5. dc-dc converters: topologies, modeling, and...
TRANSCRIPT
-
1/220
5. DC-DC Converters: Topologies, Modeling, and Control
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
2/220
Basic Structure of a Switching Power Supply
PWM
OSC
COMP
REF
60 Hz
lineinput PFC converter
and filter
PWM Controller
Highfrequencyinverter
20-200 KHz DC
output
output rectifierand filter
LoadSource
120 Hz
Feedback Sensing and
IsolatorPFC
Controller
Input EMI filter
Output EMI filter
-
3/220
Why Topologies is Important in SPS Design?
Power ratings (voltage and current) are primary concerns in determination of SPS topologies Converters are classified as nonisolated and isolated converters DC-DC converters are basic of other converters Resonant and soft-switching converters are derivatives of their corresponding PWM
converters
AC/DC Battery Charger DC/DCDC/DC
DC
AC85
26
5V
PFC Controller PWM Controller
DC/DCController
SMPS AC/DC
BatteryCharger
DC/DCController
IC
DC/DC Converter
Power Conversion, Control, and Management
4/220
ac inputsupply
dc output
voltage
Transformers in SPS +
v1(t)
i1(t)+
v2(t)
i2(t)n1:n2
+
i3(t)
v3(t)
n3
Isolation Turns Ratio Provide Wide Range Output Multiple Outputs Transformer
Input rectificationand filtering
duty cyclecontrol
controlcircuitry
HighFrequency
switch
PowerTransformer
Output rectificationand filtering
Vref
mosfet orbipolar
TPWM
OSC
Signal Coupling
-
5/220
Objectives of Transformer
Isolation of input and output ground connections, to meet safety requirements
Extend the voltage conversion range Reduction of transformer size by incorporating high
frequency isolation transformer inside converter Minimization of current and voltage stresses when a large
step-up or step-down conversion ratio is needed use transformer turns ratio
Obtain multiple output voltages via multiple transformer secondary windings and multiple converter secondary circuits
Transformer isolation is required for all circuits operating at a dc input voltage of 60 V DC or more.
6/220
Transformers for Asymmetrical and Symmetrical Converters
Asymmetrical Converter Symmetrical Converter
2Bs
B
H
Bssymmetricalconverters
asymmetricalconverters
symmetricalconverters
forwardconverter
flybackconverter
AvailableFlux swing
A
BCDN2
N1
v1
i1
v2
i2
Ll2Ll1
Lm
A
B
C
D
-
7/220
Buck Derived Isolated DC-DC Converters with Synchronous Switch
Push-Pull Converter
Forward Converter Half-Bridge Converter
Full-Bridge Converter
S
L
C RSn:1
Vg
n:n
S
S1
S2n:1
S2
S1
Vg
S1
S2n:1
S2
S1
Vg
S1
S2
S1
S2
n:1
S1
S2
Vg
LC R
LC R
LC R
8/220
Buck-Boost Derived Isolated DC-DC Converter
Flyback Converter
n:1
S
Vg Lp Ls
S
R
-
9/220
Most Commonly Used Isolated Converters for Low-Power Applications
Flyback Converter
LOAD
Forward Converter
LOAD
An isolated dc-dc converter module (MS Kennedy Corporation)
10/220
Basic Topologies of PWM DC-DC Converters
Buck
Boost
Buck-Boost L C
D
vovi
L
C
D
vovi
vi vo
L
CD
One Inductor, One Capacitor
C,uk
L1
C2D
L2C1
L1
C2
D
L2
C1
SEPIC
Zeta L1 C2D
L2C1
SEPIC: Single-Ended Primary Inductor Converter
Two Inductors, Two Capacitors
vi
vi
vi
vo
vo
vo
-
11/220
Basic DC-DC Converters in CCM
Buck Converter
Boost Converter
Buck-Boost Converter
V
Buck converter
+
V
Boost converter
+
V
Buck-boost converter
0 0.2 0.4 0.6 0.8 10
0.20.40.60.8
1
M(D
)
D
0 0.2 0.4 0.6 0.8 10
M(D
)
D
0 0.2 0.4 0.6 0.8 1
-5
M(D
)
D
-4-3-2-10
+
V
Cuk converter
+
V
Sepic
0 0.2 0.4 0.6 0.8 1
-5
M(D
)
D
-4-3-2-10
0 0.2 0.4 0.6 0.8 10
M(D
)
D
12345
12345
Transform to Isolated Converters
Buck-Boost Converter
L C
D
vovi
L1
C2
D
L2
C1
vi vo
SEPIC Converter
L1
C2D
L2C1
vi vo
Cuk Converter
LOAD
Flyback Converter
Isolated SEPIC Converter
LOAD
L1 L2
V1 V2
V0
i1 i2
C1 C2
VinLOAD
Isolated Cuk Converter
-
13/220
Power Supply Topologies (sluw001a)
14/220
Power Supply Topologies (sluw001a)
-
15/220
Selection of Isolated SPS Topologies in Watts
Topologies
1 10 100 1000 10000Output power, watts
Com
plex
ity
Flyback
Resonant resetForward
Current fedPush-pull
Single transistorforward
Tow transistorforward
Half bridge
Dual interleavedForward converter
Full bridge
Phase shiftedFull bridge
Server applications
16/220
Modeling of Switching Converters
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
-
17/220
Modeling of Switching Converters
Why Modeling ? Classification of Modeling Techniques Modeling of Switching Power Converters State-Space Averaging Technique Transfer Functions Small-Signal Equivalent Circuit Model
18/220
Objective of Modeling
Objective of Modeling
Analysis
Simulation
DesignEfficiency
Output Impedance
Static Characteristics
-
19/220
Difficulties with Modeling of SPS
Nonsmooth Systems (time and state discontinuity)Nonlinearity due to operating point Concepts of existence, uniqueness, stability not clearly defined for systems with discontinuous right half-plan zero Inherent Nonlinear Dynamic Behavior! Concept of chaotic dynamics relatively new to power electronics
20/220
Working Profile of a Switching Converter
Power-on Power-off
ov
oi
Wat
ts
% C
PU
tim
e
Dell power edge 2400 (web/SQL server)
-
21/220
Operating Region State-Variable Plane
DCM
D = 1.0
0.1
0.3
0.5
0.7
0.9
CCM
80% 100%60%40% 110%90%10% 20%
Vo/Vi = 1.0
0.25
0.5
0.75
0
1.0
22/220
Operating Region Operating Point Operating Mode?
IN
OUT
VV
)(max,LB
o
II
0 0.5 1.0 1.5 2.0
0
0.25
0.50
0.75
1.0
DCM
D = 1.0
0.1
0.3
0.5
0.7
0.9
CCMCRM
VIN = constant
(min)
(max)
IN
OUT
VV
(max)
(min)
IN
OUT
VV
(max)OUTI(min)OUTI
-
23/220
Modeling of Switching Power Converters
Modeling of Voltage-Mode DC-DC Converters- Power Stage Modeling State-space average model (Middlebrook and C'uk 1977) Discrete time-domain model (Lee 1979) Equivalent circuit model (Chetty 1981) Unified topological model (Pietkiewicz and Tollik, 1987) PWM switch model (Voperian 1988) Injected-absorbed-current model (Kisovski 1991)
- Error Processor Modeling (Chetty 1982)- Pulsewidth Modulator Modeling Describing function model (Lee 1983) Equivalent circuit model (Bello 1981)
- Larger Signal Modeling (Vicua 1992) Modeling of Current-Mode DC-DC Converters
- Equivalent Circuit Model (Chetty 1981)- y-parameter Model (Middlebrook 1989)
24/220
Modeling Techniques
State Space Averaging Method[1] R. D. Middlebrook and S. Cuk, A general unified approach to modeling switching-converter
stages, IEEE PESC Conf. Rec., pp. 18-34, 1976.[2] S. Cuk and R. D. Middlebrook, A general unified approach to modeling switching DC-to-DC
converters in discontinuous conduction mode, IEEE PESC Conf. Rec., pp. 36-57, 1977.
Modeling of Switching Converters in DCM Operation[1] D. Maksimovic and S. Cuk, A unified analysis of PWM converters in discontinuous modes, IEEE
Trans. Power Electron., vol. 6, pp. 476490, May 1991. [2] J. Sun, D. M. Mitchell, M. F. Greuel, P. T. Krein, and R. M. Bass, Averaged modeling of PWM
converters operating in discontinuous conduction mode, IEEE Trans. Power Electron., vol. 16, pp. 482-492, July 2001.
-
25/220
Modeling Techniques ..
PWM Switch Method[1] V. Vorperian, Simplified Analysis of PWM Converters Using Model of PWM Switch Part I:
Continuous Conduction Mode, IEEE Trans. on Aero. and Elec. Sys., vol. 26, no. 3, pp. 490-496, May 1990.
[2] V. Vorperian, Simplified Analysis of PWM Converters Using Model of PWM Switch Part II: Discontinuous Conduction Mode, IEEE Trans. on Aero. and Electron. Sys., vol. 26, no. 3, pp. 497-505, May 1990.
Fast Analytical Techniques for Electrical and Electronic Circuits,V. Vorperian, Cambridge Press, 2004.
26/220
Modeling Techniques ..
Discrete Time-Domain Method[1] F. C. Lee, R. P. Iwens, Y. Yu, and J. E. Triner, Generalized computer-aided discrete time-domain
modeling and analysis of DC-DC converters, IEEE Trans. IECI, vol. 26, pp. 58-69, May 1979.
Equivalent Circuit Method[1] P.R.K. Chetty, Switch-Mode Power Supply Design, TAB BOOKS, Inc., 1986.
Modeling of Current-Programmed Converter[1] R. D. Middlebrook, Modeling current programmed Buck and Boost converters, IEEE Trans. on
Power Electronics, vol. 4, pp. 36-52, January 1989.
Unified Topological Method[1] Pietkiewicz, A. and D. Tollik, Unified topological modeling method of switching dc-dc converters
in duty-ratio programmed mode, IEEE Trans. on Power Electron., vol. 2, no. 3, pp. 218-226, July 1987.
Injected-Absorbed-Current Method[1] Kislovski, A. S., R. Redl, and N. O. Sokal, Dynamic Analysis of Switching-Mode DC/DC
Converters, Van Nostrand Reinhold, 1991.
-
27/220
Recommended Books: Modeling and Simulation
Dynamic Analysis of Switching-Mode DC/DC Converters,Andre'S. Kislovski, Richard Redl, and Nathan O. Sokal,Van Nostrand Reinhold, New York, 1991.
