Professional Documents
Culture Documents
fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
AbstractA single-phase four-switch rectifier with consider- as electrical vehicles [4] and aircraft power systems [5].
ably reduced capacitance is investigated in this paper. The What is worse is that electrolytic capacitors, known to have
rectifier consists of one conventional rectification leg and one limited lifetime, are one of the most vulnerable components in
neutral leg linked with two capacitors that split the DC bus.
The ripple energy in the rectifier is diverted into the lower split power electronic systems [6], [7], [8]. As a result, in order to
capacitor so that the voltage across the upper split capacitor, enhance the reliability of power electronic systems, it is highly
designed to be the DC output voltage, has very small ripples. The desirable to minimise the usage of electrolytic capacitors and
voltage across the lower capacitor is designed to have large ripples use highly-reliable small capacitors like film capacitors if
on purpose so that the total capacitance needed is significantly possible, while maintaining low voltage ripples.
reduced and highly reliable film capacitors, instead of electrolytic
capacitors, can be used. At the same time, the rectification leg In general, the reduction of electrolytic capacitors can
is controlled independently from the neutral leg to regulate be achieved in four approaches. One approach is to inject
the input current to achieve unity power factor and also to harmonic currents to suppress fluctuations of input energy
maintain the DC-bus voltage. Experimental results are presented by changing control strategies for existing power switches
to validate the performance of the proposed strategy. in rectifiers. In [9], it was proposed to reduce the DC-bus
Index TermsSingle-phase rectifiers, voltage ripples, electro- capacitor by injecting third harmonic component to the grid
lytic capacitors, neutral leg, reliability, ripple eliminator. current. This approach benefits from fewer switches and easier
implementation, which lead to lower system costs compared
I. I NTRODUCTION to other solutions. The second approach is to add an active
More and more microgrids are now connected to the public energy storage compensator in parallel with DC-bus capacitors
grid and various loads through power converters [1]. For to bypass ripple energy that originally flows into DC-bus
both AC and DC microgrids, single-phase rectifiers are often capacitors [5], [10], [11], [12], [13], [14], [15], [16]. This
needed when supplying DC loads. Such rectifiers are expected has been extensively studied in the last few years. Normally,
to have high power density, high efficiency, high reliability and the added compensator is operated as buck/boost converters to
low costs. There are numerous topologies in the literature, aim- inject/absorb ripple currents from DC bus. The third approach
ing to have improved performance from these three aspects. is based on connecting an active compensator in series with the
Moreover, with the integration of renewable energy sources DC bus [17]. The compensator basically behaves as a voltage
into the power grid, there is a trend to have bidirectional source to offset voltage ripples. Due to the series connection,
single-phase power converter as an interface between power the compensator has lower voltage stress compared to parallel
grid and energy sources [1], [2], [3]. As a result, the study of compensators. The last approach is to introduce a ripple port
single-phase rectifiers has attracted more and more attention. terminated with a capacitor, as reported in [8], [18], to store the
Conventionally, bulky electrolytic capacitors are required for ripple power. Different from other solutions, an AC capacitor
single-phase rectifiers to produce smooth DC-bus voltage, due instead of a DC capacitor is used to handle ripple energy,
to the pulsating input power. However, the volume and weight which also reduces the voltage stress on the switches.
of bulky electrolytic capacitors could be a serious problem Following the preliminary conference version of this paper
for volume-critical and/or weight-critical applications, such [19], a 4-switch rectifier is proposed to significantly reduce
the DC-bus capacitance in the widely-adopted asymmetrical
Manuscript received October 19, 2014; revised December 19, 2014 and single-phase systems, where the midpoint of the AC side is
February 1, 2015; accepted February 18, 2015. not available. The rectifier only uses four switches, which
This work was partially supported by the Engineering and Physical Sciences
Research Council, U.K., under Grant No. EP/J01558X/1. A preliminary is similar to a conventional bridge PWM rectifier, but the
version of this paper was presented at the 5th International Symposium on switches are formed as a rectification leg and a neutral leg and
Power Electronics for Distributed Generation Systems (PEDG2014) held in operated differently from a conventional full-bridge rectifier.
Galway, Ireland, in June 2014.
