (x)
-
L2n
v k = R kik + ∑L kj (x )
dt
+ ∑i j
dt
in Rn fext j =1 j =1

+ D
Lnn transformer speed
vn
- M voltage voltage

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

1
We get Dynamic Equation of the
di1 di di Mechanical System
vk = Rkik + {Lk1(x) + Lk2(x) 2 + ... + Lkk(x) k
dt dt dt Applying Newton’s second law, we get
di
+ ... + Lkn(x) n }
dt
∑F x = Ma
Substitution gives
dL (x) dL (x) dL (x) dx d2 x
+ {i1 k1 + i2 k2 + ... + ik kk fmag + fext − D − Kx = M 2
dx dx dx dt dt
dL (x) dx
+ ... + in kn } where
dx dt n
dL kk (x) n n dL kj (x)
∑ ik ∑ ∑i i
1 2 1
fmag = 2
+ 2 k j
k = 1,2,...n k =1 dx k =1 j =1 dx
j≠ k

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Method of Analysis Lagrange Equations

Dynamic equations from energy and co-energy
1. Develop an approximate model of the electrical state functions
and mechanical systems. Define a general variable ξ , also called a
generalized coordinate to represent
2. Formulate the general dynamic equations that
describe the model. x ,θ = displacement
q = charge
3. Determine the expressions of parameters like λ = flux linkage
self- and mutual- inductances, etc.
The derivative of the generalized coordinate, ξ

4. Solve the resulting system of differential represent
equations. x
, θ
= velocity
q
= current
λ
= voltage
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

2
Energy, Coenergy and Dissipation Mechanical System
Functions Mechanical Coenergy State Function
Coenergy State Function n

∑
2
K.E ' (x
1 , x
2 ,...x
n ) =


1
2
Mix
i
T ' ( ξ
1 , ξ
2 ,... ξ
n , ξ 1 , ξ 2 ,... ξ n , t ) i =1

Mechanical Energy State Function

Energy State Function m

∑ 1 2
P.E.(x1 , x 2 ,...x m ) = 2
K jx j
U ( ξ1 , ξ 2 ,...ξ n , t ) j =1

Mechanical Rayleigh’s Dissipation Function

Rayleigh’s Dissipation Function
k

∑ 1 2
D (ξ
1 , ξ
2 ,... ξ
n ) Dm ' (x
1 , x
2 ,...x
k ) = 2
Dlx
l
l =1

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Electrical System Choice of State Functions

Electrical Coenergy State Function For Loop Formulation
W' fm (i1 , i2 ,...in ) or W' fe (e1 , e2 ,...en ) T ' = K.E. ' +W' fm (i1 , i2 ,... in )
Electrical Energy State Function U = P.E. + Wfe (q1 , q2 ,... qn )
Wfm (λ1 , λ 2 ,...λ n ) or Wfe (q1 , q2 ,...qn )
D = Dm ' + De (i1 , i2 ,... in )
Electrical Rayleigh’s Dissipation Function
n For Nodal Formulation
∑ 1 2
De (i1 , i2 ,...in ) = 2
R ji j T ' = K.E. '+W' fe (e 1 , e 2 ,... e n )
j =1

n ej
2 U = P.E. + Wfm ( λ 1 , λ 2 ,... λ n )
D'e (e1, e2 ,...en ) = ∑j=1
1
2
Rj D = Dm '+ De ' (e1 , e 2 ,...e n )
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

3
 Electrical System Magnetic Field System
Electrical Variables for Loop Formulation Magnetic field systems are generally analyzed
(Loop Currents) using loop formulation.