Complex Behavior of Switching Power Converters, Chi Kong Tse, CRC Press, 2004.
Switch-Mode Power Supply Simulation: Designing with SPICE 3Steven M. Sandler, McGraw-Hill Professional; 1 edition, Nov. 11, 2005.
Switch-Mode Power Supplies - SPICE Simulations and Practical Designs, Christophe Basso, McGraw-Hill, Feb. 1, 2008.
Modeling of DC-DC Converters
voltagereference
Pulse-widthmodulator
cv)(t sGc
refv
+ v
H
t
tsTsdT
tv
t
Complex Behavior of Switching Power Converters, Chi Kong Tse, CRC Press, 2004.
Dynamic Analysis of Switching-Mode DC/DC Converters, Andre'S. Kislovski, Richard Redl and Nathan O. Sokal, Van Nostrand Reinhold, New York, USA, 1991
SMPS Simulation with SPICE 3, Steven M. Sandler, McGraw-Hill Professional, Dec. 1, 1996.
Computer-Aided Analysis and Design of Switch-Mode Power Supplies, Yim-Shu Lee, Marcel Dekker, Inc., Feb. 23, 1993.
Switch-Mode Power Supplies -SPICE Simulations and Practical Designs, Christophe Basso, McGraw-Hill, Feb. 1, 2008. Fast Analytical Techniques for Electrical and Electronic Circuits,
V. Vorperian, Cambridge Press, 2004.
-
29/220
Modeling and Control of DC-DC Converters
Output voltage feedback only!
PWMModulator
LoopCompensator
vo
digital signal processor analog signal processor
vR
d
load
LR di~
Buck Converter Boost Converter
Buck/Boost Converter C,uk Converter
vi vo
L
CD
L
C
D
vovi
L C
D
vovi
L1
C2D
C1
vi vo
Switching power converters
oi
Define the Operating Point!
Define the Load Disturbance
vgosZ
sv~
sV
Define the Line
Disturbance
Define the Source Output Impedance!
Small-Signal Modeling of a Buck Converter
IN
OUT
VV
)(max,LB
o
II
0 0.5 1.0 1.5 2.0
0
0.25
0.50
0.75
1.0
DCM
D = 1.0
0.1
0.3
0.5
0.7
0.9
CCMCRM
Averaged Switch Modeling of Boundary Conduction Mode DC-to-DC ConvertersJ, Chen, R. Erickson, and D. Maksimovic, IECON 2001.
VIN = constant
V. Vorperian, "Simplified Analysis of PWM Converters Using Model of PWM Switch Part II: Discontinuous Conduction Mode," IEEE Trans. on Aero. and Electron. Sys., vol. 26, no. 3, pp. 497-505, May 1990.
V. Vorperian, "Simplified Analysis of PWM Converters Using Model of PWM Switch Part I: Continuous Conduction Mode," IEEE Trans. on Aero. and Elec. Sys., vol. 26, no. 3, pp. 490-496, May 1990.
-
31/220
Modeling of Switching Converters in DCM Operation
IN
OUT
VV
)(max,LB
o
II
0 0.5 1.0 1.5 2.0
0
0.25
0.50
0.75
1.0
DCM
D = 1.0
0.1
0.3
0.5
0.7
0.9
CCMCRM
VIN = constant
V. Vorperian, "Simplified Analysis of PWM Converters Using Model of PWM Switch Part II: Discontinuous Conduction Mode," IEEE Trans. on Aero. and Electron. Sys., vol. 26, no. 3, pp. 497-505, May 1990.
J. Sun, D. M. Mitchell, M. F. Greuel, P. T. Krein, and R. M. Bass, Averaged modeling of PWM converters operating in discontinuous conduction mode, IEEE Trans. Power Electron., vol. 16, pp. 482-492, July 2001.
D. Maksimovic and S. Cuk, A unified analysis of PWM converters in discontinuous modes, IEEE Trans. Power Electron., vol. 6, pp. 476490, May 1991.
Systematic View of DC-DC ConvertersEfficiency
Output Impedance
Selection of Switching Frequency amd PWM Strategies
Frequency Responses
Time Responses
Current Injection Testing
-
33/220
PWM DC-DC Power Conversion and Regulation
CLOCK RAMP
Zi
Zf
vref
vo
vx
D
vi
T
TOND T
TON
vpvx
34/220
Signal Composition of a PWM DC-DC Converters
comparator
+
D
d
Vc
v c v c
V g
v g v g
igI g
ig
clock ramp
analogamplifier
reference
load
Vv v
modulator-power-stagesubsystem
Ii i
-
35/220
Frequency Spectrum
f 2f 3f fs - f fs
outputspectrum
bandpassfilter
frequency
fs + f 2fs 3fs
36/220
AC and DC Quantities in a PWM Switching DC-DC Converter
I ii i vd ioioIo
vovoVo
RLC
L
Z f
vref
v c
V vI i
CLOCK RAMP
vd
Vc
vc vc
io vo
Z i
-
37/220
Small-Signal Modeling of a Switching Power Converter
iIi iIi
RLC
L
error processor
Zi
Z f
vref
vc
v V vi I i
duty cyclemodulator
power stage
d D d
open
oOo iIi v V vo O o
AveragedPower Stage
vi
oid
vo
vo
The concerned transfer fucntions under small perturbations can be measured under an open loop condition.
38/220
Definition of
v to( )
t
V I VO O I, ,
ioo viv ,,i to( )
v ti( )
small signal perturbation
t
TtD on
D
d(t)
d t( )
t
IOO VIV ,, Ttd x
power switch gating waveform
xt
Tont
d
-
39/220
BuckBoost
Buck/Boost
Small-Signal Modeling of Voltage-Mode DC-DC Converters
Power Stage
PulseModulator
ErrorProcessor
vr
vo
vc
io
d
vi
40/220
Modeling of Single-Loop DC-DC Converters
vo
vc
io
d
vi
; ;
vv
vi
vd
o
i
o
o
o
KPWM A(s)
( , , )v f v i do i o
vi
io
d vc
vo
Zp
Gd
-A(s)
Gv
KPWM
-
41/220
Small Signal Transfer Functions
G s vvv
o
i d io
( ) ,
0 0
: Open-loop input-to-output
G vdd
o
v ii o
,0 0
cvdk
PWM
: Open-loop output impedance
: Control to output transfer function
: PWM modulator gain
Z vip
o
o d vi
, 0 0
A s vv
c
o
( )
KVp
PWM 1 : Compensator gain: PWM dc gain
42/220
Modeling of PWM DC-DC Converters
A dc-to-dc switching regulator incorporating a three-port duty ratioprogrammed modulator-power-stage subsystem whose transfer functions aredefined in terms of ratios of small-signal ac quantities (hats) superimposedupon large-signal dc quantities (capitals).
The spectrum of the output signal contains the switching frequency, thecontrol frequency, their respective harmonics, and sidebands.
The modeling objective is to find, as function of frequency, the loop gain andthe closed properties of the regulator.
The essential prerequisite is to find the transfer function of the three-portsubsystem of the modulator-power-stage.
-
43/220
Concerned Transfer Functions
Control-to-output transfer function Line-to-output transfer function (audio susceptibility) Reference-to-output transfer function Input impedance Output impedance
A voltage sourcing power supply should have a low (zero) outputimpedance, while a current sourcing power supply should have a high(infinite) output impedance.
44/220
Closed-Loop Transfer Functions
Gv,CL (Closed-loop Audio Susceptibility)
v G v G do v i d
d AK vo PWMvv
GG K A
GT
o
i
v
d
v
1 1PWM
Zp,CL (Closed-loop Output Impedance)
v Z i G do p o d
d AK vo PWMvi
ZT
o
o
p1
Loop Gain : PWMT A s K Gd ( )
vo
vc
iod
vi
; ;
vv
vi
vd
o
i
o
o
o
KPWM A(s)
( , , )v f v i do i o
vi
io
d vc
vo
Zp
Gd
-A(s)
Gv
KPWM
-
45/220
State-Space Averaging Technique
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
46/220
State-Space Averaging Techniques
State Space Averaging Modeling Static Analysis Small-Signal Model at CCM Small-Signal Model at DCM Frequency Response Analysis
-
47/220
Concept of Time Averaging (Low Frequency Response Behavior)
I ii i v d ioioIo
v ov oV o
RLC
L
Z i
Z f
v ref
v c
V vI i
CLOCK RAMP
v d
V c
vc v c
io
v o
48/220
Linear Approximation of State Space Trajectories
x( )0
x( )dTs
x ( )Ts
x A x B u
x A x B u 1 1
dT s( )Ton
( )1 d Ts( )Toff
t
x( )t
x A x B u 2 2
-
49/220
State-Space Averaging Method (CCM)
uExCyuBxAx
uExCyuBxAx
22
22
11
11
During
During
OFF
ON
T
T
When the circuit time constant is far greater than the switching period, the above equations can be averaged as:
x d1
averaging by using state duty ratio weighting
x d2+
Switch-ON Period Switch-OFF Period
x A x B uy C x E u
2 2
2 2
x A x B uy C x E u
1 1
1 1
50/220
State-Space Averaging Method
( ) ( )( ) ( )
x A A x B B uy C C x E E u
1 1 2 2 1 1 2 2
1 1 2 2 1 1 2 2
d d d dd d d d
x Ax Buy Cx Du
A A AB B BC C CE E E
where =
1 1 2 2
1 1 2 2
1 1 2 2
1 1 2 2
d dd dd dd d
-
51/220
State Averaging by Active Duty Ratios
uExCyuBxAx
uExCyuBxAx
22
22
11
11
During
During
OFF
ON
T
T
When the circuit time constant is far greater than the switching period, the above equations can be averaged as:
d1
state averaging by using duty ratio weighting
d2
Switch-ON Period Switch-OFF Period
x A x B uy C x E u
2 2
2 2
x A x B uy C x E u
1 1
1 1
( ) ( )( ) ( )
x A A x B B uy C C x E E u
1 1 2 2 1 1 2 2
1 1 2 2 1 1 2 2
d d d dd d d d
x Ax Buy Cx Du
A A AB B BC C CE E E
where =
1 1 2 2
1 1 2 2
1 1 2 2
1 1 2 2
d dd dd dd d
52/220
Small-Signal Perturbation at a DC Operating Point
Substitute x X x y Y y u U u
; ; ; ; d D d d D d1 1 2 2
ddt
D d D d D d D d( ) ( ) ( ) ( ) ( ) ( ) ( )X x A A X x B B U u 1 1 2 2 1 1 2 2
termnonlinear ignore
2121
2121
22112211
0 termsdc
22112211
0 termsdc
)()(
)()(
)()(
)()(
uBBxAA
UBBXAA
uBBxAA
UBBXAAxX
dd
d
DDDD
DDDDdtd
dtd
i
Note:The nonlinear dynamic system is linearized around a selected operating point!