Wen-Long Ming is with the Department of Automatic Control and Systems The rectification leg is operated as a half-bridge rectifier to
Engineering, The University of Sheffield, Sheffield S1 3JD, United Kingdom. regulate the DC-bus voltage via controlling the grid current to
(e-mail: wenlongming@gmail.com) make it clean and in phase with the grid voltage to achieve
Q.-C. Zhong is with the Department of Electrical and Computer Engineer-
ing, Illinois Institute of Technology, Chicago, IL 60616, USA and also with the unity power factor. The neutral leg, consisting of two active
Department of Automatic Control and Systems Engineering, The University switches, two split capacitors and one inductor, maintains a
of Sheffield, Sheffield S1 3JD, United Kingdom. (e-mail: zhongqc@ieee.org) stable DC output voltage. The control of the two legs are
Xin Zhang is with the Department of Automatic Control and Systems
Engineering, The University of Sheffield, Sheffield S1 3JD, United Kingdom. independent from each other, which makes control design
(e-mail: xin.zhang@sheffield.ac.uk) very flexible, and the corresponding control strategies can
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
be designed according to their own objectives. Importantly, order voltage ripples on the DC bus. However, the reliability,
the neutral leg is able to divert the ripple energy from the volume and weight of electrolytic capacitors could be a serious
upper split (output) capacitor to the lower split capacitor. As a problem for high-reliability, volume-critical and weight-critical
result, the output voltage does not contain any low frequency applications [6], [7], [8]. As a result, in order to enhance
ripples and hence, the upper split capacitor can be significantly the reliability and power density of rectifiers, it is highly
reduced. Note that the voltage across the lower split capacitor desirable to reduce the usage of capacitors so that highly-
is designed to have relatively large ripples on purpose because reliable capacitors like film capacitors could be used to replace
it is not supplied to any loads. Accordingly, the total usage of electrolytic capacitors. However, for conventional single-phase
DC-bus capacitors could be reduced significantly so that it is rectifiers, there exists a trade-off between reducing required
now possible to use highly reliable film capacitors, instead capacitors and reducing the output voltage ripples because
of bulky electrolytic capacitors. This makes the rectifier very single-phase full-bridge rectifiers have only one DC voltage,
suitable for high-reliability applications. The selection criteria which is used as both the DC output and the only ripple energy
of the split capacitors are discussed with the aim to minimise buffer on the DC bus. In light of this, a lot of auxiliary circuits
their usage. This is mainly for systems without hold-up time are proposed to construct another DC or AC voltage to store
requirement. For systems with hold-up requirement, the re- the voltage ripples, which can be connected in parallel or in
quired capacitance needs to be large enough if no other means series at the AC or DC sides [5], [8], [13], [23], [24].
is applied to provide the energy required [17]. For the topology shown in Figure 1, there are two DC
The rest of the paper is organised as follows. Section II voltages because of the split capacitors. This provides a
introduces the rectifier under investigation. In Section III, how possible way to operate the rectifiers to make one of the
to significantly reduce DC-bus capacitors is discussed and, in voltages as the output voltage to supply loads and to make the
Section IV, the associated control strategies are developed. In other voltage as the ripple energy buffer. The total capacitance
order to achieve the minimal capacitance, the selection criteria could be significantly reduced because the ripple energy is
of the split capacitors are then discussed in Section V and the diverted from the output capacitor to the other capacitor, which
impact of the different voltages across the split capacitors are could have high voltage ripples. By diverting all the ripple
analysed in Section VI. Experimental results are provided in power to the lower capacitor C , the output voltage V+ can
Section VII and the conclusions are made in Section IX. become ripple free, which means the output capacitance C+
can be reduced a lot because it does not need to process any
II. T HE S INGLE - PHASE R ECTIFIER UNDER I NVESTIGATION low frequency ripple energy. Importantly, the capacitor C
The rectifier proposed in the preliminary version of this can also be significantly reduced because its voltage is not
paper [19] is investigated further in this paper. It consists of supplied to any loads so it can be designed to have large ripples
one rectification leg and one neutral leg, as shown in Figure on purpose. Accordingly, both capacitors can be significantly
1. The rectifier can be formed by adding two active switches reduced and replaced with highly-reliable film capacitors. This
into a conventional half-bridge PWM rectifier by putting a improves the system power density and reliability and reduces
neutral leg consisting of two switches across the DC bus system weight and volume. Although costly film capacitors are
with their midpoint connected to the midpoint of the split used to replace electrolytic capacitors, the cost arising from
capacitors through an inductor. The neutral leg is actually a capacitors could still be reduced because the total capacitance
typical DC/DC converter, which has been widely adopted in required is considerably reduced.
industry. In particular, the neutral leg has been applied to three-
phase four-wire power inverters as reported in [1], [20], [21], III. R EDUCTION OF THE B ULKY DC- BUS C APACITORS
[22]. According to the analysis made in [1], the neutral leg is In order to clearly show how to significantly reduce the
a stable system although the inductor is coupled with the split DC-bus capacitors, there is a need to analyse the relationship
capacitors. between the ripple energy and the required capacitors for
the investigated rectifier. For this purpose, an average circuit
I IR model is built up at first.
Q1 Q3 C+
iC+ V+ R
A. Circuit Analysis
uN iC It is assumed that the DC-bus voltage of the rectifier is
A B
Lg iL LN N VDC = V+ + V (1)
ig C-
V- where V+ and V are the voltages across the split capacitors
~ Q2 Q4 iC- C+ and C with respect to the neutral point N and the
vg negative point of the DC bus, respectively. Suppose that the
grid current is
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
Q3
V+ * - 1
PI PWM Q4 vg sin(t)
VDC PLL
- ig* Q1
V+*+ V-max* Repetitive
H(s) PI PWM
Controller Q2
- -
V+
+ ig
Repetitive Controller
0 - V-max
Kr
- i + +
i
BPF e d s LPF LPF Peak
s + i Extraction
I of V-max
KR(s)
0 KR(s)
- V+ V-
V-
Figure 4. Controller for the rectification leg.
