Special Topics in Power Electronics

EEE-569 Lecture Notes

Lecture 9 Resonant Converters

A. Prof. Dr. Canras Batunlu

METU Northern Cyprus Campus
 Resonant and Soft-switching Converters

Typical switching trajectories of power switches

During the turn-on and turn-off processes, the power device has to withstand high voltage and current simultaneously,
resulting in high switching losses and stress. Dissipative passive snubbers are usually added to the power circuits so that
the dv/dt and di/dt of the power devices could be reduced, and the switching loss and stress be diverted to the passive
snubber circuits. However, the switching loss is proportional to the switching frequency, thus limiting the maximum
switching frequency of the power converters. Lots of research efforts were diverted towards the use of resonant
converters. The concept was to incorporate resonant tanks in the converters to create oscillatory
(usually sinusoidal) voltage and/or current waveforms so that zero-voltage switching (ZVS) or zero-current switching
(ZCS) conditions can be created for the power switches. The reduction of switching loss and the continual improvement
of power switches allow the switching frequency of the resonant converters to reach hundreds of kilo-Hertz
(typically 100–500 kHz).
Unlike the resonant converters, new soft-switched converters usually utilize
the resonance in a controlled manner. Resonance is allowed to occur just before and during
the turn-on and turn-off processes so as to create ZVS and ZCS conditions. Other
than that, they behave just like conventional PWM converters. With simple modifications,
many customized control integrated circuits (ICs) designed for conventional converters
can be employed for soft-switched converters. Because the switching loss and stress have
been reduced, soft-switched converter can be operated at the very high frequency
(typically 500 kHz to a few Mega-Hertz). Soft-switching converters also provide an effective
solution to suppress EMI and have been applied to DC–DC, AC–DC, and DC–AC converters.

Typical: (a) ideal and (b) practical switching waveforms.

 Second-Order Circuits
• Second-order circuits contain 2 storage elements
– Of different type, or
– Of the same type but not parallel or series

• The total response is still the sum of natural and forced

response (like the first order circuits)

• The natural response of a RLC circuit may consist of

functions other than decaying exponentials (unlike simple
RL and RC)

• To solve second-order circuits two initial conditions are

required
10
Ali Muhtaroğlu
 Circuits with 2 Storage Elements
• First-order circuits were represented by:
dx
 ax  f (t ) with tot al response : x(t )  xn (t )  x f (t )
dt
where xn(t) is the natural response obtained by setting f(t)=0
(homogeneous solution with a constant solved using initial condition)
and xf(t) is the forced response that satisfies the forced differential
equation.
d 2x dx
• The second-order circuits are represented by: 2  a1  a0 x  f (t )
dt dt
• A similar procedure as before applies:2
d xn dxn
1. Unforced (homogeneous) equation:  a1  a0 x n  0
dt 2 dt
d 2xf dx f
2. Forced equation to find the particular solution:  a1  a0 x f  f (t )
dt 2 dt
d 2 ( xn  x f ) d ( xn  x f )
3. Superposition produces:  a1  a0 ( xn  x f )  f (t )
dt 2 dt

11
Ali Muhtaroğlu
 Natural Response
d 2 xn dxn
• Consider the homogeneous equation: 2
 a1  a0 x n  0
dt dt

• Take xn(t)=Kest as the trial. Substitute into above: Ks e  Ksa1e  Ka0e  0

2 st st st

Ke st (s 2  a1s  a0 )  0
If K=0, no stored energy. More generally: s 2  a1s  a0  0...(1)
Solving, get:  a1  a12  4a0
s1, 2 
2
• Two natural solutions: xn1  K1e 1 and xn 2  K2e s2t
st

The two distinct solutions individually form a solution. For a linear equation,
any linear combination is also a solution: xn  xn1  xn 2  K1e 1  K2e 2
st st

• If we rewrite (1) as: s 2  20 s  02  0 i.e. a1  20 ; a0  02

where 0  a 0 : undamped natural frequency
a1
  : damping ratio

 
2 a0
then characteristic exponents: s1, 2     ( 2  1) 0
12
Ali Muhtaroğlu
 3 Cases for the General Solution : s=σ+jω
1. Overdamped Case:   1

• s1 and s2 are real and distinct : s1   1 , s2   2

t  t
• Natural response: xn (t )  K1e  K2e
1 2

• For circuits with passive elements driven by independent sources, σ1 and

σ2 cannot be positive. i.e. with independent sources killed (natural
response) the energy stored cannot increase beyond the initially stored
energy. The energy reduces with time or gets ‘damped’.