-
53/220
DC Model and AC Model
DC Model
X A BU1
Y ( CA B E)U1
where
ddt
d
d
x Ax Bu F
y Cx Eu G
AC ModelA A AB B BC C CE E EF A A X B B UG C C X E E U
1 1 2 2
1 1 2 2
1 1 2 2
1 1 2 2
1 2 1 2
1 2 1 2
D DD DD DD D
( ) ( )( ) ( )
54/220
Transfer Function Matrix
)()()(
)()(
)()()()(
)()(
2
1
2221
1211 sdsGsG
sisv
sGsGsGsG
sisv
d
d
d
i
uu
uu
i
o
)()()()()()()()()( 11
sdssssdssss
du HuHFAIBuAIx
y C A B E u C A F GG u G
( ) [ ( ) ] ( ) [ ( ) ] ( )( ) ( ) ( ) ( )
s sI s sI d ss s s d su d
1 1
G C A B Eu s sI( ) ( ) 1
G C A F Gd s sI( ) ( ) 1
H I A BH I A F
u
d
s ss s d s
( ) ( )( ) ( ) ( )
1
1
-
55/220
Open Loop Transfer Functions
control-to-output (open-loop transfer function): v sd s
G s s so d d( )( )
( ) ( ) ( ) 1 11
1G C I A F G
line-to-output (audio susceptibility):
output impedance:
v sv s
G s s soi
u u( )( )
( ) ( ) ( ) 11 111
11G C I A B E
v si s
G s s sod
u u( )( )
( ) ( ) ( ) 12 121
12G C I A B E
56/220
Example: Buck Converter
Step 1. Draw the linear equivalent circuit for each switching state of the converter.
L
CD
Q iL io
vc
icrc
rL
R vo id
iRu1 vi u2
iiy2
y1
x1
x2
Q closed, D open
L
C
iL io
vc
icrc
rL
vi R vo
id
iRii
D closed, Q open
L
C
iL io
vc
icrc
rL
R vo
id
iRii
-
57/220
Select State Variables
Select the inductor current iL and capacitor voltage vL state variables.
c
L
vi
xx
2
1x
i
o
iv
yy
2
1y
Select the input dc voltage and output disturbance current as input variables.
Select the output voltage and input current as output variables.
d
i
iv
uu
2
1u
58/220
Derive Circuit Equations: State Equations
Step 2. Write the circuit equations for each equivalent circuit in a state-variable format.
L
C
iL io
vc
icrc
rL
vi R vo
id
iRiiQ-ON and D-OFF State:
v L didt
i r r C dvdt
vi L L L C c C
r Cdvdt
v i i i R
i Cdvdt
i R
i R RCdvdt
i R
Cc
C L C d
Lc
d
Lc
d
( )
( )
( )R r Cdvdt
i R v i RC c L C d
dvdt
RR r C
iR r C
v RR r C
icC
LC
CC
d
( ) ( ) ( )1
-
59/220
State-Space Average Modeling: State Equations
dC
CiC
CLL
C
C
CdC
CC
C
CL
C
CLLi
L
irR
RrvvrR
RirrR
Rr
virR
RrvrR
rirR
RrrivdtdiL
)()()(
dC
CiC
CLL
C
CL irR
RrL
vL
vrR
RL
irrR
RrLdt
di
1111
Replace in with R
R ri
R rv R
R ri
CL
CC
Cd( ) ( ) ( )
1C dv
dtc
CdC
CC
LC
CLLL
i virRRv
rRi
rRRrri
dtdiLv
)()(
1)(
We can obtain:
60/220
State-Space Average Modeling: State Equations
i ii L
dC
CC
CL
C
C
ddC
CC
LC
L
dCL
dCLo
irR
RrvrR
RirR
Rr
RiiCrR
RvCrR
iCrR
RRCRi
RiRiRiRiiiv
)()()(
)()(1
)(
)(
v RrR r
i RR r
v RrR r
io CC
LC
CC
Cd
( ) ( ) ( )
d
iC
C
C
LCC
C
i
o
d
i
C
C
C
C
L
CC
CL
C
C
C
L
iv
rRRr
vi
rRR
rRRr
iv
iv
rRR
C
rRRr
LLvi
rRCrRR
C
rRR
Lr
rRRr
Lvi
00
0
01
10
11
111
11
Q-OFF and D-ONState Equation
-
61/220
State-Space Average Modeling: State Equations
Q-OFF and D-ON State:
0 Ldidt
i r r Cdvdt
vL L L C c C
RidtdvRCRi
RidtdvCi
RiiivdtdvCr
dc
L
dc
L
dCLCc
C
)(
)(
L
C
iL io
vc
icrc
rL
R vo
id
iRii
RivRidtdvCrR dCLcC )(
dC
CC
LC
c iCrR
RvCrR
iCrR
Rdtdv
)()(1
)(
62/220
State-Space Average Modeling: State Equations
dC
CiC
CLL
C
C
CdC
CC
C
CL
C
CLL
L
irR
RrvrR
RirrR
Rr
virR
RrvrR
rirR
RrridtdiL
)()()(
dC
CiC
CLL
C
CL irR
RrL
vrR
RL
irrR
RrLdt
di
111
( )R r Cdvdt
i R v i RC c L C d
dvdt
RR r C
iR r C
v RR r C
icC
LC
CC
d
( ) ( ) ( )1
Replace in with RR r
iR r
v RR r
iC
LC
CC
d( ) ( ) ( )
1C dv
dtc
CdC
CC
LC
CLLL vi
rRRv
rRi
rRRrri
dtdiL
)()(
1)(
0We can obtain:
-
63/220
State-Space Average Modeling: State Equations
ii 0
dC
CC
CL
C
C
ddC
CC
LC
L
dCL
dCLo
irR
RrvrR
RirR
Rr
RiiCrR
RvCrR
iCrR
RRCRi
RiRiRiRiiiv
)()()(
)()(1
)(
)(
v RrR r
i RR r
v RrR r
io CC
LC
CC
Cd
( ) ( ) ( )
d
iC
C
C
LCC
C
i
o
d
i
C
C
C
C
L
CC
CL
C
C
C
L
iv
rRRr
vi
rRR
rRRr
iv
iv
rRR
C
rRRr
Lvi
rRCrRR
C
rRR
Lr
rRRr
Lvi
00
0
00
10
10
111
11
Q-ON and D-OFFState Equation
64/220
State-Space Average Modeling: State Equations
d
iC
C
C
LCC
C
i
o
d
i
C
C
C
C
L
CC
CL
C
C
C
L
iv
rRRr
vi
rRR
rRRr
iv
iv
rRR
C
rRRr
LLvi
rRCrRR
C
rRR
Lr
rRRr
Lvi
00
0
01
10
11
111
11
Q Conducted:
D Conducted:
d
iC
C
C
LCC
C
i
o
d
i
C
C
C
C
L
CC
CL
C
C
C
L
iv
rRRr
vi
rRR
rRRr
iv
iv
rRR
C
rRRr
Lvi
rRCrRR
C
rRR
Lr
rRRr
Lvi
00
0
00
10
10
111
11
-
65/220
State-Space Average Modeling: State Equations
00
0 ,01
10
11
,111
11
11
11
C
C
CC
C
C
C
C
CC
CL
C
C
rRRr
rRR
rRRr
rRR
C
rRRr
LL
rRCrRR
C
rRR
Lr
rRRr
L
EC
BA
00
0 ,00
10
10 ,
111
11
22
22
C
C
CC
C
C
C
C
CC
CL
C
C
rRRr
rRR
rRRr
rRR
C
rRRr
L
rRCrRR
C
rRR
Lr
rRRr
L
EC
BA
Q Conducted:
D Conducted:
66/220
State-Space Average Modeling: Averaging
Step 3. Average each state by using duty ratio as a weighting factor and then combine thetwo sets of equations into a single set.
x A x B uy C x E u
1 1
1 1
x A x B uy C x E u
2 2
2 2
Q : D :x d1
averaging by using state duty ratio weighting
x d2+
( ) ( )( ) ( )
x A A x B B uy C C x E E u
1 1 2 2 1 1 2 2
1 1 2 2 1 1 2 2
d d d dd d d d
x Ax Buy Cx Eu
A A AB B BC C CE E E
where =
1 1 2 2
1 1 2 2
1 1 2 2
1 1 2 2
d dd dd dd d
-
67/220
State-Space Average Modeling: Perturbation
Step 4. Perturb the averaged equation set to produce DC and small signal terms and eliminatenonlinear product terms.
Substitute
into the averaged equations (in Steps 3)
x X x y Y y u U u
; ; ; ; d D d d D d1 1 2 2
( ) ( )( ) ( )
x A A x B B uy C C x E E u
1 1 2 2 1 1 2 2
1 1 2 2 1 1 2 2
d d d dd d d d
uBuBUBUB
uBuBUBUB
xAxAXAXA
xAxAXAXA
uUBB
xXAAxX
)))(()((
)))(()(()(
222222
111111
222222
111111
2211
2211
dDdD
dDdD
dDdD
dDdD
dDdD
dDdDdtd
68/220
State-Space Average Modeling: Perturbation
ddt
D D D D
D D D D
d
d d
( ) ( ) ( )
) ( )
) ( )
) ( )
X x A A X B B U
A A x B B u
A A X B B U
A A x B B u
1 1 2 2 1 1 2 2
1 1 2 2 1 1 2 2
1 2 1 2
1 2 1 2
(
(
(
Y y C C X x
E E U uC C X E E U
C C x C C X E E U
E E u C C x E E u
( ( ) ( ))( )
( ( ) ( ))( )( ) ( )
( ) ( ) ( )
( ) ( ) ( )
1 1 2 2
1 1 2 2
1 1 2 2 1 1 2 2
1 1 2 2 1 2 1 2
1 1 2 2 1 2 1 2
D d D d
D d D dD D D D
D D d
D D d d
Perturbation of the State Equations
x1
x1Operating Point New Coordinates
1x
2x
-
69/220
State-Space Average Modeling: DC Analysis
DC Analysis:
Set all variational terms to zero, we can obtain
X A BU1
Y ( CA B E)U1
0 1 1 2 2 1 1 2 2
( ) ( )A A X B B UAX BU
D D D D
Y C C X E E UCX EU
( ) ( )1 1 2 2 1 1 2 2D D D D
Therefore
70/220
State-Space Average Modeling: DC Model
AX BU 0Eliminate the DC term ddt
X 0 and Y CX EU
LS
CC
LCo
CC
LC
iCC
LCL
DII
VrR
RIrRV
VrR
IrR
R
DVVrR
RIrRr
CD
)//(
10
)//(0
..