Figure 3. Controller for the neutral leg.
In general, the adoption of the BPF does not lead to any B. Control of the Rectification Leg
resonance of the controllers with the rectifier. This is mainly
due to the fact that the BPF behaves as a low-pass filter at The control of the rectification leg is very similar to that
high frequencies and is cascaded with the repetitive controller, of conventional half-bridge rectifiers, which is mainly used
which again is a low-pass filter. to regulate the grid current and to control the whole DC-
3) Removal of fundamental component in iC : The control bus voltage. To be more precise, the grid current is expected
of the DC-bus ripple current i to 0 leads to the fact that the to be in phase with the grid voltage and also to be clean
ripples are now diverted to the lower capacitor C . In this with low harmonics. For this purpose, the grid current ig
case, the current of the capacitor C is expected to only have should be measured as a feedback to form a current tracking
a second-order component. However, according to (6), when controller. Here, the repetitive controller shown in the dashed
i = 0, there is box of Figure 3 is adopted again. In order to generate the grid
ig = iL iC + IR , current reference ig , an outer-loop voltage controller can be
constructed.
which means that the grid current ig could flow through the There can be different ways to construct this voltage con-
inductor LN and the capacitor C if not controlled properly. troller; see e.g. [33]. In this paper, the voltage controller is
Hence, there is a need to make sure that no fundamental com- designed to maintain the maximum voltage Vmax of V
ponent flows through the capacitor C otherwise it would lead constant. Hence, the total DC-bus voltage is maintained at
to increased voltage ripples without providing any benefits. V+ + Vmax
, with a PI controller. The output of the controller
This can be achieved by forcing the fundamental component can be used as the peak value of the grid current reference
of V to be zero, as shown in Figure 3. The following resonant ig , as shown in Figure 4. This is multiplied with the phase
controller signal of the grid voltage, which can be obtained from a phase-
Kh 2hs
KR (s) = 2, (16) locked-loop, to form the grid current reference ig . As a result,
s2 + 2hs + (h) the grid current is in phase with the grid voltage to achieve
with = 0.01, h = 1, and = 2f , can be adopted. the unity power factor. Here, the phase-locked-loop proposed
The output of the resonant controller is then added onto the in [34] is adopted.
outputs of the other two controllers before sending to the The above-mentioned control strategy of the half-bridge
PWM conversion block, as shown in Figure 3. The gain Kh of rectification leg is now somewhat standard [26], [28], [29],
the resonant controller can be selected by fine tuning through [35]. What is different here is that the maximum DC-bus
trial-and-error in practice; it is chosen as Kh = 10 for the voltage, instead of the average DC-bus voltage, is selected
experimental system to be tested. In general, a large gain as the controlled output. As a result, the objective here is set
should improve the performance of the control but may lead to control the maximum DC-bus voltage. For this purpose,
to a large charging current when starting up the system that the maximum DC-bus voltage should be extracted at first.
might trigger the current protection and also may introduce Since the voltage V+ is controlled to be more or less pure
noticeable disturbance into the current controller. These should DC without second-order components, it is only required to
be avoided. Note that the output of this controller is + extract the maximum value of the voltage V , which can be
because the voltage under control relates to V . obtained by adding the DC component with the peak voltage
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
of the ripple component, as shown in Figure 4. The hold filter capacitor is mainly limited by the allowed maximum voltage
(14) is again used to obtain the DC component. In order to Vamax and the required minimum capacitance Cmin is
extract the second-order component, the resonant filter (16) Vg I g
is again adopted with = 0.01, h = 2, and = 2f . A C = 2 2 , (20)
(Vamax Vmin )
Peak block in Figure 4 is used to calculate the peak value of
the ripple component. The sum of the average voltage and the which can be small if Vamax is high enough.