Generalized Coordinate: ∫ idt = q Generalized

λ
Coordinate = q
Electrical Variables for Nodal Formulation W'fm (i, x)
(Node Voltages)

Generalized Coordinate: ∫ vdt = λ For a linear system, i=q

Note: For Lagrange formulation, we define e = λ
. W'fm (i, x) = 1

2
L(x)i2 = 1
2
L (x)q
2

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Magnetic Field System Electric Field System

Using Loop Formulation An electricic field system may be analyzed using
loop or nodal formulation.
T ' = K.E.'+W' fm (q
1 , q
2 ,...q
n ) Wfe (q, x)
q
U = P.E. + W fe ( q1 , q 2 ,... q n ) Generalized
Coordinate = q or λ
D = Dm '+ De (q
1 , q
2 ,...q
n ) W'fe (e, x)
Note: Since we are dealing with a magnetic field
system, Wfe will be non-zero only if there is a For a linear system, e = λ

constant capacitor C in the loop. Under this 1 q2
Wfe (q, x) = 2
condition, 2 C(x)
q
Wfe = 1
2
C W'fe (e, x) = 1
2
C(x)e 2 = 1
2
C(x)λ
2
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

4
 Electric Field System Electric Field System
Using Loop Formulation Using Nodal Formulation
T ' = K.E.'+W' fm (q
1 , q
2 ,...q
n ) T ' = K.E. ' + W' fe ( λ
1 , λ
2 ,... λ
n )
U = P.E. + W fe ( q 1 , q 2 ,... q n ) U = P.E. + Wfm (λ1 , λ 2 ,...λn )

D = Dm '+ De (q
1 , q
2 ,... q
n ) D = Dm '+ De ' ( λ
1 , λ
2 ,... λ
n )

Note: Since we are dealing with an electric field Note: Since we are dealing with an electric field
system, W’fm will be non-zero only if there is a system, Wfm will be non-zero only if there is a
constant inductor L in the loop. Under this constant inductor L connected to the node. Under
condition, this condition, 2
λ
W'fm = 1
2
L q
2 Wfm = 1
2
L
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Choice of State Functions General Form of the Lagrange

For Loop Formulation Equation
T ' = K.E. '+W' fm (q
1 , q
2 ,... q
n ) Define the Lagrangian function, L

U = P.E. + W fe ( q 1 , q 2 ,... q n ) L (ξ
, ξ, t) = T ' (ξ
, ξ, t) − U (ξ, t)

D = Dm ' + De ( q
1 , q
2 ,... q
n ) Write a Lagrangian equation for each generalized

coordinate ξk
For Nodal Formulation
d  ∂L (ξ
, ξ, t)  ∂L (ξ
, ξ, t) ∂D(ξ
)
T ' = K.E. ' + W' ( λ
1 , λ
2 ,... λ
n )  − + = Pk
fe dt  ∂ξ
k  ∂ξ k ∂ξ
k
U = P.E. + Wfm (λ 1 , λ 2 ,...λ n ) k = 1,2,...n
D = Dm '+ De ' (λ
1 , λ
2 ,...λ
n ) Note: Pk is an externally-applied forcing function,
which is acting on the kth generalized coordinate.

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

5
Example 1: In the electromagnetic plunger shown, Calculate the coil inductance
the area of center leg is twice the area of the outer x Rg
legs. The sleeve is non-magnetic. When the spring Rg =
µo A
is unstretched, the air gap is g. Assume that the φ
d Rd Rd
iron is infinitely permeable. Determine the dynamic Rd = Ni
equations that describe the system. µ o (A / 2)
sleeve
x+d
d Rtot = Rg + 12 Rd =
R µo A
N
+ K We get
v q
M Ni Niµ o A
-
φ= =
x fext Rtot x+d
D Nφ N2µ o A
L(x) = =
g i x+d
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

The electrical part of the system will be analyzed Energy State Function
using loop analysis. The generalized coordinates
are q and x. U = P.E. + W fe ( q1 , q 2 ,... q n )
Coenergy State Function P.E. = 12 K(x − g)2
T ' = K.E.+
' W'fm (q
, x) Note: (1) The spring is stretched by a length (g-x).

