ECE 463/ECE 663

Spring 2024

Mehrdad Kazerani
Electrical & Computer Engineering
University of Waterloo
©Mehrdad Kazerani, 2024
 LECTURE 7-1
Non-Isolated DC/DC Converter
Topologies and Control Techniques

DC/DC Converter Modeling (Cont.)

4-May-24 2
Boost Converter Equation-Based Average Modeling-1
Continuous-Conduction Mode of Operation

 Topological Mode I: Switch ON

 diL 1
 dt = L Vd
 (I )
 dvC = − 1 v
 dt RC
C

 This system of first-order differential equations is valid during

ON-state of the switch, ton.

4-May-24 3
Boost Converter Equation-Based Average Modeling-2

 Topological Mode II: Switch OFF

 diL 1 1
 dt = − L vC + L Vd
 ( II )
 dvC = 1 i − 1 v
 dt C
L
RC
C

 This system of first-order differential equations is valid during

OFF-state of the switch, toff.

4-May-24 4
Boost Converter Equation-Based Average Modeling-3
diL di di
, ave. = d  L (on − state) + (1 − d )  L (off − state)
dt dt dt
1 1 1
= d  ( Vd ) + (1 − d )  (− vC + Vd )
L L L
1− d 1
=− vC + Vd
L L
dvC  dv   dv 
, ave. = d   C (on − state)  + (1 − d )   C (off − state) 
dt  dt   dt 
1 1 1
= d  (− vC ) + (1 − d )  ( iL − vC )
RC C RC
1− d 1
= iL − vC
C RC

4-May-24 5
Boost Converter Equation-Based Average Modeling-4
 For simplicity of results, let
 diL diL
 dt = dt , ave.
 dv dv
 C = C , ave.
 dt dt
 diL 1− d 1
 dt = − L vC + L Vd
Therefore,


 dvC = 1 − d i − 1 v
 dt C
L
RC
C

This is a system of first-order nonlinear differential equations of the

general form:
d x
 = f ( x, u )  iL  Vd 
 dt where x= ,u = 
 y = g ( x, u ) vC  d 

4-May-24 6
Boost Converter Equation-Based Average Modeling-5
 Let’s linearize the system of nonlinear differential equations
around a steady-state operating point. The perturbed system is
represented by:

d f f
x = x + u
dt x * u *
f f
where and are the Jacobean matrices evaluated at
x *
u *

the steady-state operating point and are given as:

 f1 f1   f1 f1 
 f  iL vC   f  Vd d 
=  , = 
 x  f 2 f 2  u  f 2 f 2 
 iL vC   Vd d 

4-May-24 7
Boost Converter Equation-Based Average Modeling-6
 The elements of the Jacobean matrices are as follows:

f1 f1 1
=0 =
iL Vd L
f1 1− d f1 1
=− = vC
vC L d L
f 2 1 − d f 2
= =0
iL C Vd
f 2 1 f 2 1
=− = − iL
vC RC d C

4-May-24 8
Boost Converter Equation-Based Average Modeling-7
Steady-State Operating Point
 At steady-state, one can write:
 diL  1− d* * 1 *
 dt = 0  −
L
VC + Vd = 0
L
  
 dv 1 − d *
1
C
=0 I L* − * VC* = 0
 dt  C RC
 Given Vd , VC and R * , d * and I L can be found as follows:
* * *

 V *
V *

 1− d* * 1 *  1 − d * = d*  d * = 1 − d*
 − VC + Vd = 0  VC VC
L L
  * VC* VC* VC*2
1 − d I * − 1 V * = 0  I L = R(1 − d * ) = V *  I L = RV *
* *

 C L R*C C  R( d* ) d
 VC

Note that the steady-state operating point could have been obtained by looking at
steady-state equations relating average output to input voltage (VC = Vd (1 − d ) ),
and average output power to average input power ( Pi = Po  I L = VC2 / ( RVd )).