peak voltage of the ripple component then forms the maximum In addition, another important factor for selecting capacitors
voltage of the voltage V , which is denoted as Vmax in is the maximum allowable ripple currents [4], [7], [36]. This
Figure 4 and is added with V+ to obtain the maximum DC-bus is very important for the reliability of capacitors. In general,
voltage for feedback. large current ripples lead to short lifetime. The current ripples
are closely related to the voltage ripples and the equivalent
C. Stability of the System impedance of capacitors. In order to evaluate the level of
voltage ripples, (20) can be rewritten as
Because of the decoupled nature of the controllers for the
two legs, the stability of the system can be easily guaranteed. Vg I g
C =
The controller for the rectification leg has a very typical V (Vmax + Vmin )
structure, which is very mature and widely used in industry. Vg I g
= (21)
The controller for the neutral leg has a very special structure 2V Vave
with one current loop and two voltage loops. What is special is
that these three loops are in parallel rather than cascaded so the where V = Vmax Vmin and Vave = Vmax +V 2
min
stability of each loop can be treated individually. The current are the peak-peak ripple voltage and the average voltage of
loop is designed to regulate the AC components of iC+ (or i) V , respectively. Since the capacitor impedance at the second-
1
to be around zero, i.e., to remove any non-DC components in i. order frequency is 2C
, the peak-peak value iC of the
At the same time, the voltage loop related to V+ is to maintain second-order ripple current flowing through C is
the DC component of the voltage V+ while the voltage loop V
iC = 1 = 2C V . (22)
related to V is designed to reduce the fundamental component
2C
of voltage V . Hence, the functions of the three loops are
decoupled in the frequency domain for current or voltage. Substitute (21) into (22), then there is
The three loops consist of simple PI, repetitive and resonant Vg I g
controllers, which have been widely analysed in the literature. iC = . (23)
Vave
See, for example, [1]. Hence, detailed analysis of the stability
This is consistent with (13). The average voltage should
of the loops is not repeated in this paper.
be increased in order to reduce the ripple current. For the
proposed strategy, the voltage V can be different or the same
V. S ELECTION OF C OMPONENTS as V+ . As a result, the voltage Vave can be maintained at
A. Selection of Capacitor C a higher value in order to reduce the ripple current. This
As demonstrated in [9], [24], the total ripple energy stored can be naturally achieved when the maximum voltage Vmax
in the split capacitors over a charging period for single-phase is controlled at the allowable value because the higher the
rectifiers with the unity power factor is maximum voltage is, the higher the average voltage Vave is.
Of course, some other factors such as hold-up time require-
Vg I g
Er = . ment [17], current stress and limited voltage rating of the
2 capacitors and switches, should be taken into account when
With the proposed strategy, all the ripple energy is now stored selecting capacitors. If the maximum voltage of the capacitor
on the lower capacitor C instead of both capacitors C+ and is determined, then increased capacitance means increased
C . Hence, hold-up time and reduced current stress, which is preferred
2Er in practical applications. As a result, there are several trade-
C = 2 2 (17) offs when selecting capacitors for a certain application.
Vmax Vmin
where Vmax and Vmin are the maximum and minimum
voltages of V , respectively. A small capacitor means that B. Selection of Inductor LN
high Vmax and/or low Vmin is needed. However, in order The fast switching of the neutral leg leads to switching
to ensure the proper boost operation of the rectifier, ripples over the current flowing through the inductor LN .
Since the two switches Q3 and Q4 are operated complement-
V > Vg | sin t| (18)
arily, the on time of Q3 is dfs3 and the on time of Q4 is 1d fs
3
should be satisfied. In other words, in one PWM period. Since the switching frequency is much
higher than the line frequency, it can be assumed that the
Vmin > Vg . (19) current increased LV+Ndf3s (to withstand the positive voltage V+ )
At the same time, Vmax has an upper bound as well because and the current decreased VL(1dN fs
3)
(to withstand the negative
of the limit on the devices and/or the applications. Hence, the voltage V ) in these two modes are the same. According to (8),
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
Since V+ , V > Vg , then d2min > 0 and d2max < 1 can be 1) Switches and diodes of the rectification leg: For the
achieved for any combinations of V+ and V . According to rectification leg, there are two switches and two diodes in
the average model, the duty cycle of the switch Q2 is total. The current flowing through the rectification leg mainly
depends on the grid current ig . The positive cycle of the grid
V+ Vg current ig flows through the Switch Q2 and the corresponding
d2 = sin t
VDC VDC free-wheeling diode is D1 . On the other hand, the negative
1 cycle of the grid current ig flows through the Switch Q1
= (V+ Vg sin t) (28)
V+ + V and the the corresponding free-wheeling diode is D2 . As
demonstrated in [26], the average currents flowing through
If all the ripple power is provided by the lower capacitor, then
active switches Q1 and Q2 and diodes D1 and D2 are
(28) becomes
Z 2
1 1 2V
d2 = (V+ Vg sin t) I Q1 = ig (1 d2 )dt = IR ( 0.5)
q
Po
2 Vg
V+ + 2 +
Vmin C (1 sin 2t) Z
1 2V+
I Q2 = ig d2 dt = IR ( 0.5) (29)
2 0 Vg
where the derivation of V can be found in [24] for a Z
1 2V
given load power Po . It is clear that the obtained duty cycle I D1 = ig (1 d2 )dt = IR ( + 0.5)
contains a second-order ripple component coming from V . 2 0 Vg
Z 2
The existence of a second-order ripple component is common 1 2V+
I D2 = ig d2 dt = IR ( + 0.5).