K.E.' = 1
M x
2 (2) Wfe (q, x) = 0 since there is no capacitor
2
2 in the system.
N µoA
2 We get
W' fm (q
, x) = 1
2
L(x)i2 = 1
2
q
x+d U = 12 K(x − g)2
We get Rayleigh’s Dissipation Function
N2µ o A
2
T' = 1
M x
2 + 1
q
2 2
x+d D (q
, x
) = 1
2
R q
2 + 1
2
D x
2

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

6
Lagrangian Function  N2 µ o A   N2 µ o A 
 q

− q
 2 
x
+ R q
= v
L = T '−U x+d  (x + d) 
N2 µ o A
2 which can be rewritten as
L (q
, x
, x ) = 1
2
M x
2 + 1
2
q − 1
2
K (x − g)2 di dL (x) dx
x+d L (x ) +i + Ri = v
Lagrange Equation - Electrical System dt dx dt
Lagrange Equation - Mechanical System
d  ∂L  ∂L ∂D
q:   − + = Pq d  ∂L  ∂L ∂D
dt  ∂ q  ∂q

∂q
x: − + = Px
 
dt  ∂x  ∂x

∂x

d  N2 µ o A 
q
 − 0 + R q
= v
[M x
] − ∂  12 N µ o A q
2  + K(x − g) + D x
= fext
2
 d
dt  x + d 
dt ∂x  x + d 
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

 2
 Free-body Diagram: (Check)


+  12 N µ o A2 q
2  + K(x − g) + D x
= fext
Mx
 (x + d)  fmag M
K(g-x)
fext
which can be written as
x D x

 1 N2 µ o A 2 
M x + D x + K(x − g) = −  2



2
q
 + fext g
 ( x + d) 
Applying Newton’s second law, we get
or
Mx


+ D x
+ K(x − g) = fmag + fext ∑F x = Ma
fmag + fext + K(g − x) − D x
= M x



Note: The assumed direction of motion defines the
direction of the magnetic force. Also from the or
diagram, the spring has been stretched. Mx


+ Dx
+ K(x − g) = fmag + fext

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

7
Example: The electric field transducer shown has Capacitance Calculation
a solid copper plunger with a mass M and a radius
Assume a cylindrical Gaussian surface of radius r.
b. The radius of the outer copper tube is a. When
x = 0, the plunger is halfway into the tube and l
From Gauss’ Law x
l 2
the spring is l → → -Q
unstretched. x 2
Copper Q = ∫D • dA a r
Plunger s b
Neglect fringing +Q
of the flux and K  l
fext M = D(2πr )  x + 
assume that  2
the insulating Cylindrical Gaussian Surface
bushing has a R l
damping
Solving for the flux density, we get
+ λ - Insulating
coefficient D. Bushing
→
Q
D= âr
π(2x + l) r
i(t)
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

The electric field intensity is The electrical part of the system will be analyzed
→
→
D Q using nodal analysis. The generalized coordinates
E= = âr are λ and x.
εo πε o (2x + l) r
Coenergy State Function
The voltage drop between the plunger and the outer
tube is a → → T ' = K.E.'+W' fe (λ
, x)
Vba = ∫ b
E⋅ d L
K.E.' = 1
2
M x
2
Q dr a Q a
= ∫ = ln  πε o (2x + l)
2
πεo (2x + l) b r πεo (2x + l) b W'fe (λ
, x) = 1
2
C(x)e2 = 1
2  ln(a b)  λ
 
The capacitance of the system is
We get
 πε o (2x + l)
2
Q πε (2x + l) T' = 1
M x
2 + 1
 ln( a b)  λ
C(x) = = o 2 2
 
Vba ln(a b)
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

8
Energy State Function Lagrangian Function

U = P.E. + Wfm (λ, x) L = T ' −U

P.E. = 12 Kx2  πε o (2 x + l)
2
L = 1
2
M x
2 + 1
2  ln( a b)  λ −
1
2
Kx 2
Note: The spring is stretched by a length x.
 