4-May-24 9
Boost Converter Equation-Based Average Modeling-8
 The state-space representation of the linearized system becomes:

 1− d*  1 1 * 
 −
L   iL   L
0 VC 
d  i  L  Vd 

L
 =  v  +  
dt  vC  1 − d * 1  C  1 *   d 
 C −  0 − IL
RC  
 C 
or:

 Vd*  1 1 * 
 0 −   VC  V 

d  L 
i LVC*   iL   L L
  =   + 2  d 
dt vC   Vd
*
1  vC   0 VC*   d 
− −
 CV * RC   RCVd* 
2
 C 

4-May-24 10
Boost Converter Equation-Based Average Modeling-9
 The general form for the linearized system is:
d x
 = Ax + Bu
 dt
 y = C x + Du
where 
 Vd*  1 1 * 
 0 − * L VC 
 iL   LV C  L
x=  ; A= ;B= ;
 vC  V 1  0 − C * 
*2
* V
 d
−  
 CVC R*C   RCVd 
*

 V  0 1  0 0 
u =  d;C =   ; and D = 0 0
  d   1 0   

 With the duty ratio-to-output voltage transfer function in mind,

the output of interest is vC and the input of interest is d .

4-May-24 11
Boost Converter Equation-Based Average Modeling-10
Duty Ratio to Output Voltage Transfer Function
 The goal is now to find the desired control input-to-output
transfer function of the converter. From linear control theory,
Transfer Function H ( s) = C ( sI − A) −1 B + D
 The desired control input-to-output transfer ratio vC / d can be
found in the following way.
 1 * 
 L VC 
B' =   , C
'
=  0 1 , and D
'
= 0,
− C 
*2
V
 RCV * 
 d 
−1
 Vd*   1 * 
 s *   VC 
v LVC  L
H ' ( s ) = C ( s ) =  0 1   
d  V* 1   − VC 
*2

− d * s+ *   *
 CVC R C   RCVd 

4-May-24 12
Boost Converter Equation-Based Average Modeling-11
a b 
 Inverse of a 22 Matrix: If M =   ,
c d 
then,
−1 1  d −b 
M =   , where det .( M ) = a  d − b  c
det .( M )  −c a 
Therefore,
−1
 V *
  1 Vd* 
 s
d
  s + − 
 LVC*  1  R *
C LVC* 
=
 V* 1  1 Vd*2  Vd* 
− d * s+ *  s +
2
s+ *2  s 
 CVC R C  C  
RC LCV *
 CVC
Vd* VC*2
− s
v LC RCVd *

and H ' ( s) = C ( s) =
d 1 Vd*2
s + * s+
2

RC LCVC*2

4-May-24 13
 Boost Converter Equation-Based Average Modeling-12
Vd* VC*2
− s
v LC RCVd *
H ' ( s) = C ( s) =
d 1 Vd*2
s + * s+
2

RC LCVC*2
 This transfer function shows that boost converter, in continuous mode
of operation, has one right-half-plane zero and 2 left-half-plane poles.
 R*Vd*2
 sZ = *2
 0  RHP zero!  non-minimal phase System!Stability problem.
 LVC
 1 1 2 Vd*2
 − *  ( * ) −4 *2
s , s = R C R C LCV C
 P1 P 2 2
1 Vd*2
Note that in the denominator of the transfer function, s + * s +
2
*2
,
RC LCVC
1 Vd*2
 0 and  0  sP1 & sP 2  0  2 LHP poles
R*C LCVC*2
Note: There are a few ways to get around the RHP zero issue: (i) cancel the effect of RHP zero through
proper controller design; (ii) implement input (inductor) current control rather than output (capacitor)
voltage control; (iii) implement a nested-loop control structure, with inductor control as the inner loop and
capacitor voltage control as the outer loop; (iv) operate to converter in discontinuous conduction mode.
4-May-24 14
 Boost Converter Modeling, Continuous-13
Duty Ratio to Input Current Transfer Function
 With the duty ratio-to-input current transfer function in mind, the output of
interest is iL and the input of interest is d .
 The desired control input-to-output transfer ratio iL / d can be found as follows.
 1 * 
 L VC 
B' =   , C
'
= 1 0  , and D
'
= 0,
 VC 
*2

 − RCV * 
 d 
−1
 Vd*   1 * 
 s *   VC 
i LVC  L
H ' ( s ) = L ( s ) = 1 0   
d  Vd *
1   − VC 
*2