• Overdamped case is when there is sufficient damping for each term in the
natural response to steadily decay to zero. In lightly damped case, the
damping may have some oscillations.

13
Ali Muhtaroğlu
 3 Cases for the General Solution : s=σ+jω
2. Underdamped Case:   1
• s1 and s2 are complex, and are complex conjugates of each other.
s1    j , s2    j   0

• Natural response: xn (t )  et K1e jt  K2e jt 

• For a circuit with purely real element laws and sources, xn is real as well. K1
and K2 have to be complex conjugates.
   
xn  K1e jt  K1*e jt et  2 Re K1e jt et
• Using Euler’s Identity: e jt  cost  j sin t
ReK1e jt   ReReK1   j ImK1 cost  j sin t   ReK1  cost  ImK1 sin t
Therefore, ReK1e jt   1 B1 cost  1 B2 sin t where B1  2ReK1  ; B2  2ImK1 
2 2
Then, xn (t )  B1et cost  B2et sin t
• The damping is oscillatory in this case (multiplied by sinusoids) with
frequency ω rad/s. Therefore, this is called the underdamped case. The
ampitude of the oscillations decays as eσt, so the damping rate varies
14
inversely with σ.
Ali Muhtaroğlu
 3 Cases for the General Solution : s=σ+jω
2. Underdamped Case:   1

• Undamped case: When σ=0 (no decay)

– Occurs when daming ratio   0

– This case is a constant amplitude sinusoid with frequency ω=ω0

– Hence ω0 is called the undamped frequency.

15
Ali Muhtaroğlu
 3 Cases for the General Solution : s=σ+jω
3. Critically Damped Case:   1

• s1 and s2 are real and equal : s1  s2  

t t
• Natural response: xn (t )  K1e  K2te

t-multiplied form

16
Ali Muhtaroğlu
 Example : Parallel RLC Circuit Natural Response
v

ig
R L C

17
Ali Muhtaroğlu
 Example : Parallel RLC Circuit Natural Response
v Nodal Equation (KCL):
v 1 t dv
  v( )d  iL (t0 )  C  ig
ig R L t0 dt
R L C 1 dv 1 d 2 v dig
Differentiate:  vC 2 
R dt L dt dt
d 2v 1 dv 1 1 dig
Divide by C:   v
dt 2 RC dt LC C dt
1 1
Characteristic Equation: s  s 0
2

RC LC 1
The undamped natural frequency of the (parallel RLC) circuit: 0   a 0 
LC
a1 1 L
Parallel RLC Damping Ratio:  P  
2 a0 2 R C

18
Ali Muhtaroğlu
 Example : Parallel RLC Circuit Natural Response
v Nodal Equation (KCL):
v 1 t dv
  v( )d  iL (t0 )  C  ig
ig R L t0 dt
R L C 1 dv 1 d 2 v dig
Differentiate:  vC 2 
R dt L dt dt
d 2v 1 dv 1 1 dig
Divide by C:   v
dt 2 RC dt LC C dt
1 1
Characteristic Equation: s  s 0
2

RC LC 1
The undamped natural frequency of the (parallel RLC) circuit: 0   a 0 
LC
a1 1 L
Parallel RLC Damping Ratio:  P  
2 a0 2 R C
If  P  1, two real characteristic exponents
If  P  1,  P2  1  0, the characteristic exponents are complex
If ζ P  1, this is called critical damping - the two real exponents are the same
1 L
Critical parallel resistance can be found from : R cp 
2 C
For R  R cp (  1), underdampe d case; for R19  R cp (  1), overdamped case
Ali Muhtaroğlu
 Example : Series RLC Circuit Natural Response
R

i
vg L

- vC +

20
Ali Muhtaroğlu
 Example : Series RLC Circuit Natural Response
R
1 t di
i KVL: Ri 
C 
t0
i ( )d  vC (t0 )  L
dt
 vg
vg L

C Differentiate and divide by L:

d 2i R di 1 1 dvg
- vC +   i 
dt 2 L dt LC L dt
R 1
Characteristic Equation: s2 s 0
L LC 1
The undamped natural frequency of the (series RLC) circuit:  0  a0 
LC
a1 R
Series RLC Damping Ratio:  S  
2 a0 2 L C
R CS L
Critical series resistancecan be found from :  S   1  R CS  2
2 LC C
For R  R CS (  1), overdamped case; for R  R CS (  1), underdampe d case