W fm ( λ , x ) = 0 since there is no inductor Lagrange Equation - Electrical System

in the system.
We get d  ∂L  ∂L ∂D
λ:   − + = Pλ
U = 1
2
Kx 2

dt  ∂λ  ∂λ ∂λ

Rayleigh’s Dissipation Function
d  πε o (2 x + l)
 λ

λ
2 λ − 0 + = i(t)
D (λ
, x
) = 1
2
+ 1
2
D x
2 dt  ln( a b) 
 R
R
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

 πε o (2x + l)


 2 πε o  λ
πε o
2
 ln( a b)  λ + λ  ln( a b)  x

+ = i(t) Mx


− λ + Kx + D x
= − fext
    R ln( a b)
which can be rewritten as
which can be written as
dv dC (x) dx v
C(x) +v + = i(t) πε o
2
dt dx dt R Mx


+ D x
+ Kx = λ − fext
Lagrange Equation - Mechanical System ln( a b)
d  ∂L  ∂L ∂D or
x:   − + = Px
dt  ∂x
 ∂x ∂x
Mx


+ D x
+ Kx = felec − fext

∂  1 πε o (2 x + l)
2 
Mx


− λ  + Kx + D x
= − fext
∂ x  2 ln( a b) 
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

9
Example: In the electric field transducer shown, The electrical part of the system will be analyzed
the area of one plate is A. When the spring is using nodal analysis. The generalized coordinates
relaxed, the separation between the plates is Go. are λ and x.
Write the dynamic equations that describe the Coenergy State Function
system.
T ' = K.E.'+W' fe (λ
, x)
K
M
K.E.' = 12 M x
2
+
Go  ε o A 
2
x λ
R i(t) W'fe (λ
, x) = 1
2
C(x)e2 = 1
2  λ
-  Go − x 
We get
The capacitance is
 ε o A 
2
εoA T' = 1
M x
2 + 1
 λ
C(x) = 2 2
 Go − x 
Go − x
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Energy State Function Lagrangian Function

U = P.E. + Wfm (λ, x) L = T '−U

P.E. = 12 Kx2  ε o A 
2
L = 1
2
M x
2 + 1
2  λ −
1
2
Kx 2
Note: The spring is stretched by a length x.  Go − x 
Lagrange Equation - Electrical System
W fm ( λ , x ) = 0 since there is no inductor
in the system. d  ∂L  ∂L ∂D
We get 2
λ:   − + = Pλ
U = Kx 1

dt  ∂λ  ∂λ ∂λ

2

Rayleigh’s Dissipation Function d  εoA  λ

λ
2  λ
 − 0 + = i(t)
D ( λ
, x
) = 1 dt  G o − x  R
2
R
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

10
  ε o A 


 ε o A  λ
εo A
  λ + λ  2 
x

+ = i(t) Mx


− 1
λ
2 + Kx = Mg
 Go − x   (G o − x)  R 2
(G o − x)2

which can be rewritten as

which can be written as
dv dC (x) dx v
C(x) +v + = i(t) εoA
dt dx dt R Mx


+ Kx = 1
2 2
λ
2 + Mg
(G o − x)
Lagrange Equation - Mechanical System
or
d  ∂L  ∂L ∂D
x:   − + = Px Mx


+ Kx = felec + Mg
dt  ∂x  ∂x

∂x

 


− ∂  12 ε o A λ
2  + Kx = Mg
Mx Note: The term Mg represents the weight of the
∂x  G o − x  movable plate.