for all rectifiers based on the half-bridge structure and does 2 Vg
not constitute any problem because the switching frequency
is much higher than the second-order frequency. The only It can be seen that most of the currents flow through the diodes
difference here is that the ripples are stored in the lower rather than the active switches. More importantly, the currents
capacitor only. of the active switches are different if V+ 6= V . The same is
Because the rectification leg is independently controlled, true for the diode currents. For example, if V > V+ , which is
the input power factor and the THD of the input current can preferred in order to reduce C , then IQ1 > IQ2 and ID1 >
be regulated as usual and are not affected by the difference ID2 . As a result, the power loss of the upper switch Q1 is
between V+ and V . As a result, the regulation of the input higher than that of the lower switch Q2 if V > V+ . (29) can
current is not affected by the voltage difference between V+ be used as a principle to select the active switches and diodes.
and V . Of course, the two voltages can be controlled to be the same.
In this case, the average currents of the switches become the
same and also, the average currents of the switches become
B. Impact on the Neutral Leg the same too.
2) Switches and diodes of the neutral leg: For the neutral
The neutral leg is used for two purposes, i.e. splitting the leg, there are also two switches and two diodes in total. The
DC-bus voltage to V+ and V and diverting the ripple power current flowing through the neutral leg mainly depends on the
to the lower capacitor C . According to the average model, inductor current iL . In order to analyse the average currents,
the duty cycle of the switch Q3 can be given as similar analysis can be done. The only difference here is the
conduction periods of Q3 , Q4 and D3 , D4 . For example, the
V+
d3 = 1 conduction period of Switch Q1 in the rectification leg is from
VDC to 2, which is not affected by other factors like the input
V+ or output power. This is because the periods of the positive
= 1 q ,
2 Po
V+ + Vmin + C (1 sin 2t) and negative cycles of the current ig are the same, which is
. However, it is obvious that those periods of the current iL
which is also affected by a second-order ripple component. As are not the same according to (12), which leads to the fact
long as V < VDC , which is always true because VDC = V+ + that the conduction periods of Q3 , Q4 and D3 , D4 are not the
V > V+ , V , the duty cycle d3 can be always achieved by same. In order to calculate their average currents, there is a
controlling the two switches Q3 and Q4 in an complementary need to first know these periods. Let iL = 0, then
way. Hence, the voltage difference does not cause any problem
Vg I g Vg I g
to the control of the neutral leg either. Ig sin t cos 2t = 0,
2V 2V+
or
C. Impact on the Current Stress of the Switches V 1 V
sin2 t + sin t = 0.
Vg 2 2V+
The voltage difference may lead to different current stresses
to the switches. The average currents are very important when Hence,
selecting a switch or a diode [26]. As a result, they are s
calculated here in order for selecting suitable switches and V 1 V2 2V
sin t = 2
+2+ .
diodes for both legs. 2Vg 2 Vg V+
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
Because of (19), the sign is not valid. There are two can be obtained according to (20). The higher the voltage stress
solutions within [0, ], which are can be, the smaller the capacitance C can be. It is worth
s mentioning that high voltage stress leads to high switching
1 1 V2 2V V loss, which decreases the efficiency. When large electrolytic
t1 = arcsin( +2+ ),
2 Vg2 V+ 2Vg capacitors are used, the voltage stress could be lower because
s Vmax could be reduced. In this case, the switching loss of the
1 1 V2 2V V rectifier is lower. However, the loss caused by the ESR of the
t2 = arcsin( +2+ ).
2 Vg2 V+ 2Vg capacitors becomes higher because the required capacitance
As a result, the average currents flowing through active can be very large and electrolytic capacitors have to be used.
switches Q3 and Q4 and diodes D3 and D4 can be calculated The additional loss caused by the high voltage stress may
from have been well compensated by the reduction of the loss in
Z t2 capacitors.
1
I Q3 = iL d3 dt
2 t1
Z t1 +2 E. Impact on Switching Ripples of the Grid Current
1
I Q4 = iL (1 d3 )dt (30) Since the switching frequency is much higher than the
2 t2
Z t1 +2 fundamental frequency, the average grid current over each
1 switching period can be controlled to track its reference, which
I D3 = iL d3 dt
2 t2 is a sinusoidal signal. Apart from controlling the average grid
Z t2
1 current, it is also desirable to maintain the switching ripple of
I D4 = iL (1 d3 )dt. the grid current under a certain level, in order not to introduce
2 t1
power pollution to the grid. According to [26], the peak-peak
The analytical solutions are complicated but, in principle, the switching ripple of the grid current can be given as
currents ID3 and ID4 are again higher than the currents IQ3
and IQ4 . Moreover, because V is set higher than V+ , IQ3 V
and ID3 are higher than IQ4 and ID4 , respectively. VDC VDC + Vg sin t
ig = d2
For certain applications with known system parameters, the Lg f s
average currents flowing through the switches and diodes of 1
both legs can be easily obtained based on the above analysis. = (V+ V Vg2 sin2 t) +
Lg fs VDC
Together with the voltage stress, i.e. the DC-bus voltage VDC , (V+ V )Vg
suitable switches and diodes can be selected. sin t. (31)
Lg fs VDC
Apparently, increasing the inductor and/or increasing the
D. Impact on the Voltage Stress of the Switches
switching frequency could reduce the switching current ripple.