 − s + *   RCV * 
 CV C
*
R C   d 

 1 Vd*   1 *  * *
 s + *
− * V C 
2V C
+
VC
s
1  R C LV  L
= 1 0 = RLC L
C

1 Vd *2
 V *   VC  *2
1 Vd*2
s + s+ s  − s + * s+
2 2
*2 
d
*
RC LCVC  CVC *
  RCV d  RC LCVC*2

4-May-24 15
 Boost Converter Equation-Based Average Modeling-14
 The resulting transfer function therefore is
2VC* VC*
+ s
iL R *
LC L
H (s) =
'
(s) =
d 1 Vd*2
s + * s+
2

RC LCVC*2
 This shows that the duty ratio-to-input current transfer function
transfer function of boost converter, in continuous mode of
operation, has one left-half-plane zero and 2 left-half-plane poles.
 2
 Z s = −  0  LHP zero!  minimal phase System! No stability problem.
R*C

 1 1 2 Vd*2
 − *  ( * ) −4 *2
 sP1 , sP 2 = R C R C LCV C
 2
1 Vd*2
Note that in the denominator of the transfer function, s + * s +
2
,
RC LCVC*2
1 Vd*2
*
 0 and *2
 0  sP1 & sP 2  0  2 LHP poles
RC LCVC
Note: This approach is commonly used for the control of power flow through a boost converter. For
example, this scheme can be used for the control of power supplied by a fuel cell stack in a fuel cell
vehicle, where a boost converter is placed between the fuel cell stack and vehicle’s high-voltage dc bus.
4-May-24 16
Boost Converter Equation-Based Average Modeling-15
Discontinuous Mode of Operation

iL

S: ON S: OFF
dTs (1-d)Ts

dTs d1Ts d2Ts

4-May-24 17
Boost Converter Equation-Based Average Modeling-16

 Topological Mode I: Switch ON

 diL 1
 dt = L Vd
 (I )
 dvC = − 1 v
 dt RC
C

 This system of first-order differential equations is valid during

ON-state of the switch, ton.

4-May-24 18
Boost Converter Equation-Based Average Modeling-17

 Topological Mode II: Switch OFF, Period d1Ts

 diL 1 1
 dt = − L vC + L Vd
 ( II )
 dvC = 1 i − 1 v
 dt C
L
RC
C

 This system of first-order differential equations is valid during

d1Ts.

4-May-24 19
Boost Converter Equation-Based Average Modeling-18

 d1 has to be found in terms of d. IL,peak

iL
di V
I L , peak =(during d Ts )  d Ts = d  d Ts
dt L
di Vd − vC 0 − I L , peak S: ON S: OFF
(during d1Ts ) = = dTs (1-d)Ts
dt L d1Ts
Vd d Ts V d
=− =− d dTs d1Ts d2Ts
L d1 Ts L d1
− LVd d − Vd
 d1 = = d
L (Vd − vC ) (Vd − vC )

4-May-24 20
Boost Converter Equation-Based Average Modeling-19

 Topological Mode III: Switch OFF, Period d2Ts

 diL
 =0
dt
 ( III )
 dvC = − 1 v
 dt RC
C

 This system of first-order differential equations is valid during

d2Ts.

4-May-24 21
Boost Converter Equation-Based Average Modeling-20

IL,peak
 d2 has to be found in terms of d. iL

Vd
d 2 = 1 − (d + d1 ) = 1 − (d − d) S: ON S: OFF
Vd − vC dTs (1-d)Ts
Vd
= 1 − (1 − )d
Vd − vC dTs d1Ts d2Ts
vC
 d2 = 1 + d
Vd − vC

4-May-24 22
Boost Converter Equation-Based Average Modeling-21
diL di di di
, ave. = d  L (on − state) + d1  L (during d1Ts ) + d 2  L (during d 2Ts )
dt dt dt dt
1 1 1
= d  ( Vd ) + d1  (− vC + Vd ) + d 2  0
L L L

dvC  dv   dv   dv 
, ave. = d   C (on − state)  + d1   C (during d1Ts )  + d 2   C (during d 2Ts ) 
dt  dt   dt   dt 
1 1 1 1
= d  (− vC ) + d1  ( iL − vC ) + d 2  (− vC )
RC C RC RC