21
Ali Muhtaroğlu
 Example : Series RLC Circuit with Numbers
R

i Let L=1 H , C=1/4 F:

vg L

- vC +

22
Ali Muhtaroğlu
 Example : Series RLC Circuit with Numbers
R
d 2i di
i Let L=1 H , C=1/4 F: 2
 R  4i  0
vg L
dt dt
Characteristic Equation: s  Rs  4  0
2
C
1 R R R
0   2 rad/s    
- vC + LC 2 LC 2 4 4
Critically damped case:   1  R  4 

Overdamped case:   R / 4  1 , for example assume R  5 

Underdamped case:   R / 4  1 , for example assume R  2 

Undamped case:   R / 4  0  R  0 

23
Ali Muhtaroğlu
 Example : Series RLC Circuit with Numbers
R
d 2i di
i Let L=1 H , C=1/4 F: 2
 R  4i  0
vg L
dt dt
Characteristic Equation: s  Rs  4  0
2
C
1 R R R
0   2 rad/s    
- vC + LC 2 LC 2 4 4
Critically damped case:   1  R  4 
s 2  4s  4  0  s    2 (real and identical roots)  i n  K1e 2t  K 2te2t
Overdamped case:   R / 4  1 , for example assume R  5 
s 2  5s  4  0  s1  4, s2  1 (real and distinct roots)  i n  K1e t  K 2e 4t
Underdamped case:   R / 4  1 , for example assume R  2 
 b  b 2  4c
s  2s  4  0  s1, 2 
2
 s1  -1 j 3 , s2  -1 j 3 (complex conjugates)
2
 i n  K1e t cos 3t  K 2e t sin 3t
Undamped case:   R / 4  0  R  0 
s 2  4  0  s1  j 2, s2   j 2 (imaginary roots)  i n  K1 cos 2t  K 2 sin 2t
24
Ali Muhtaroğlu
 Example : Series RLC Natural Response Sketches
R

i
vg L

- vC +
Overdamped case: Critically damped case:

t t
Underdamped case: Undamped case:

t
25
Ali Muhtaroğlu
 Forced Response
• The method is similar to the first2 order circuits
d xf dx f
• Forced differential equation: dt 2 1 dt  a0 x f  f (t )
 a

• The forced response is identified through the general linear

combination of f(t) and all of its derivatives.
– If f(t) is a constant, e.g. an independent source, use a trial forced solution
that is a constant, A. [This is what was done with first order circuits so far.]
– What if f(t) is not constant ?
Examples in Table 7.1 Forcing Term f(t) Trial Forced Solution xf(t)
k A
(Jonhnson & Johnson)
t At+B
– In addition:
tn Atn + Btn-1 +...+ Ft + G
When the trial forced solution
est Aest
matches a term in the natural
sin(ωt), cos(ωt) Asin(ωt)+Bcos(ωt)
solution, it is necessary to
.... ....
t-multiply the trial forced
solution.
26
Ali Muhtaroğlu
 SRC half-bridge configuration.

Depending on the ratio between the switching frequency ωS and the

converter resonant frequency ωr , the converter has several possible
operating modes.
Circuit waveforms under different operating
conditions:
(a) ωS < 0.5 ωr ;
(b) (b) 0.5 ωr < ωS < ω; and
(c) (c) ωr < ωS .
A. Discontinuous Conduction Mode (DCM) with ωS < 0.5ωr
Figure shows the waveforms of iLr and the resonant capacitor voltage vCr in this mode of
operation. From 0 to t1, T1 conducts. From t1 to t2, the current in T1reverses its direction. The
current flows through D1 and back to the supply source. From t2 to t3, all switches are in the off
state. From t3 to t4, T2 conducts. From t4 to t5, the current in T2 reverses its direction. The
current flows through D2 and back to the supply source. T1 and T2 are switched on under ZCS
condition and they are switched off under zero-current and zero-voltage conditions. However,
the switches are under high current stress in this mode of operation and thus have higher
conduction loss.
B. Continuous Conduction Mode (CCM) with 0.5ωr < ωS < ωr