In order to choose suitable switches for both legs, the According to (31), the switching ripple current is related to al-
voltage stress of the switches is another factor to be con- most all system parameters. Compared to most of the analysis
sidered. in the literature, the only difference here is that the voltages
Compared to the current stress of the switches, it is more V+ and V are designed to be different on purpose. The first
straightforward to analyse the voltage stress of the switches. Vg2
part of the switching ripple, i.e. Lg1fs ( VV+DC
V
VDC sin2 t), is
Regardless of other system parameters, the voltage stress of
always positive in the positive and negative half cycles of the
the switches is always the sum of the voltages V+ and V . It (V V )V
grid current. However, the second part, i.e. L+g fs VDC g sin t,
is well known that low voltage stress generally leads to low
changes its sign during the positive and negative cycles of the
costs and high efficiency. As a result, it is always hoped to
grid current. For example, if V+ < V , it is negative in the
maintain the voltage stress within a reasonable level. Either
positive half cycle of the grid current but is positive in the
decreasing V+ or V can help to reduce the voltage stress.
negative half cycle of the grid current. The resulted effect is
The level of the voltage V+ is determined by the requirement
that the switching ripples of the grid current in the negative
of the DC load and cannot be changed. However, the voltage
half cycle are larger than those in the positive half cycle. If
V does have some freedom to be decreased. According to
V+ > V , then the switching ripples of the grid current in the
the discussion before, the proposed rectifier can still perform
negative half cycle are smaller than those in the positive half
well to control the grid current and the DC output voltage as
cycle. However, the slightly different switching ripples does
long as the minimum voltage of V is higher than the peak
not constitute any noticeable problems to the grid. If needed,
of the grid voltage. Hence, there is Vmin = Vg . Note that
an LCL filter, instead of an inductor Lg , can be adopted so
the ripple level of the output voltage V+ is still very low even
that the switching ripple currents can flow through the filter
if Vmin = Vg . The only compromise here is the DC-bus
capacitor instead of the grid.
capacitance. In order to meet a given maximum voltage stress
Vmax , the maximum voltage of V is then fixed, which is
Vmax = Vmax V+ . It is clear that low voltage stress means VII. E XPERIMENTAL VALIDATION
low Vmax . Since both maximum and minimum values of the In order to validate the design and operation of the rectifier,
voltage V are fixed now, the minimum required capacitance experiments were conducted on a test rig in the lab. The test
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
10
A. Steady-state Performance
t: [10 ms/div]
1) The grid current ig and the DC voltages V+ and V :
The system steady-state performance with V+ = 200 V is (b)
given in Figure 5(a)-(d) for Vmax = 600, Vmax = 650,
Vmax = 700 and Vmax = 750, respectively. It is clear that V-: [250 V/div]
the DC output voltage V+ is always maintained around its
350 V
reference 200 V while the ripple voltage of V varies from 700 V
V+: [250 V/div]
337 V to 431 V depending on the maximum voltages of V . 5V
vg: [103 V/div] ig: [8.33 A/div]
Importantly, the voltage ripples of the voltage V+ are only
about 5 V when Vmax = 700 V and 750 V. As a result,
nearly all the ripple power is now stored on the lower capacitor
C instead of both C+ and C over a wide range of V . It t: [10 ms/div]
is worth again pointing out that only two 5 F are used in the
system. The reduction of capacitors and ripples on the output (c)
V+ have been achieved at the same time.
In order to clearly illustrate the relationship between the V-: [250 V/div]
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
11
500
450
402 V 600 V
400 9
350 V iC-: [1 A/div] V-: [250 V/div]
337 V
Ripple Voltage (V)
350
1.4 A
300 V
iC+: [2 A/div]
0.04 A i: [0.54 A/div]
250 185000/Vave
0.1 A
200 V+
150 t: [10 ms/div]
100
(a)
50 12 V 7V 5V 5V
0
350 400 450 500 550 600
Average Voltage Vave (V) 650 V
Current 700 V
Distortion t: [10 ms/div] V-: [250 V/div]
iC-: [1 A/div]
1.1 A
Figure 7. Deteriorated system performance with V+ = 200 V when iC+: [2 A/div]
i: [0.54 A/div]
0.02 A
Vmax
= 500 V.