 For simplicity of results, let

 diL diL
 dt = dt , ave.
 dv dv
 C = C , ave.
 dt dt

4-May-24 23
Boost Converter Equation-Based Average Modeling-22
 After replacing d1 and d2 with their expressions in terms of d, one
gets:
 diL 1 1 1 Vd
=
 dt L d V d + ( − v + V )( − )d
Vd − vC
C d
 L L

 dvC = − 1 v d + ( 1 i − 1 v )(− Vd )d + (− 1 v )(1 + vC d )
 dt RC
C
C
L
RC
C
Vd − vC RC
C
Vd − vC
 diL 1 1 vCVd 1 Vd 2
 = Vd d + d− d
 dt L L Vd − vC L Vd − vC
 2
 dvC = − 1 v d − 1 i Vd d + 1 vCVd d − 1 v − 1 vC d
 dt RC
C
C Vd − vC
L
RC Vd − vC RC
C
RC Vd − vC

4-May-24 24
Boost Converter Equation-Based Average Modeling-23
 diL 1  vC Vd 
 = V d  1 + − 
− −
d
 dt L  V v V vC 

d C d

 dvC 1  Vd vC  1 Vd
 dt = − vC  d − d + 1 + d  − i d
Vd − vC Vd − vC  C Vd − vC
L
 RC 
 Vd vC   vC Vd 
Note that :  d − d +1+ d  = 1 and 1 + − =0
 Vd − vC Vd − vC   Vd − vC Vd − vC 
 diL
 dt = 0

 dvC = − 1 vC − 1 iL Vd d
 dt RC C Vd − vC
This is a system of first-order nonlinear differential equations of the
general form:
d x
 = f ( x, u )  iL  Vd 
 dt where x= ,u = 
 y = g ( x, u ) vC  d 

4-May-24 25
Boost Converter Equation-Based Average Modeling-24
 Let’s linearize the system of nonlinear differential equations
around a steady-state operating point. The perturbed system is
represented by:
d f f
x = x + u
dt x * u *

f f
where and are the Jacobean matrices evaluated at
x *
u *

the steady-state operating point and are given as:

 f1 f1   f1 f1 
 f  iL vC   f  Vd d 
=  , = 
 x  f 2 f 2  u  f 2 f 2 
 iL vC   Vd d 

4-May-24 26
Boost Converter Equation-Based Average Modeling-25
 The elements of the Jacobean matrices are as follows:

f1 f1
=0 =0
iL Vd
f1 f1
=0 =0
vC d
f 2 1 Vd f 2 1 vC
=− d = iL d
iL C Vd − vC Vd C (Vd − vC ) 2

f 2 1 1 Vd f 2 1 Vd
=− − iL d = − iL
vC RC C (Vd − vC ) 2
d C Vd − vC

4-May-24 27
Boost Converter Equation-Based Average Modeling-26
Steady-State Operating Point
 At steady-state, one can write:
dvC 1 * 1 * Vd*
= 0  − * VC − I L * d *
=0
dt RC C (Vd − VC )
*

 Given Vd* , VC* and R * , d * and I L* can be found. Since there is

only one equation with two unknowns, one has to find
another relation. Note that I L* can be found by taking the average
of iL when d = d * , Vd = Vd* , and vC = VC* .
1 * I *peak
I peak (d + d1 )Ts
* * *
IL
iL

I L* = 2
Ts
But, S: ON S: OFF
* *
V V
I *peak = d d *Ts and d1* = − * d * d * dTs (1-d)Ts
L Vd − VC
dTs d1Ts d2Ts
Ts VC*Vd* *
I = −
*
d
2 L Vd* − VC*
L

4-May-24 28
Boost Converter Equation-Based Average Modeling-27
dvC 1 * 1 * Vd*
= 0  − * vC − I L * * d * = 0
dt RC C (Vd − vC )
1 *
− *
vC
d = * RC
1 * Vd*
IL * *
C (Vd − vC )
Substituting I L* in the above equation,
1 1
− * vC* *
d =
* R C = R
* * *
1 Ts vCVd Vd Ts Vd* 2 *2
− d *2
( * *) d
C 2 L Vd − vC
* *
(Vd − vC ) 2 L Vd − vC
* *

1
R * 2 L Vd* − vC* 2
d =
*3
*
 d *
= 3
*
( *
)
Ts V R Ts Vd
( * d * )2
2 L Vd − vC

4-May-24 29
Boost Converter Equation-Based Average Modeling-28
 The state-space representation of the linearized system becomes:

  0 0 
 d  iL     iL 
 dt  v  =  − 1 Vd d * − 1 − 1 I * +
* *
Vd *  
 C d  vC 
  C Vd − VC
* *
RC C (Vd − VC )
L * * 2




  0 0 
  * *   Vd 
  1 I* VC 1 * Vd   
d *
− I   d 
  C L (Vd* − VC* ) 2 C Vd* − VC* 
L


  iL 
y =  0 1  v 
  C

4-May-24 30
Boost Converter Equation-Based Average Modeling-29
 From linear control theory,
Transfer Function H ( s) = C ( sI − A) −1 B + D
 Using
 0 
B ' =  1 * Vd*  , C =  0 1 , and D = 0,
' '

− I
 C L Vd* − VC* 
−1
 s 0   0 

H ' ( s) = C ( s ) =  0 1  1 Vd*   
v
1 1 * Vd* *  1 * Vd* 
d d s + * + IL *
*
d − IL
 C Vd* − VC* R C C (Vd − VC* )2   C Vd* − VC* 
 1 1 * Vd* 
s + +
 R*C C (V * − V * ) 2
I L d *
0 
 d C 
 1 Vd* 
 − d *
s  0 
*
− *
  
=  0 1 
C V d V C
 1 * Vd* 
 1 1 V *
 − I
s 2 +  * + I L* * d * 2 d *  s  C L Vd* − VC* 
 R C C (Vd − VC ) 
1 * Vd* 1 * Vd*
− IL * s − IL *
C Vd − VC* C Vd − VC*
= =
 1 1 V *
  1 1 * Vd* *
s +  * + IL *
2 * d
d *
 s s + +
 R*C C L (V * − V * )2 
I d
 R C C (Vd − VC )
* 2
  d C 
4-May-24 31
Boost Converter Equation-Based Average Modeling-30
1 * Vd*
IL *
vC C VC − Vd*
=
d  1 1 * Vd* *
s +  * + IL * d 
 R C C (Vd − VC )
* 2

 The transfer function of the boost converter in the discontinuous
mode of operation shows that:
– There is no zero.  No stability problem duo to RHP zero.
– There is only 1 pole in the LHP.  system is stable.
 The average model of the boost converter shows that the converter is
stable in discontinuous mode of operation, while it is non-minimal
phase and has stability problems in the continuous mode of operation.
Operating in the discontinuous mode ensures stability. However, if
continuous mode of operation is desired, let’s say to lower the input
current ripple contents, the controller should stabilize the system.
 It was also seen that input current control in continuous mode of
operation does not result in any stability problems, as the duty ratio to
input current transfer function does not have a RHP zero.

4-May-24 32
 Boost Converter Circuit-Based Average Modeling-1
Note that the objective here is to derive the circuit-based average model, not the
equation-based average model of boost converter. As a result, the relations developed to
derive the average model will not include the first-order differential equations for the
inductor current and capacitor voltage, as these components are integrated with the
circuit-based average model for simulation. The model focuses on the switch voltage
and diode current (indicated by the dashed rectangle) only.

Given that S has two states (ON and OFF), two topological modes can be identified for
the boost converter.
Topological Mode 1: In this case, S is
ON, resulting in the following relations: iL L D iD io
vS = 0 + + +
iD = 0 vi
S
vS vo
− − −
Topological Mode 2: In this case, S is
OFF, resulting in the following relations:: Boost Converter Circuit Diagram
vS = vo
iD = iL

4-May-24 33
 Boost Converter Circuit-Based Average Modeling-2
By averaging the equations for vS and iD over one switching period, one can derive the
following relations, where d is the duty cycle of S, and all variables are average-per-
switching-period quantities:
vS = d ( 0 ) + (1 − d )( vo ) = (1 − d )( vo )
iD = d ( 0 ) + (1 − d )( iL ) = (1 − d )( iL )

L iL iD io

+ +
iL L D iD io
+ + + vi vo
vi vS vo
S
− − − − −

Boost Converter Circuit Diagram Boost Converter Average Model

Note that the circuit-based average model was derived based on the assumption of
operation in continuous conduction mode (CCM). To derive a circuit-based average
model valid for discontinuous conduction mode (DCM), a third sub-period must be
considered in each switching period during which inductor current stays at zero.

4-May-24 34