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Example: In the electric field transducer shown, The electrical part of the system will be analyzed
the area of one plate is A. When the spring is using loop analysis. The generalized coordinates
relaxed, the separation between the plates is Go. are q and x.
Write the dynamic equations that describe the Coenergy State Function
system.
T ' = K.E.+
' W'fm (q

, x)
K R

M +
K.E. ' = 1
2
M x
2
Go
x q
v(t) W' fm = 1
2
Li 2 = 1
2
L q
2
-
Note: We have a W’fm because of the inductor.
The capacitance is L We get
εo A
C(x) = T' = 1
2
M x
2 + 12 Lq
2
Go − x
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

11
Energy State Function Lagrangian Function

U = P.E. + Wfe (q, x ) L = T '−U

 Go − x  2
P.E. = 12 Kx2 L = 1
2
M x
2 + 1
2
Lq
2 − 12 Kx 2 − 1
2  q
1 q2 1
Go − x  2  εo A 
W fe ( q, x ) = 2
= 2  q
C(x )  εoA  Lagrange Equation - Electrical System
We get
Go − x  2 d  ∂L  ∂L ∂D
U = 1
2
Kx 2 + 1
2  q q:   − + = Pq
 εoA  dt  ∂ q
 ∂q ∂q

Rayleigh’s Dissipation Function ∂  1 Go − x 2 
2
Lq


− − 2
q  + R q
= v(t)
D (q
, x
) = 1
2
R q
∂q  εoA 

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Go − x 1 1 2
Lq


+ q + R q
= v(t) Mx


− 2
q + Kx = Mg
εoA εo A
which can be rewritten as
which can be written as
di 1
dt C(x) ∫
L + idt + Ri = v(t) 1 2
1
Mx


+ Kx = 2
q + Mg
Lagrange Equation - Mechanical System εoA
d  ∂L  ∂L ∂D or
x:   − + = Px
dt  ∂x
 ∂x ∂x
Mx


+ Kx = felec + Mg
 


− ∂  − 12 G o − x q2  + Kx = Mg
Mx
∂x  εoA 

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

12
Example: The iron may be assumed to be The electrical part of the system will be analyzed
infinitely permeable. Flux fringing and friction are using loop analysis. The generalized coordinates
negligible. The cross-sectional area of the gap is are q and x.
A. When the coil current is zero, the length of the Coenergy State Function
air gap is Go.
Coil Inductance K T ' = K.E.'+W'fm (q

, x)
Go − x frictionless (no gap) K.E. ' = 1
M x
2
Rg = M
2
µoA R N2 µ o A
2
+ W' fm = 1
2
L (x )i2 = 1
2
q
Ni Niµo A Go q

Go − x
x v(t)
φ= = We get
Rg Go − x -
N2µ o A
2
N φ N2µ o A C T' = 1
Mx
2 + 1
q
L (x) = = 2 2
Go − x
i Go − x
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Energy State Function Lagrangian Function

U = P.E. + Wfe (q, x ) L = T '−U

N2 µ o A
2 q2
P.E. = Kx 1
2
2
L = 1
2
M x
2 + 1
2
q − 1
2
2
Kx − 1
2
Go − x C
q2
W fe ( q) = 12
C Lagrange Equation - Electrical System
We get d  ∂L  ∂L ∂D
q2 q: − + = Pq
U = 1
2
Kx 2 + 1
2  
dt  ∂ q  ∂q

∂q

C
Rayleigh’s Dissipation Function d  N2 µ o A  q
 q
 + + R q
= v(t)
D (q
) = 1
2
R q
2 dt  Go − x  C

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

13
  N2 µ o A   2
 N2 µ o A
2
 q

+
 N µ o A 2  x
+ q + R q
= v(t)
q Mx


− 1
q + Kx = 0
2
 Go − x   (Go − x)  C (G o − x)2
which can be rewritten as
which can be written as
di dL (x) dx 1
L (x ) +i + ∫ idt + Ri = v(t) N2 µ o A
2
dt dx dt C Mx


+ Kx = 1
2
q
Lagrange Equation - Mechanical System (G o − x )2
d  ∂L  ∂L ∂D or
x:   − + = Px
dt  ∂x  ∂x