We obtain the dc model equation:
Comments:1. DC model gives DC information (steady-state behavior). 2. DC model can be used for loss estimation.
-
71/220
State-Space Average Modeling: AC Model
ddt
d
d
x A x B u F
y C x E u G
Neglect the nonlinear product term d d x u and
where
A A AB B BC C CE E EF A A X B B UG C C X E E U
1 1 2 2
1 1 2 2
1 1 2 2
1 1 2 2
1 2 1 2
1 2 1 2
D DD DD DD D
( ) ( )( ) ( )
dIiDi
vrR
RirRv
vrRL
irR
RCdt
vd
dLVv
LDv
rRR
Li
LrRr
dtid
CA
LLs
CC
LCo
CC
LC
C
iiC
CL
CLL
)//(
111
1//
..
We obtain the ac model (small-signal) equation:
Comments:1. AC model gives small signal information.2. AC model parameter value depends on dc operating point.
72/220
Time-to-Frequency Transform
In the following, the hat above the variables will be neglected for simplicity. The small signal(or ac term) of a variable is denoted using lower case letter. Thus, the dynamic equation of aPWM dc-dc converter can be represented as:
Step 5. Transform ac state-space model into frequency domain (Laplace transform, s-domain).
ddt
d
d
x Ax Bu F
y Cx Eu G
ddt
d
d
x Ax Bu F
y Cx Eu G
-
73/220
Time-to-Frequency Transform
Take Laplace transform:
s s s s d ss s s d s
x Ax Bu Fy Cx Eu G
( ) ( ) ( ) ( )( ) ( ) ( ) ( )
x I A Bu I A FH u H
( ) ( ) ( ) ( ) ( )( ) ( ) ( ) ( )
s s s s d ss s s d su d
1 1
y C I A B E u C I A F GG u G
( ) [ ( ) ] ( ) [ ( ) ] ( )( ) ( ) ( ) ( )
s s s s d ss s s d su d
1 1
G C I A B Eu s s( ) ( ) 1
G C I A F Gd s s( ) ( ) 1
H I A BH I A F
u
d
s ss s d s
( ) ( )( ) ( ) ( )
1
1
)()()()(
)()(2221
12111
sGsGsGsG
ssuu
uuu EBAICG
74/220
State-Space Average Modeling: Transfer Functions
dsGsG
iv
sGsGsGsG
iv
d
d
d
i
uu
uu
i
o
)()(
)()()()(
2
1
2221
1211
Interpretation of the transfer function matrix
Input-to-output voltage gain(audio susceptibility)
Output impedance
Input AdmittanceLoad-to-line current gain
Control-to-output voltage gain
Control-to-input current gain
-
75/220
State-Space Average Modeling: Transfer Functions
vo
vc
io
d
vi
; ;
vv
vi
vd
o
i
o
o
o
K PWM A(s)
iidi
ii
vi i
o
i
i
i
;
;
Loop compensator
Disturbances Results
Control action
76/220
State-Space Averaging: Time-to-Frequency Transform
LC
C
LLC
C
i
o
LC
C
LLC
Ci
o
rrRrRLCs
rRLCrRCrs
CsrDsvsv
rrRrRLCs
rRLCrRCrs
CsrVsdsv
2
2
]//[1
1)()(
]//[1
1)()(
Z s v si s
G so od dv
u
i
( ) ( )( )
( ) 00
12
-
77/220
State-Space Averaging: Time-to-Frequency Transform
( )( )
v sd s
V sr Cs
oi
C
o
1 2
LCsCrrRLs
CsrVsdsv
LC
Ci
o
2)(1
1)()(
CrrRLQ
LC
sQ
ss
CLo
o
oo
)(
11
1
22
Note: In most conditions, because R >>(rL + rC), the above equations can be approximated as
LCsCrrRLs
CsrVsdsv
LC
Ci
o
2)(1
1)()(
LCsCrrRLs
CsrDsvsv
LC
C
i
o
2)(1
1)()(
78/220
State-Space Average Modeling
The small signal control-to-output , and line-to-output are transferfunctions of two poles and one zero
( ) ( )
v sd s
o ( ) ( )v sv s
o
i
K s zs p s p
( )( )( )
1
1 2
with its parameters depending upon component values and operating point. p1, p2 can be complex poles or real poles. But p1, p2 and z1 all lie on LHP.
Note: The equivalent series resistance of the capacitor will introduce a LHP zero.
-
79/220
State-Space Averaging: Output Impedance
Output Impedance: Z s v si s
so od du
( ) ( )( )
( )
00
1c I A B du
( ) ( )
( )v si s
R
sQ
s
so
d o
1
1 1 1
2
2
s sQ
s
LCQ L
Rr r C
LCrr
Q Lr
r C
oo
oo
L C
L
C
CL
2 2
1 11
1 1 1
1 1 1
,( )
;
80/220
Modeling of Load Disturbances
L
CD
Q iL io
vc
icrc
rL
vi R vo
id
iR
-
81/220
Dynamic Responses for Step Load Changes
Intel: VRM (Voltage Regulator Module) and Enterprise Voltage Regulator-Down (EVRD) 11.0 Design Guidelines, Nov. 2006.
Adaptive voltagepositioning offset
VOFFSET (40mV)
Nominal setpoint voltageVSET (2.0V)
Dynamic voltagetolerance, VDYN-(100mV for 2s)
Initial voltage drop is mainly due to the product
of the load current step and ESR of the capacitors.
V = I ESR.(ESL effects are ignored)
Output voltageVOUT (50mV/Div)
Steady state voltage athigh current is approximatelyVSET VOFFSET IOUT RSENSE
Output current transientstep, I = 0 to 14A(5A/Div)
m5.2 GX;-MV Sanyo F15006 ;H5.2 SENSEOUT RCL
82/220
Buck Converter with Load Current DisturbanceBlock Diagram Representation
D
Q
vi
iL iC vo
io
1sL rL
rsCC
1
1R
id
iR
-
83/220
Control-to-Output Transfer FunctionsDC-DC Converters in CCM Operating Mode
Buck
( )( ) ( )
v sd s
V r CssQ
so
iC
o o
11 12
CrrRLQ
LC
CLo
o
)(
11
1
vi vo
L
CD
Boost
( )( )
( )( ( ))
( )'
'v sd s
VD
r Cs sL D Rs
Qs
o i C
o o
2
2
2
1 11 1
Cr
DCr
RDL
DQ
LCD
CLo
o
1
L
C
D
vovi
Buck/Boost
L C
D
vovi ( )( )
( )( ( ))
( )'
'v sd s
VD
r Cs sDL D Rs
Qs
o i C
o o
2
2
2
1 11 1
CDr
Dr
RDL
DQ
LCD
LCo
o
))(
()(
1
2 2
84/220
Transfer Functions of Buck Converters
Transfer Function Transfer Function
( ) ( )
v sd s
V r Css
Foi
C
o
D
1
21
( ) ( )v sv s
D r Css
Foi
C
o
U
1
211
( ) ( )
( )v si s
R
sQ
s
sFo
deq
o
U
1
1 1 1
2
212
( ) ( )
i sd s
Vr
r Css
FL iL
L
o
D
1
22
( ) ( )i sv s
DR
RCss
FLi o
U
1
221
( ) ( )i si s
r Css
FLd
C
o
U
1
222
s sQ
s R r
LCQ L
Rr r C
LCrr
Q Lr
r C
oo eq L
oo
L C
L
C
CL
2 2
1 11
1 1 1
1 1 1
,
,( )
;
-
85/220
Transfer Functions of Boost Converters
Transfer Function Transfer Function
( ) ( ) ( )
( )( )
'v sd s
VD
s s
sFo i z a
o
D
2 21
1 1
( ) ( ) 'v sv s D s
Foi o
U 1 1
211
( ) ( )
( )v sv s
R
sQ
s
sFo
ieq
o
U
1
1 1 1
2
212
( ) ( ) ( )'
i sd s
VD R
RC
sFL I
o
D
2 1 2
3 22
( ) ( ) ( )'i sv s
VD R
RCs
FLi
i
o
U
2 1
3 221
( ) ( ) 'i si s D
r Css
FLd
C
o
U 1 1 222
s sQ
s Q D LD R
r CD
D r C
DLC
DLC
Rr
Q LD R
r C
r CD R
LR r
DDD
r
oo
o LC
o
eq
C
eqC
zC
a eqL
C
2 2
1 11
2
2
2
1
1 1
1
,
;( )
; ( )
'
' ''
'
'
'
'
' '
86/220
Transfer Functions of Buck/Boost Converters
Transfer Function Transfer Function
( )( ) ( )
( )( )
'v sd s
VD
s s
sFo i z a
o
D
2 21
1 1
( ) ( )
( )'
v sv s
DD
r Css
Foi
C
o
U
1
211
( ) ( )
( )v si s
R
sQ
s
sFo
deq
o
U
1
1 1 1
2
212
( ) ( ) ( )'
i sd s
VD R
RCss
FL io
D 2 221
( ) ( ) ( )'i sv s
DD R
RCss
FLi o
U
2 2211 1
( ) ( ) 'i si s D
r Css
FLd
C
o
U 1 1 222
s sQ
s
DLC
Q D LD R
rD
rD
C
r CD RDL
R rD
DD
r
Rr
Q LD R
r C
oo
oo C L
zC
a eqL
C
oeq
C
eqC
2 2
2 2
2
2
1 11
2
1
1
1 1
, ;
' '
' ' '
'
' '
'
,
( )(
( ))
( )( )
;
-
87/220
Summary: Buck Converters at CCM
Transfer Function Transfer Function
( ) ( )
v sd s
V r Css
Foi
C
o
D
1
21
( ) ( )v sv s
D r Css
Foi
C
o
U 1 211
( ) ( )
( )v si s
R
sQ
s
sFo
deq
o
U
1
1 1 1
2
212
( ) ( )
i sd s
Vr
r Css
FL iL
L
o
D
1
22
( ) ( )i sv s
DR
RCss
FLi o
U 1 221
( ) ( )i si s
r Css
FLd
C
o
U
1
222
L
CD
Q iL io
vc
icrc
rL
vi R vo
idiR
L
C
iL io
vc
icrc
rL
vi R vo
idiRii
L
C
iL io
vc
icrc
rL
R vo
idiR
ii
x d1
averaging by using state duty ratio weighting
x d2
Switch-ON Period Switch-OFF Period
x A x B uy C x E u
2 2
2 2
x A x B uy C x E u
1 1
1 1
x Ax Buy Cx Du
A A AB B BC C CE E E
where =
1 1 2 2
1 1 2 2
1 1 2 2
1 1 2 2
d dd dd dd d
D ONQ ON
88/220
PWM Switch Model
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
-
89/220
PWM Switch Model
Introduction PWM Switch and its Invariant Properties CCM Analysis
DC and Small-Signal Model of PWM SwitchPWM Switch Model of Buck ConvertersPWM Switch Model of Boost ConvertersPWM Switch Model of Buck/Boost ConvertersPWM Switch Model of Cuk ConvertersAnalysis of PWM ConvertersRight-half Plane Zero of the ConvertersPWM Switch Model Including Storage-Time Modulation
DCM AnalysisDC and Small Model of PWM SwitchAnalysis of PWM ConvertersZero of Control-to-Output Transfer Function in DCM
90/220
Pulse-Width Modulator
For natural sampling,
v c
vp
d ts ( )
vpD T
Tvv
ON
S
c
p
( ) ( )d s K v sm c
Kvm p
1 constant
-
91/220
PWM Switch Model
( )i ta
( )v tcp
cd
1-d
p
( )v tap
a)(~ tic
(common)
(passive)
(active)
Instantaneous value
PWM Switch Modeling[1] V. Vorperian, Simplified Analysis of PWM Converters Using Model of PWM Switch Part I: Continuous
Conduction Mode, IEEE Trans. on Aero. and Electron. Sys., vol. 26, no. 3, pp. 490-496, May 1990.[2] V. Vorperian, Simplified Analysis of PWM Converters Using Model of PWM Switch Part II: Discontinuous
Conduction Mode, IEEE Trans. on Aero. and Electron. Sys., vol. 26, no. 3, pp. 497-505, May 1990.[3] E. Van Dijk, J. N. Spruijt, D. M. O'Sullivan, and J. B. Klaassens, PWM-switch modeling of DC-DC converters,
IEEE Transactions on Power Electronics,, vol. 10, no. 6, pp. 659 -665, Nov 1995.