Figure shows the circuit waveforms. From 0 to t1, iLr transfers from D2 to T1. T1 is
switched on with finite switch current and voltage, resulting in turn-on switching loss.
Moreover, the diodes must have good reverse recovery characteristics in order to
reduce the reverse recovery current. From t1 to t2, D1 conducts and T1 is turned off
softly with zero voltage and zero current. From t2 to t3, T2 is switched on with finite
switch current and voltage. At t3, T2 is turned off softly and D2 conducts until t4.
C. Continuous Conduction Mode (CCM) with ωr < ωS
Figure shows the circuit waveforms. From 0 to t1, iLr transfers from D1 to T1. Thus, T1
is switched on with zero current and zero voltage. At t1, T1 is switched off with finite
voltage and current, resulting in turn-off switching loss. From t1 to t2, D2 conducts.
From t2 to t3, T2 is switched on with zero current and zero voltage. At t3, T2 is
switched off. iLr transfers from T2 to D1. As the switches are turned on with ZVS,
lossless snubber capacitors can be added across the switches.
M vs γ in SRC.
A family of quasi-
resonant converter
with ZCS.
A family of quasi-resonant converter with ZVS.
 Equations for p-pulse Converter (Continued)

A general expression for the mean load voltage of a p-pulse fully-controlled

rectifier, including the effects of the overlap angle y, can be determined by
reference to Figure below;
 Regulation

The term regulation is used to describe the characteristic

of equipment as it is loaded. In the case of the rectifier,
regulation describes the drop in mean voltage with load
relative to the no-load or open-circuit condition.

There are three main sources contributing to loss of

output voltage:

1. The voltage drop across the diodes and/or thyristors.

2. The resistance of the a.c. supply source and
conductors.
3. The a.c. supply source inductance.
These three voltage drops can be represented respectively by the three
resistors R1, R2 and R3 in the equivalent circuit of Figure below. The open-
circuit voltage is given by Vo and the actual load voltage by VL.

If the load current IL is taken to be level, that is, a pure direct current, then any
voltage drop can only be represented by resistors.
Regulation for Three Phase Half Wave Rectifier

Voltage Drop Due to overlap 3ωL/2π

 Load Aspects
A d.c. load fed from a rectifier will in its voltage have a harmonic voltage
content, the lowest order of which is the pulse-number of the rectifier. Harmonics
at multiples of the pulse-number will also exist.

The controlled rectifier will have a higher harmonic content in the load voltage than
when the rectifier is uncontrolled.

The harmonic voltages present in the voltage waveforms will inevitably

give rise to harmonic current of the same frequency in the load.

The presence of harmonics can distort the supply waveforms and cause
problems in respect to firing circuits which rely on the supply voltage zero as
reference to the firing delay angle. Such distortion can cause slight differences
in the firing delay from one device to the next, setting up a phenomenon known
as jitter. The jitter frequency would be in the radio-frequency range, and hence
worsen the radio-interference problem.
 Supply Aspects

The switching action of the rectifying device inevitably results in non-sinusoidal

current being drawn from the a.c. supply system. In essence, the a.c. supply
delivers a sinusoidal voltage with power flow relating only to the fundamental
(mains) frequency. The load then converts some of this power to higher frequencies,
and transmits harmonic power back into the supply system. Hence a rectifying
load acts in part as a harmonic generator.

Taking the case of an infinite load inductance, and consequently level load current, the
magnitudes of the harmonic current components are inversely proportional to their
frequency an uncontrolled rectifier.
 Supply Aspects

For the ideal p-pulse rectifier, taking level load current, no losses,
and no overlap, then only harmonics exist of the order

r = mp ± 1

Controlled Rectifier

The assumption of level load current is frequently not justified sufficiently

for the above statements to be accurate, and in practice, particularly at large
firing delays, the harmonic components differ from the simple relationship of
l/r to the fundamental.
Example 1

A single-phase diode bridge circuit is supplied at 50 V, 50 Hz to feed a d.c. load

taking a level current of 60 A. Determine the volt-drops due to: (i) the supply
having an inductance of 0.1 mH, (ii) each diode having a forward volt-drop of
(0.6 + 0.002i) volts, (iii) supply and lead resistance of 0.002 Ω.
Draw an equivalent circuit to represent the rectifier.
Example 2

A three-phase fully-controlled bridge converter is connected to a 415 V supply,

having a reactance of 0.3 Ω/phase and resistance of 0.05 Ω /phase. The converter
is operating in the inverting mode at a firing advance angle of 35°. Determine
the mean generator voltage, overlap angle, and recovery angle when the current
is level at 60 A. Assume a thyristor volt-drop of 1.5V.
Example 3

Determine the Fourier series for the load voltage of the 2-pulse
uncontrolled rectifier taking the mean level to be 100 V. Neglect overlap and
losses. Also determine the ripple factor for this configuration.
Example 4
Example 5
Example 6
Example 7