∂x
Mx


+ Kx = fmag
∂  1 N2 µ o A
2 
Mx


−  q  + Kx = 0
∂x  2 G o − x 
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Linearization Consider the transducer shown. The inputs to the

For incremental-motion transducers, the dynamic electrical and mechanical systems consist of a
equations may be linearized. constant part plus an incremental component.
X o + ∆x
1. If the incremental changes in the variables are
small, we can assume a linear region of operation. Io + ∆i
•
2. The equilibrium position must be stable; i.e. the Vo + ∆v
moving element should return to the equilibrium •
position after a small disturbance. o F + ∆f
Electrical
Under these conditions, the differential equations System Mechanical
describing the system can be linearized. The System
resulting equations will involve the incremental The dynamic equations for this system may be
variables. linearized.
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

14
Example: The transducer shown represents a The equivalent circuit is shown below. The
telephone receiver. The input signal ∆v results in a generalized coordinates are q and x.
changing current in the coil, which causes the
diaphragm to vibrate. The elasticity of the x
Let L(x) be the
diaphragm is represented by K. Air friction is inductance of
represented by D. the coil. R
1 2 +
K.E. ' = 2
M x
K q


Sound M ∆v
R ∆f
Waves W' fm = 1
2
L (x )q
2 -
+ D - +
i ∆v Vo
-
Diaphragm - + Coenergy State Function
Vo
T' = 1
2
Mx
2 + 12 L(x)q
2
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Energy State Function Lagrange Equation - Electrical System

P.E. = 12 Kx2
d  ∂L  ∂L ∂D
W fe ( q) = 0 q:   − + = Pq
dt  ∂q
 ∂q ∂q

We get
U = 1
2
Kx 2


+ q
dL (x) x
+ R q
= Vo + ∆ v
L(x)q
dx
Rayleigh’s Dissipation Function
Lagrange Equation - Mechanical System
D(q
, x
) = 1
2
Rq
2 + 12 Dx
2 d  ∂L  ∂L ∂D
x:   − + = Px
dt  ∂x  ∂x

∂x

Lagrangian Function
dL (x)
2
L = 1
2
M x
2 + 1
2
L(x)q
2 − 1
2
Kx 2 Mx


− 1
2
q + D x
+ Kx = ∆ f
dx
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

15
Summary of Lagrange Equations First-order Taylor series approximation
di dL(x) dx dL
L(x)+i + Ri = Vo + ∆v L (Xo + ∆x) ≈ L (Xo ) + (Xo )∆x
dt dx dt dx
d2 x dL (x) dx dL 2

M 2 − 12 i2 +D + Kx = ∆ f (X o + ∆x ) ≈ dL (X o ) + d L2 (X o )∆x
dt dx dt dx dx dx
Assume x = Xo + ∆x Substitution gives
i = Io + ∆i  dL  •
For the electrical system, we get L(Xo ) + (Xo )∆x∆ i+ R(Io + ∆i)
 dx 
d
L (Xo + ∆x) (Io + ∆i) + R(Io + ∆i)
dt  dL d2L  •
dL + (Io + ∆i) (Xo ) + 2 (Xo )∆x∆ x = Vo + ∆v
+ (Io + ∆i) (Xo + ∆x) d (Xo + ∆x) = Vo + ∆v  dx dx 
dx dt
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Neglecting products of incremental terms, we get Simplify further

dL M ∆x


+ D∆x
+ K(X o + ∆x)
L(Xo )∆
i + R(Io + ∆i) + Io (Xo )∆x
= Vo + ∆v
dx
 2 dL 2
2 dL dL 
Similarly for the mechanical system, we get − 12 Io (Xo ) + Io 2
(Xo )∆x + 2Io (Xo )∆i = ∆f
d2  dx dx dx 
M 2
(Xo + ∆x ) + D d (Xo + ∆x )
dt dt
1 2 dL
Finally, we get
+ K(Xo + ∆x) − (Io + ∆i) (Xo + ∆x) = ∆f 2 dL
2 dx M∆x