92/220
Modeling of the Switch
( )i ta )(~ tic
( )v tap ( )v tcp
vcp
vap
i Dia c
v Dvcp ap
)(~ of valueaverage thedenotes :Note tii aa
During a PWM switching interval:
ss
sea TtDT
DTttiti
,00),(~
)(~
ss
sapcp TtDT
DTttvtv
,00),(~
)(~
-
93/220
Small-Signal Perturbation
ccccaa
ccaa
ca
idiDdIDIiI
iIdDiI
dii
))((
v dv
V v D d V v
V v DV V d Dv dv
cp ap
cp cp ap ap
cp cp ap ap ap ap
( )( )
v V vv V v
cp cp cp
ap ap ap
Make a small-signal perturbation at a DC operating point
i I ii I ia a a
c c c
d D d
i Dia c
v Dvcp ap
We can obtain
94/220
DC and AC Analysis
For dc analysis, eliminate all small signal perturbation terms, we can get itsequivalent dc model.
I DIV DV
a c
cp ap
For ac analysis, set all dc terms to zero and neglect second order nonlinear terms, we can get its equivalent small-signal ac model.
cca iDdIi
v V d Dvcp ap ap
-
95/220
PWM Switch Model
DC Model:
apcp
ca
DVVDII
a c
p
1 D
I a Ic
Vap Vcp
AC Model:
dVvDv
dIiDi
apapcp
cca
VD
dap
a c
p
1 DI dc
ai ciciD
96/220
DC Analysis of a Buck Converter
DC Analysis
rL
rCR
L
C
a
p
c
Li
o
rRRD
VV
p
a c
1Vi Vo
DR
rL
-
97/220
AC Analysis of a Buck Converter
Note:In the following, the hat above the small signal variables are neglected for simplicity.
Small-signal equivalent circuit model of the buck converter.
voRrC
C
rLL iL io
vc
ic
id
iR
I dc
dDVi
p
1iv
Dii
aia
cic
ii
Vi
iS
iDvoR
rC
C
rLL iL io
vc
ic
id
iR
98/220
AC Analysis: Control-to-Output Transfer Function
Short the unrelated voltage source and open the unrelated current source
voRrC
C
rLL iL io
vc
iR
VD
dap
p
a c
1 Dii ic
'
-
99/220
AC Analysis: Control-to-Output Transfer Function
v VD
d D
v v R r sCr sL R r sC
V d R sr Cs R r LC s r r C r RC L r RC R r
cpi
o cpC
L C
iC
C L C L C L
/ /( / )( ) / /( / )
( )( ) ( ) ( )
11
12
G s v sd s
V R sr Cs RLC s r r C L r RC r RC R rd
oi
C
L C L C L
( ) ( )( )
( )( ) ( )
1
2
if rL
-
101/220
AC Analysis: Line-to-Output Transfer Function
)()()()1(
)1()1)(()1(
)/1()/1)(()/1(
/1)/1()(
/1)/1(
)/1//()()/1//(
2LCLCLC
Ci
CCL
Ci
CCL
Ci
C
CL
C
C
iCL
Ccpo
icp
rRRCrLRCrCrrsLCrRsCsrRDv
CsrRCsrsRCsLrCsrRDv
sCrRsCrRsLrsCrRDv
sCrRsCrRsLr
sCrRsCrR
DvsCrRsLr
sCrRvv
Dvv
G s v sv s
D R sr Cs RLC s r r C L r RC r RC R rv
o
i
C
L C L C L
( ) ( )( )
( )( ) ( )
1
2
if rL
-
103/220
AC Analysis: Disturbance-to-Output Transfer Function
LLLCCC
LCLC
LCC
LCLC
LCC
LC
LCC
LC
LC
C
LC
C
LC
CLC
d
o
rsLRCsrCrsrLCrsRLCsRRCsrrsLCrsrLCrsR
sLrCsrsRCRRCsrrsLCrsrLCrsR
sLrCsrsRCCsrRsLrCsrR
sLrsCrRsCrRsLrsCrR
sLrsCrR
sCrR
sLrsCrR
sCrR
sLrsCrR
sCrRsLrsCrRiv
22
2
2
))(1())(1()1())(1(
))(/1()/1())(/1(
/1)/1(
)(/1
)/1(
)//(/1
)/1()//()/1//(
vi
R s r LC s L r r C rs R r LC s L r r RC r r C R r
o
d
C L C L
C C L C L L
2
2
( )( ) ( ) ( )
104/220
AC Analysis: Disturbance-to-Output Transfer Function
if rL
-
105/220
AC Analysis: Disturbance-to-Output Transfer Function
LCLrr
RCss
LCr
Csrs
sisvsZ
LC
LC
d
oo 11
1
)()()(
2
2
o C LQ RC
r rL
1
o LC2 1
LrrRL
Lrr
RCLCL
rrRC
QLC
oLCoLC
o
o )(
111
111
1 2
1LC
106/220
AC Analysis: Disturbance-to-Output Transfer Function
Q LR
r r Lo C L
1 1 ( )
22
2 11)(
oo
LCop
Qss
LCLrr
RCsssZ
s s sQ
oo
2 2
111
1
1)(
11
221
2
2
2
sQ
sR
srLLCs
rr
LCr
LCr
CsrssZ
oeq
LL
CL
LCoq
R req L
11
rr LC
L
C
Q rLL
11
1
-
107/220
AC Analysis: Disturbance-to-Output Transfer Function
The exact output impedance is derived as:
LCRr
Rrrrr
LRCs
Rrs
LCr
Lrr
Csrs
Rr
RCrrCrr
RLsLC
Rrs
rCrrLsLCrsrRCrrRCrrLsLCrRs
rCrrLsLCrsRiv
LCLLC
C
LCLC
LLCLC
C
LCLC
LLCLCC
LCLC
d
o
1)1()(11)1(
)1(
)1()()1(
)(
)()()()(
2
2
2
2
2
2
108/220
AC Analysis: Disturbance-to-Output Transfer Function
111
1)(1)(
11
221
2
22
sQ
sR
sCrrLLCs
rr
LCr
LCr
Lrr
CsrssZ
oeq
CLL
CLLCLCoq
R req L
11
rr LC
L
C
Q Lr
r CL
C
11
1 1
-
109/220
Output Impedance of the Buck Converter
LCRr
Rrrrr
LRCs
Rrs
LCr
Lrr
Csrs
sisvsZ
LCLLC
C
LCLC
d
oo 1)1()(11)1(
)1(
)()()(
2
2
Z s v si s
R
sQ
s
soo
deq
o
( ) ( )( )
( )
11 1 1
2
2
Output Impedance of the Buck Converter:
s sQ
s R r
LCQ L
Rr r C
LCrr
Q Lr
r C
oo eq L
oo
L C
L
C
CL
2 2
1 11
1 1 1
1 1 1
,
,( )
;
110/220
Difficulties with Modeling and Control of SPS
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
-
111/220
6. Difficulties with Modeling and Control of SPS
Nonsmooth Systems (time and state discontinuity)Concepts of existence, uniqueness, stability not clearly defined for systems with discontinuous right hand sideConcept of chaotic dynamics relatively new To power electronics
112/220
Complex Behavior of Switching Power ConvertersChi Kong Tse, Complex Behavior of Switching Power Converters, CRC Press, 2004.
(c) Simulation results
(d) Experimental results
E = 22-33 V, L = 20 mH, C = 47 F, R = 22 , Vref = 11 V, A = 8.4, T = 400 s, VL = 3.8 V, VU = 8.2 V.
(a) Voltage-mode controlled buck converter.
(b) Operation waveforms.
u
VL
VU Vrampvcon
t
t3T2T1T0
-
113/220
Bifurcation of Voltage Mode Buck Converter in CCM
RLC
Z i
Z f
vref
v c
V vI i
CLOCK RAMP
vd
Vc
vc vc
io vo
rampv
u
VL
VUVramp
vcon
t
t3T2T1T0
Li
114/220
Vcon & Vramp at Different Gains
L = 20 mH, C = 47 F, R = 22 , T = 400 s, VL = 3.8 V, VU = 8.2 V, Vs = 33 V, Vref = 11 V.