0.06 A
t: [10 ms/div]
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
12
1.6 start-up are shown in Figure 12. The grid current first increased
to charge the capacitors and then the current was maintained
1.4 1.3 A
1.4 A well back to its steady-state value after the DC output voltage
1.1 A
1.2
1.05 A was settled. The system start-up took about 200 ms, which is
Ripple Current (A)
t: [400 ms/div]
(a)
Figure 13. Transient response when the reference of the voltage V+ was
V+: [250 V/div] changed from 200 V to 300 V.
f=100 Hz
FFT spectrum of I
2) Change of the voltage reference: When the reference
of the voltage V+ was changed from 200 V to 300 V, the
results are shown in Figure 13. The voltage V+ was smoothly
I: [2 A/div] increased from 200 V to 300 V without any spikes. It is worth
f: [50 Hz/div] t: [400 ms/div] highlighting that the ripple level of the output voltage V+ is
always small during the transient period. However, the ripples
(b) of the voltage V became larger in order to tackle the increased
ripple power caused by the increased voltage reference (and
Figure 11. Comparison of (a) without and (b) with the repetitive current
controller for the neutral leg.
the power). This transient response took about 2 s, which
is limited by the allowable maximum neutral current of the
experimental system, and could be made much faster if the
Moreover, the spectra of the DC-bus current I are shown in allowable maximum neutral current is increased. For the test
Figure 11(a) and Figure 11(b) to demonstrate the performance rig, the neutral current is limited by the neutral inductor, which
of reducing the second-order ripples in the current I for the would be saturated if the neutral current exceeds about 5 A.
cases without and with the repetitive controller, respectively. 3) Hold-up time: Although the DC-bus capacitors are
It is obvious that the second-order harmonic component, i.e. designed for systems without hold-time requirement, it is
100 Hz, in the current I is significantly reduced when the still interesting to see how the proposed rectifier responds
repetitive controller is enabled. Most of the 100 Hz component to a sudden AC power outage. Here, two experiments were
is diverted to the neutral leg from the output capacitor. Due conducted in order to show the system performance regarding
to the diverted 100 Hz current, both the ripples of the output to the hold-up time under different capacitors. In order for a
voltage V+ and DC-bus current I are considerably reduced as fair comparison, only the capacitor C+ was changed while the
shown in Figure 11(b). other system parameters were kept unchanged. With C+ = 5
F, the time for the voltage V+ decreased from 200 V to 0 V
is about 14 ms as shown in Figure 14(a). Of course, this time
B. Transient Performance is too short for systems with hold-up requirement. A simple
1) System start-up: In order to demonstrate the transient way to increase this time is to use a larger capacitor. The
response of the proposed system, the results during the system experimental result with a larger capacitor (C+ = 100 F) is
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
13
Ripple eliminator
DP
Ib Lb I ir IR
Qb1 Q1 Q3
V+ Qb2 Cb Vb Rb La
T D1 R
uN N C
A B i VDC
Lg L
LN +
ig
Q1
(a) ~
Q2 D2 Ca Va iC
vg Q2 Q4
- DN
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
14
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
15
[11] X. Zhang, X. Ruan, H. Kim, and C. K. Tse, Adaptive active capacitor [35] B.-R. Lin and T.-L. Hung, Single-phase half-bridge converter topology
converter for improving stability of cascaded DC power supply system, for power quality compensation, IEE Proceedings-Electric Power Ap-
IEEE Trans. Power Electron., vol. 28, no. 4, pp. 18071816, 2013. plications, vol. 149, no. 5, pp. 351359, 2002.
[12] Q.-C. Zhong, W.-L. Ming, X. Cao, and M. Krstic, Reduction of DC- [36] M. MAKDESSI, A. Sari, P. Venet, P. Bevilacqua, and C. Joubert,
bus voltage ripples and capacitors for single-phase PWM-controlled Accelerated ageing of metallized film capacitors under high ripple
rectifiers, in Proc. IECON 2012 - 38th Annual Conf. IEEE Industrial currents combined with a DC voltage, IEEE Trans. Power Electron., to
Electronics Society, 2012, pp. 708713. be published, early Access.
[13] S. Wang, X. Ruan, K. Yao, S.-C. Tan, Y. Yang, and Z. Ye, A flicker- [37] M. Salcone and J. Bond, Selecting film bus link capacitors for
free electrolytic capacitor-less acdc LED driver, IEEE Trans. Power high performance inverter applications, in Electric Machines and
Electron., vol. 27, no. 11, pp. 45404548, 2012. Drives Conference, 2009. IEMDC 09. IEEE International, 2009, pp.
[14] S. H. Pini and I. Barbi, A single-phase high-power-factor rectifier, based 16921699. [Online]. Available: http://ieeexplore.ieee.org/stamp/stamp.
on a two-quadrant shunt active filter, IEEE Trans. Power Electron., jsp?arnumber=5075431
vol. 26, no. 11, pp. 31313143, Nov. 2011. [38] N. Mohan, Power electronics: converters, applications, and design.