+ D∆x
+ K(Xo + ∆x) − 12 Io (Xo )
or dx
M ∆x

+ D∆x
+ K(X o + ∆x)
2 d2L dL
2  dL d2L  − 1
2
Io 2
(X o )∆x − Io (X o )∆i = ∆f
− 12 (Io + 2Io ∆i) (X o ) + 2
(X o )∆x  = ∆f dx dx
 dx dx 
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

16
If ∆v and ∆f are both zero, the system will be at We get the resulting linear differential equations
the quiescent point with
dL
i = I o and x = Xo L(Xo )∆
i + R∆i + Io (Xo )∆x
= ∆v
dx
The equation for the electrical system and
dL
L(Xo )∆
i + R(Io + ∆i) + Io (Xo )∆x
= Vo + ∆v M∆x


+ D∆x
+ K∆x
dx
reduces to
1 2 d2L dL
RI o = Vo − 2
Io 2
(X o )∆x − Io (X o )∆i = ∆f
dx dx
Similarly for the mechanical system, we get
1 2 dL
KXo − Io (Xo ) = 0 Note: Given ∆v and ∆f, these linear differential
2 dx equations can be solved for ∆i and ∆x.

Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Example: The figure below shows a condenser The equivalent circuit is shown x
microphone. When ∆f = 0 and E=0, the separation below. The generalized
between the plates is G. When ∆f = 0 and the coordinates are q and x.
G
voltage E is applied, the separation between the If the area of one plate is A,
M K
plates is g. the capacitance is
Insulator εoA ∆f
∆f
C(x ) =
G−x
D
Air Escape Holes Diaphragm K.E. ' = 1
2
M x
2
R W'fm = 0 q


R
E Coenergy State Function
E
Output T' = 1
2
M x
2 + er −
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

17
Energy State Function Lagrange Equation - Electrical System
1 2
P.E. = Kx 2 d  ∂L  ∂L ∂D
q2 G − x 2 q:   − + = Pq
W fe ( q, x ) = 12 = 12 q dt  ∂q
 ∂q ∂q

C(x ) εoA
We get G−x
G − x 2 R q
+ q=E
U = 12 Kx 2 + 12 q εoA
εoA
Rayleigh’s Dissipation Function Lagrange Equation - Mechanical System

D (q
, x
) = 1
R q
2 + 1
D x
2 d  ∂ L  ∂L ∂D
2 2 x:   − + = Px
dt  ∂x
 ∂x ∂x

Lagrangian Function
L = 1
M x
2 − 1
Kx 2 − 1 G−x 2
q

− 12 1 q2 + D x
+ Kx = ∆ f
Mx
2 2 2
εoA εoA
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

Let x = Xo + ∆x and q = Qo + ∆q. For the electrical At the equilibrium point,

system, we get
Qo G − Xo
G − Xo − ∆x E= = Qo
R∆q
+ (Qo + ∆q) = E Co εo A
εo A
1 2
felec = KX o = Qo

+ G − Xo Qo + G − Xo ∆q − Qo ∆x = E
R∆q 2ε o A
εo A εo A εo A
The linearized equations are
For the mechanical system, we get
1
+ G − Xo ∆q − Qo ∆x = 0
R∆q
M∆x


− 1
2
(Qo + ∆q)2 + D∆x
+ K(Xo + ∆x) = ∆f εo A εo A
εo A


+ D∆x
+ K∆x − Qo ∆q = ∆f
2


+ D∆x
+ K∆x − Qo ∆q = ∆f + Qo − KXo
M∆x M∆x
εo A 2εo A εo A
Department of Electrical and Electronics Engineering Department of Electrical and Electronics Engineering

18