KP = 5.0
KP = 3.0
KP = 7.0
KP = 9.0
8 msec
RL
vrefv c
vd
io vo
rampv
Li
PK
C
-
115/220
Phase Portraits
KP = 6.8KP = 3.0
KP = 7.4 KP = 12.0
116/220
Inductor Current and Output Voltage at Kp = 12.0
KP = 12.0Inductor Current
Output Voltage
-
117/220
Simulation Results with Varying Source Voltage
Source Voltage
Output Voltage
Inductor Curent
118/220
Experimental Setup for Measuring of a SPS
Active LoadDC Power Supply Buck Converter
Experimental Setup Experimental Buck Converter
-
119/220
Typical Waveforms
IL
VC
IL
VC
Coupling switching noise
5s
1ms
What happens?
120/220
Bifurcation Occurs in a Switching Power Supply
IL
IL
IL
IL
IL
IL
VC
VC
VC
VC
IL
VC
-
121/220
Intermittency, Parasitic and Common Mode Effects
IL
IL
VC
VC
Icom
122/220
Homework Assignments
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
-
123/220
Simulation of a Buck Converter
S. A. Shirsavar, "Teaching Practical Design of Switch-Mode Power Supplies," IEEE Trans. on Education, vol. 47, no. 4, pp. 467-473, Nov. 2004.
Vinvca
iL
LvL vo
io
C R
low-pass filteris
ic
Input voltage Vin = 12 VNominal output voltage Vout = 5 V
Switching Frequency fs = 76.5 kHzInductor L = 100 HCapacitor C = 2.4 FLoad resistor R => 25
id
124/220
Buck Converter
v VDL HrC FrRf KHz
i
L
C
s
10502000 14700 05550
%
.
.
ohm
ohm ohm
L
RvorC
C
rL
vi D
Q
iL
io
vc
iciSiD
Control-to-Output Transfer Function
( )( ) ( )
v sd s
V r CssQ
so
iC
o o
11 12
o o
L CLC
Q LR
r r C
1 1 1 ,( )
-
125/220
Control Loop Design
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Power Electronic Systems & Chips Lab.
~
Voltage Mode Control of DC-DC Converters
PWMModulator
LoopCompensator
v o
digital signal processor analog signal processor
vR
d
Output voltage feedback only!
load
LR di~
Buck Converter Boost Converter
Buck/Boost Converter C,uk Converter
vi vo
L
CD
L
C
D
vovi
L C
D
vovi
L1
C2D
C1
vi vo
Switching power converters
oi
vgosZ
sv~
sV
-
127/220
Control Loop Design for DC-DC Converters
L. H. Dixon, Closing the Feedback Loop, Unitrode, 1988.
Feedback Control to Achieve: Good Voltage Regulation (Load Variation & Line Variation) High Stability (Load Variation & Line Variation)
vref
d
ve
v v f do i ( )
Fm
ConverterPower Stage
)(sA
Rvi
Single-Loop-Controlled Switching Regulator
128/220
Small Signal Modeling of Voltage Mode ControlSwitching Mode DC-DC Regulator
BuckBoost
Buck/Boost
Power Stage
PulseModulator
ErrorProcessor rv
~
ov~
cv~
oi~
d~
iv~
LoadDynamics
-
129/220
Small Signal Transfer Functions
G s vvv
o
i d io
( ) ,
0 0
: Open-loop input-to-output
G vdd
o
v ii o
, 0 0
K dv c
PWM
: Open-loop output impedance
: Control to output transfer function
: PWM modulator gain
Z vip
o
o d v i
, 0 0
A s vv
c
o
( )
KV p
PWM 1
: Compensator gain
: Power stage transfer function
: PWM dc gain
G Z Gv p d, ,
DC gain: DV Vc p
1
A(s) : Compensator to be designed
130/220
Small Signal Model of Voltage Mode Control
dv
iv
vv o
o
o
i
o ~~
;~~
;~~
KPWM A(s)
)~,~,~(~ divfv oio Zp
Gd
-A(s)
Gv
KPWM
)(1Z
sLZ
Z popcpo
)()()( sGKsAsL dPWM)(1
sL
GGG vovcvo
oi~
d~
iv~
oi~
d~
iv~ov~
cv~cv~
ov~
The design problem is to synthesize a loop compensator L(s) to satisfy given specifications of Gvc(s) and Zpc(s).
-
131/220
Small Signal Transfer Functions
G s vvv
o
i d io
( ) ,
0 0
: Open-loop input-to-output
G vdd
o
v ii o
,0 0
cvdk
PWM
: Open-loop output impedance
: Control to output transfer function
: PWM modulator gain
Z vip
o
o d vi
, 0 0
A s vv
c
o
( )
KVp
PWM 1 : Compensator gain: PWM dc gain
132/220
Control Concept for Switching Regulators
vo
vc
io
d
vi
; ;
vv
vi
vd
o
i
o
o
o
K PWM A(s)
iidi
ii
vi i
o
i
i
i
;
;
Loop compensator
Disturbances Results
Control action
-
133/220
Single-Loop Control Example
vovi
PWM Modulator
Zi
Zf
vref
DriverStage
Buck Converter
Power Stage
Error Amplifierand Compensation
VoltageDividerNetwork
vc
d
R
VP
134/220
Single-Loop-Controlled PWM Converter
Z i
Z f
v ref
Logic&
DRIVE
+
A
Se
vi
vo
B
FvFm
RL
Features of Voltage Mode Control: Error voltage compared to external ramp Se. Unique loop gain for control loop design Analog signal can be measured at B. Digital signal can be measured at A.
-
135/220
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Control Architecture
Power Electronic Systems & Chips Lab.
~
136/220
The Purpose of Feedback Control
PWMModulator
LoopCompensator
vgvo
vR
load
RL di~
Buck Converter
Power Flow Control Efficiency & Stability & Power Quality Output Voltage Regulation Zero Output Impedance Input Voltage Disturbance Rejection Zero Audio Suspetibility Control with Estimated Feedback Signals
-
137/220
Control of Basic PWM DC-DC Converters
PWMModulator
LoopCompensator
vg
vo
vR
Efficiency
load
RL di~
Switching power converters
Output Impedance
GateDrive
osZ
sv~
sV
Buck ConverterBuck-Boost ConverterBoost Converter
138/220
Multiple-Loop Control Scheme
PWMModulator
Voltage LoopCompensator
vo
vRCurrent LoopCompensator
CurrentSensing
CurrentLoop
VoltageLoop
load
RL di~Buck ConverterBuck-Boost ConverterBoost Converter
vgosZ
sv~
sV
Switching Power Converters
CarrierGenerator
-
139/220
Pioneer Voltage-Mode PWM Control IC SG1524
The First PWM IC: SG1524, Bob Mammano, 1976.
SG1524: -55C ~ 125CSG2524: -25C ~ 85CSG3524: 0C ~ 70C
Pioneer Current-Mode PWM Control IC uc3842
7/12
5/9
4/7
2/3
1/1
3/5
Vcc
VFB
ground
compCurrent
sense
OSC
34V
2.50V
UVLO
Vrefgoodlogic
TS2R
R 1VPWMlatch Power
ground
output
CurrentSense
comparator
Vref5.0V
50mAInternalbias
5VrefRS /
Erroramp
RT /CT
R
Vc
8/14
7/11
6/10
5/8
Block Diagram of UC3842/3/4/5
ccV
RS
Q
Latch
Clock
PWMcomparator
Erroramplifier
reference
Li
eV
sR
sV
RV
oV
RESET
LatchOutput(PWM)
Clock (SET)
Li
eV
sV
-
141/220
Voltage-Mode Control vs. Current-Mode Control
REF: Bob Bell, National Semiconductor, "Virtual-current mode: current-mode control without the noise," EDN, Feb. 16, 2006.
Voltage Mode Control Current Mode Control
VIN VINVOUT VOUTL1 L1
D1 D1
Q1 Q1
FET DRIVER
CLOCKSAWTOOTHGENERATOR
1.25VREFERENCE
Q RS
Q RS
CLOCK1.25V
REFERENCE
SWITCH DRIVER
SWITCH-CURRENTMEASURE
PWMCOMPARATOR
PWMCOMPARATOR
ERRORAMPLIFIER
ERRORAMP
142/220
Control Block Diagram of a Current-Mode Controller
-
143/220
Peak-Current-Mode-Control of a Buck Converter
144/220
Current-Mode Control vs. Voltage-Model Control
A. J. Forsyth and S. V. Mollov, "Modelling and control of DC-DC converters," IEEE Power Engineering Journal, vol. no. 5, pp. 229-236, Oct. 1998.
(a) Voltage-Mode Control (b) Current-Mode Control
-
145/220
Current-Mode vs. Voltage-Mode Control
Control-to-output bode plots for voltage and current mode converters. (The current mode converter has an extra 90 degrees of phase.)
146/220
Voltage Mode Control of DC-DC Converters
PWMModulator
LoopCompensator
vg vo
digital signal processor analog signal processor
vR
d
Output voltage feedback only!
load
LR di~
Buck Converter Boost Converter
Buck/Boost Converter C,uk Converter
vi vo
L
CD
L
C
D
vovi
L C
D
vovi
L1
C2D
C1
vi vo
Switching power converters
oi
-
147/220
Voltage Mode Control
REF: Voltage-mode control and compensation - Intricacies for buck regulators (Timothy Hegarty, EDN 2008)
Converter power stage
ModulatorType III compensator
PWMcomparator
Erroramp
COMP
SW
VIN
Vout
IoutRLRESRCO
LORDCR
RFB1FB
RC1CC1
CC2
RC2
RFB2
CC3
Vref
Driver&
deadtime
D
PWM rampVramp
148/220
Robert W. Erickson and Dragan Maksimovic, Fundamentals of Power Electronics, Kluwer Academic Publishers, 2nd Ed., February 2001.
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Control Loop Design Basics
Power Electronic Systems & Chips Lab.
-
149/220
Control Loop Design Basics
1. Introduction2. Effect of Negative Feedback
Feedback reduces the transfer function from disturbances to the output Feedback causes the transfer function from the reference input to the
output to be insensitive to variations in the gains in the forward path ofthe loop
3. Construction of the important quantities 1/(1+T) and T/(1+T) and theclosed-loop transfer functions
4. Stability The phase margin test The relation between phase margin and closed-loop damping factor Transient response vs. damping factor
150/220
Control Loop Design Basics ..
5. Regulator design Feedback Lead (PD) compensator Lag (PI) compensator Combined (PID) compensator Design example
6. Measurement of loop gains The Voltage injection Current injection Measurement of unstable systems
7. Summary of key points
-
151/220
Modeling of Switching Converters
tvg tvo
tiload
Pulse-widthmodulator
Switching converter
transistor gate driver
Load
tvc t t
tsTsdT tvo
tvg
tiload td
divftv loadg ,,
disturbances
Control input
Switching converter
Output voltage of aswitching converterdepends on duty cycle d,input voltage vg, and loadcurrent iload
td
152/220
The DC Regulator Application
Objective: maintain constantoutput voltage vo(t) = V, in spiteof disturbances in vg(t) andiload(t).Typical variation in vg(t) : 100Hzor 120Hz ripple, produced byrectifier circuit.
Load current variations: a significant step-change in load current, suchas from 50% to 100% of rated value, may be applied.A typical output voltage regulation specification: 5V 0.1V.
Circuit elements are constructed to some specified tolerance. In highvolume manufacturing of converters, all output voltages must meetspecifications.
tvo tvg
tiload td
divftv loadgo ,,
disturbances
Control input
Switching converter
-
153/220
The DC Regulator Application
So we cannot expect to set the duty cycle to a single value, and obtain agiven constant output voltage under all conditions.
Negative feedback: build a circuit that automatically adjusts the duty cycleas necessary, to obtain the specified output voltage with high accuracy,regardless of disturbances or component tolerances.
154/220
Negative Feedback: A Switching Regulator System
gv v
tiload
Pulse-widthmodulator
Switching converter
transistor gate driver
Load
cv
Powerinput
H(s)sensorgain
error signal
compensator
sGc ev
refvreference
input
+
tvo
-
155/220
Negative Feedback
tvo tvg
tiload td
divftv loadgo ,,
disturbances
Control input
Switching converter
Pulse-widthmodulator
cv
error signal
sensorgain
tverefvreference
input
+ compensator
156/220
Power Electronic Systems & Chips Lab., NCTU, Taiwan
Effect of Negative Feedback
Power Electronic Systems & Chips Lab.
~
-
157/220
Effect of Negative Feedback
C tvg svo sdsj
eL
R
sdse
1 : M(D)
Small signal model: open-loop equivalent circuit mode of a buck converter
Output voltage can be expressed as
where
tiload
sisZsvsGsdsGtv loadoutgvgvdo
000
00
0
gloadload
g
vdload
oout
id
g
ovg
sisv
ovd si
svsZsvsvsG
sdsvsG
158/220
Voltage Regulator System Small-Signal Model
Use small-signal converter model
Perturb and linearize remainder of feedback loop
.
etctvvtv
tvvtv
eee
refrefref
C tvg svo sdsj
eL
R
sdse 1 : M(D)
tiload
error signal
sve svref
referenceinput
+
svc
H(s)
sGccompensator Pulse-width
modulator
sd
svsH o
sensorgain
MV1
-
159/220
H(s)
Regulator System Small-Signal Block Diagram
error signal
sve svref
referenceinput
svc sGc
compensator Pulse-widthmodulator
svsH o
Sensor gain
MV1 sd sGvd
sGvd svgac line variation
svo
tiloadload current variation
output voltagevariation
duty cyclevariation
Converter power stage
sZout
Mvdc
outload
Mvdc
vdg
Mvdc
Mvdcrefo VGHG
ZiVGHG
GvVGHG
VGGvv/1
/1
/1
/
160/220
Solution of Block Diagram
Manipulate block diagram to solve for . Result is
Loop gain T(s) = products of the gains around the negative feedback loop.
which is of the form
svo
Mvdc
outload
Mvdc
vdg
Mvdc
Mvdcrefo VGHG
ZiVGHG
GvVGHG
VGGvv/1
/1
/1
/
" "/with 1
1
1
1
gainloopVsGsGsHsTT
ZiT
Gv
TT
Hvv
Mvdc
outload
vggrefo
-
161/220
Feedback Reduces Sensitivity to Load Disturbances
Original (open-loop) line-to-output transfer function:
With addition of negative feedback, the line-to-output transfer function becomes:
Feedback reduces the line-to-output transfer function by a factor of
If T(s) is large in magnitude, then the line-to-output transfer function becomessmall.
0 0
laodid
g
ovg sv
svsG
sTsZ
sisv out
vv
load
o
gref
1
00
sT11
162/220
Closed-loop Output Impedance
Original (open-loop) output impedance:
With addition of negative feedback, the output impedance becomes:
Feedback reduces the output impedance by a factor of
If T(s) is large in magnitude, then the output impedance is greatly reduced in magnitude.
0 0
gv
dload
oout si
svsZ
sTsZ
sisv out
vv
load
o
gref
1
00
sT11
-
163/220
Feedback Causes Reduction of Sensitivity
If the loop gain is large in magnitude, i.e., || T || >> 1, then (1+T) T and T/(1+T) T/T = 1. The transfer function then becomes
which is independent of the gains in the forward path of the loop.This result applies equally well to dc values:
0
101
00
1HT
THV
Vref
o
Closed-loop transfer function from to is : svrefv
sT
sTsHsv
sv
laod
giv
ref
o
1
1
00
sHsvsv
ref
o 1
Feedback causes the transfer function from the reference input to the output to be insensitive to variations in the gains in the forward path of the loop
164/220
Construction of 1/(1+T) and T/(1+T)
cf
At the crossover frequency fc, ||T || = 1
80dB
60dB
40dB
20dB
0dB
-20dB
-40dB1Hz 1KHz 10kHz100Hz 100kHz10Hz
f
-40dB/dec
-20dB/dec
-40dB/dec
zf
2pf
1pf
dB0T dBQ
2
2
11
0
11
1
ppp
z
s
s
Qs
s
TsT
crossoverfrequency
T
Example
-
165/220
Approximating 1/(1+T) and T/(1+T)
1for
1for 11 TT
TT
T
1for 1
1for 1
11
T
TsT
sT
166/220
Construction of T/(1+T)
cf
80dB
60dB
40dB
20dB
0dB
-20dB
-40dB1Hz 1KHz 10kHz100Hz 100kHz10Hz
f
-40dB/dec
-20dB/deczf
2pf
1pf T
crossoverfrequency
1for
1for 11 TT
TT
T
TT1
-
167/220
Derive Reference to Output Transfer Function
This is the desired behavior: the output follows the reference according to the idealgain 1/H(s) . The feedback loop works well at frequencies where the loop gain T(s)has large magnitude.At frequencies above the crossover frequency, || T ||
-
169/220
Interpretation: How the Loop Gain Rejects Disturbances?
1for 1
1for 1
11
T
TsT
sT
Bellow the crossover frequency: f < fc and || T || > 1Then 1/(1+T) 1/T, and disturbances are reduced in magnitude by 1/ || T ||
Above the crossover frequency: f < fc and || T || < 1Then 1/(1+T) 1, and the feedback loop has essentially no effect ondisturbances
170/220
Terminology: Open-loop vs. Closed-loop
Upon introduction of feedback, these transfer functions become (closed-looptransfer functions):
The loop gain:
Original transfer functions, before introduction of feedback (open-loop transferfunctions):
sZsGsG outvgvd
sTsZ
sTsG
sTsT
sHoutvg
1
1
11
sT
-
171/220
Stability
Even though the original open-loop system is stable, the closed-loop transfer functionscan be unstable and contain right half-plane poles. Even when the closed-loop system isstable, the transient response can exhibit undesirable ringing and overshoot, due to thehigh Q -factor of the closed-loop poles in the vicinity of the crossover frequency.When feedback destabilizes the system, the denominator (1+T(s)) terms in the closed-loop transfer functions contain roots in the right half-plane (i.e., with positive realparts). If T(s) is a rational fraction of the form N(s) / D(s), where N(s) and D(s) arepolynomials, then we can write
sDsN
sD
sDsNsT
sDsNsN
sDsN
sDsN
sTsT
1
11
1
11
Could evaluate stability by evaluatingN(s) + D(s), then factoring to evaluateroots. This is a lot of work, and is notvery illuminating.
172/220
Nyquist stability theorem: general result. A special case of the Nyquist stability theorem: the phase margin
test
Allows determination of closed-loop stability (i.e., whether 1/(1+T(s)) containsRHP poles) directly from the magnitude and phase of T(s).
A good design tool: yields insight into how T(s) should be shaped, to obtaingood performance in transfer functions containing 1/(1+T(s)) terms.
Determination of Stability Directly from T(s)
-
173/220
The Phase Margin Test
If there is exactly one crossover frequency, and if T(s) contains no RHP poles, then
A test on T(s), to determine whether 1/(1+T(s)) contains RHP poles. The crossoverfrequency fc is defined as the frequency where
dB012 cfjTThe phase margin m is determined from the phase of T(s) at fc, as follows:
cm fjT 2180
the quantities T(s) / (1+T(s)) and 1/(1+T(s)) contain no RHP poles wheneverthe phase margin jm is positive.
174/220
Example: a loop gain leading to a stable closed-loop system
68112180
1122
m
c
fjT
60dB
40dB
20dB
0dB
-20dB
-40dB
1Hz 1KHz 10kHz100Hz 100kHz10Hz
cfzf
1pf crossoverfrequency
T
T TT
m
0
-90
-180
-270
f
-
175/220
Example: a loop gain leading to an unstable closed-loop system
502301802302
m
c
fjT
60dB
40dB
20dB
0dB
-20dB
-40dB
1Hz 1KHz 10kHz100Hz 100kHz10Hz
cfzf
1pf crossoverfrequency
T
T TT
0m
0
-90
-180
-270
f
176/220
Relation Between Phase Margin and Closed-loop Damping Factor
A small positive phase margin leads to a stable closed-loop systemhaving complex poles near the crossover frequency with high Q. Thetransient response exhibits overshoot and ringing.
Increasing the phase margin reduces the Q. Obtaining real poles, withno overshoot and ringing, requires a large phase margin.
The relation between phase margin and closed-loop Q is quantified inthis section.
How much phase margin is required?
-
177/220
A Simple Second-order System
f
40dB
20dB
0dB
-20dB
-40dB
2f
0f
T
TTT
m
0
-90
-180
-270
-90-40dB/dec
-20dB/dec
10/2f
210 f
2f
ff0
220
fff
Consider the case where T(s) can be well-approximated in the vicinity of thecrossover frequency as
20
1
1
s
s
sT
178/220
Closed-Loop Response
if
Then
20
1
1
s
s
sT
20
2
0
1
111
11
s
s
sTsT
sT
or,
2
1
11
cc s
QssT
sT
where2
002 2
Qf
cccc
-
179/220
Low-Q case
202
00
Q :ionapproximat -low QQ
Q cc
c
f
40dB
20dB
0dB
-20dB
-40dB
2f
0f
T
-40dB/dec
-20dB/dec
ff0
220
fff
20 fffc
20 / ffQ T
T1
180/220
High-Q case
2
0
c
020 2
Q fQ c
f
40dB
20dB
0dB
-20dB
-40dB
2f
0f
T
-40dB/dec
-20dB/dec
ff0
220
fff
20 fffc
cffQ /0
TT1
60dB
-
181/220
Q vs. m
Solve for exact crossover frequency, evaluate phase margin, express as function of m. Result is:
4
41
2411
tan
sincos
QQ
Q
m
m
m
182/220
Q vs. m
-20dB
20dB
10dB
0dB
-10dB
-15dB
-5dB
15dB
5dB
Q
0 10 20 30 40 50 60 70 80 90
Q=1 0dB
Q=0.5 -6dB
m
52m
76m
-
183/220
Transient response vs. damping factor
Unit-step response of second-order system T(s)/(1+T(s))
22121
1
12
2
212
2
2/
4112
,
5.0 1
5.0 14tan2
14sin
1421
21
QQ
Qe
e
tv
QQtQ
QQ
Qetv
c
tt
c
Qtc
For Q>0.5, the peak value is
14/ 21peak Qetv
184/220
Transient Response vs. Damping Factor
2
0 5radianst c ,
1.5
1
0.5
010 15
tv Q=50Q=10
Q=4
Q=2
Q=1Q=0.75Q=0.5
Q=0.3
Q=0.2
Q=0.1Q=0.05Q=0.01
-
185/220
Typical specifications: Effect of load current variations on output voltage regulation
This is a limit on the maximum allowable output impedance Effect of input voltage variations on the output voltage regulation
This limits the maximum allowable line-to-output transfer function Transient response time
This requires a sufficiently high crossover frequency Overshoot and ringing
An adequate phase margin must be obtained
The regulator design problem: add compensator network Gc(s) tomodify T(s) such that all specifications are met.
5. Regulator Design
186/220
Lead (PD) Compensator
f
zf
0cG
cG
cG
+45/decade
10/zf
zf10
z
pc f
fG 0
pz fff max
Improves phase margin
p
zcc
s
s
GsG1
1
0
-45/decade
10/pf
0
pf
-
187/220
Lead Compensator: Maximum Phase Lead
ff
ff
ff
fG
fff
z
p
p
z
z
p
c
pz
sin1sin1
2tan 1max
max
0
15
30
45
60
75
90
1 10 100 1000
maximumphase lead
zp ff /
188/220
Lead (PD) Compensator
f
zf
0cG
cG
cG
+45/decade
10/zf
zf10
z
pc f
fG 0
pz fff max
-45/decade
10/pf
0
pf
To optimally obtain a compensator phase lead of at frequency fc, the pole andzero frequencies should be chosen as follows:
ff
ff
cp
cz
sin1sin1
sin1sin1
If it is desired that the magnitude of the compensator gain at fc be unity, then Gc0 should be chosen as
p
zc f
fG 0
-
189/220
Example: Lead Compensation
60dB
40dB
20dB
0dB
-20dB
-40dB
cfzf
pf
original gain
T
TT
T
m
0
-90
-180
-270
f
0
compensated gain
compensated phase asymptotes
original phase asymptotes
0f
0T
00 cGT
190/220
Lag (PI) Compensation
f
zf
cG
cG
cG
10/Lf
Lf10
Improves low-frequencyloop gain and regulation
s
GsG Lcc 1
+45/decade-90
-20dB/dec
0
-
191/220
Example: Lag Compensation
0
0
1s
TsT uu40dB
20dB
0dB
-20dB
-40dB
cf
010 fT
T
m
0
-90
-180
f
0f0uT
0uc TG
1Hz 1KHz 10kHz100Hz 100kHz10Hz
90
Original (uncompensated)loop gain is
Compensator:
s
GsG Lcc 1
Design strategy:chooseGc to obtain desiredcrossover frequencyL sufficiently low tomaintain adequate phasemargin
uT
uT
0f
Lf
Lf10
192/220
Example, Continued
Construction of 1/(1+T), lag compensator example:
40dB
20dB
0dB
-20dB
-40dB
cf
T
f
0f0uc TG
1Hz 1KHz 10kHz100Hz 100kHz10Hz
0f
Lf
Lf
0
1uc TG T1
1
-
193/220
Combined (PID) Compensator
40dB
20dB
0dB
-20dB
-40dB
f1Hz 1KHz 10kHz100Hz 100kHz10Hz
zfcmG
cG
cG
90/dec
10/Lf110 pf
45/dec
10/1pf-90
1pf
21
11
11
pp
Z
L
cmc
s
s
s
s
GsG
cG
cfLf
2pf
10/zf
10/2pfLf10 zf10
-90/dec
cG
0
90
-180
-90
194/220
5. An Illustrated Design Example
V28tvg tvo
loadi
Pulse-widthmodulator
transistor gate driver
cv
H(s)sensorgain
error signal
compensator
sGc ev
V5refv
referenceinput
+
F50 L
3R
F500C
V4mV
vH
100kHzsf
-
195/220
Quiescent Operating Point
Input voltageOutputQuiescent duty cycleReference voltageQuiescent value of control voltageGain H(s) 3/115/5/
14.25
536.028/153,5,15
28
VVHVDVV
VVD
RAIVVVV
ref
Mc
ref
load
g
196/220
Small-Signal Model
C tvg svodRV
L
R
1 : D
siload
error signal
sve 0 refv + svc
H(s)
sGccompensator
sd
svsH o
MV1
dDV
2
V4mV
31
H
sT
-
197/220
2
000
0
1
1
s
Qss
GsG dvd
Open-Loop Control-to-Output Transfer Function Gvd(s)
LCs
RLsD
VsGvd21
1
f
40dB
20dB
0dB
-20dB
-40dB
vdG
vdG
0
-90
-180
-270
0fStandard form:
Salient features:
dB5.195.9
kHz12
12
28
0
00
0
LCRQ
LCf
VFVGd
60dB
vdG
vdG
dBV29V280 dGdB5.199.50 G
kHz1.110 02/1 0 fQ
1Hz 1KHz 10kHz100Hz10Hz 100kHz
900Hz10 02/1 0 fQ
198/220
Open-loop line-to-output transfer function and output impedance
Same poles as control-to-output transfer function standard form:
2
000
0
1
1
s
Qss
GsG dvd
LCs
RLs
DsGvd21
1
Output impedance:
LCs
RLs
sLsLsC
RsZout21
1
-
199/220
System Block Diagram
sve 0 refv + svc
H(s)
sGcMV1 sd sGvd
sGvg svg
ac linevariation
sZout
+
+
svo
tiloadload current
variation
duty cyclevariation
Converter power stage
sHsGV
sGsT vdM
c
1
V4mV
31
H
2
000
1
1
s
QsD
VV
sHsGsTM
c
200/220
dB4.733.2
1
1
0
2
000
0
Mu
uu
DVHVT
s
Qs
TsT
Uncompensated Loop Gain (with Gc=1)
f
40dB
20dB
0dB
-20dB
-40dB
uT
0
-90
-180
-270
0f
With Gc=1, the loop gain is
uT
dB4.723.20 uT dB5.199.50 Q
kHz1.110 02/1 fQ
1Hz 1KHz 10kHz100Hz10Hz 100kHz
900Hz10 02/1 fQ
uTuT
0kHz1
5 ,kHz8.1 mc f
-
201/220
Lead Compensator Design
Obtain a crossover frequency of 5kHz, with phase margin of 52 Tu has phase of approximately -180 at 5kHz, hence lead (PD)
compensator is needed to increase phase margin. Lead compensator should have phase of +52 at 5kHz Tu has magnitude of -20.6dB at 5kHz Lead compensator gain should have magnitude of +20.6dB at 5kHz Lead compensator pole and zero frequencies should be
kHz5.1452sin152sin1kHz5
kHz7.152sin152sin1kHz5
p
z
f
f
Compensator dc gain should be dB3.117.310
2
00
p
z
u
cc f
fTf
fG
202/220
Lead Compensator Bode Plot
f
zf0cG
cG
cG
10/zfzf10
z
pc f
fG 0
pz
c
ff
f
10/pf0
pf
40dB
20dB
0dB
-20dB
-40dB 0
-180
1Hz 1KHz 10kHz100Hz10Hz 100kHz
cG
-90
90
cG
-
203/220
2
000
00
11
1
s
Qs
s
s
CTsT
p
zcuLoop Gain with Lead Compensator
f
40dB
20dB
0dB
-40dB 0
-90
-180
-270
0f
T
dB7.186.80 TdB5.199.50 Q
1Hz 1KHz 10kHz100Hz10Hz 100kHz
0
T
-20dB
T
zfcf
pf
52m
170kHz
900kHz
1.4kHz
1.1kHz
17kHz
14kHz
5kHz
1.7kHz1kHz
204/220
1/(1+T) with Lead Compensator
need more low-frequency loop gainhence, add inverted zero (PID controller)
f
40dB
20dB
0dB
-40dB
0f
dB7.186.80 T dB5.199.50 Q
1Hz 1KHz 10kHz100Hz10Hz 100kHz
T
-20dB
zfcf
pfT1
1dB7.1812.0/1 0 T
0Q
-
205/220
Improved Compensator (PID)
p
L
Zcmc
s
s
s
GsG1
11
40dB
20dB
0dB
-20dB
-40dB
f1Hz 1KHz 10kHz100Hz10Hz
zfcmG
cG
cG
90/dec
10/Lf
45/dec
10/pf-90
pf
cG
cfLf
10/zf
Lf10 zf10 -45/dec
cG
0
90
-180
-90
add inverted zero toPD compensator,without changing dcgain or cornerfrequencies
choose fL to be fc/10,so that phase marginis unchanged
206/220
T(s) and 1/(1+T(s)), with P