[15] S. Dusmez and A. Khaligh, Generalized technique of compensating Hoboken, NJ : John Wiley and Sons, 2003.
low-frequency component of load current with a parallel bidirectional [39] H. Fujita, A high-efficiency solar power conditioner using a zigzag-
DC/dc converter, IEEE Trans. Power Electron., vol. 29, no. 11, pp. connected chopper converter, in Power Electronics Conference (IPEC),
58925904, 2014. 2010 International, 2010, pp. 16811687.
[16] Y. Tang, Z. Qin, F. Blaabjerg, and P. Loh, A dual voltage control [40] D. Karschny, Wechselrichter, German Patent DE19 642 522 C1, Apr.,
strategy for single-phase PWM converters with power decoupling func- 1998.
tion, IEEE Trans. Power Electron., to be published, early Access. Wen-Long Ming received the
[17] H. Wang, H.-H. Chung, and W. Liu, Use of a series voltage com-
pensator for reduction of the DC-link capacitance in a capacitor- B.Eng. and M.Eng. Degrees in Auto-
supported system, IEEE Trans. Power Electron., vol. 29, no. 3, pp. mation from Shandong University,
11631175, 2014. Jinan, China, in 2007 and 2010, re-
[18] S. Harb, M. Mirjafari, and R. Balog, Ripple-port module-integrated
inverter for grid-connected PV applications, IEEE Trans. Ind. Appl.,
spectively. He is currently pursuing
vol. 49, no. 6, pp. 26922698, 2013. the Ph.D. degree in the Department of
[19] W.-L. Ming and Q.-C. Zhong, Single-phase half-bridge rectifiers with Automatic Control and Systems En-
extended voltage ranges and reduced voltage ripples, in Power Elec-
tronics for Distributed Generation Systems (PEDG), 2014 IEEE 5th
gineering, the University of Sheffield,
International Symposium on, 2014, pp. 16. Sheffield, U.K.
[20] Q.-C. Zhong, T. Green, J. Liang, and G. Weiss, H control of the He was with Center for Power Elec-
neutral leg for 3phase 4wire DC-AC converters, in Proc. of the 28th tronics Systems (CPES), Virginia Tech, Blacksburg, USA in
Annual IEEE Conference of Industrial Electronics (IECON), 2002.
[21] T. Hornik and Q.-C. Zhong, Parallel PI voltageH current controller 2012 as an academic visiting scholar. His research interests
for the neutral point of a three-phase inverter, IEEE Trans. Ind. focus on reliability of DC-bus capacitors, technology to reduce
Electron., vol. 60, no. 4, pp. 13351343, 2013. passive components, traction power systems, transformerless
[22] J. Liang, T. Green, C. Feng, and G. Weiss, Increasing voltage utilization
in split-link, four-wire inverters, IEEE Trans. Power Electron., vol. 24, PV inverters and neutral line provision in power electronic
no. 6, pp. 15621569, Jun. 2009. systems.
[23] H. Wang and F. Blaabjerg, Reliability of capacitors for DC-link Qing-Chang Zhong (M04-
applications in power electronic convertersan overview, IEEE Trans.
Ind. Appl., vol. 50, no. 5, pp. 35693578, 2014. SM04) received the Ph.D. degree
[24] L. Gu, X. Ruan, M. Xu, and K. Yao, Means of eliminating electrolytic in control and engineering from
capacitor in AC/DC power supplies for LED lightings, IEEE Trans. Shanghai Jiao Tong University,
Power Electron., vol. 24, no. 5, pp. 13991408, May 2009.
Shanghai, China, in 2000, and the
[25] H. K. Khalil, Nonlinear Systems. Prentice Hall, 2001.
[26] R. Srinivasan and R. Oruganti, A unity power factor converter using Ph.D. degree in control theory and
half-bridge boost topology, IEEE Trans. Power Electron., vol. 13, no. 3, power engineering (awarded the Best
pp. 487500, 1998. Doctoral Thesis Prize) from Imperial
[27] R. Tymerski, V. Vorperian, F. Lee, and W. Baumann, Nonlinear
modeling of the PWM switch, IEEE Trans. Power Electron., vol. 4,
College London, London, U.K., in
no. 2, pp. 225233, 1989. 2004.
[28] Y.-K. Lo, T.-H. Song, and H.-J. Chiu, Analysis and elimination of He holds the Max McGraw Endowed Chair Professor in
voltage imbalance between the split capacitors in half-bridge boost
rectifiers, IEEE Trans. Ind. Electron., vol. 49, no. 5, pp. 11751177,
Energy and Power Engineering at the Dept. of Electrical
2002. and Computer Engineering, Illinois Institute of Technology,
[29] Y.-K. Lo, C.-T. Ho, and J.-M. Wang, Elimination of the output voltage Chicago, USA, and the Research Professor in Control of
imbalance in a half-bridge boost rectifier, IEEE Trans. Power Electron., Power Systems at the Department of Automatic Control and
vol. 22, no. 4, pp. 13521360, 2007.
[30] T. Hornik and Q.-C. Zhong, H current control strategy for the neutral Systems Engineering, The University of Sheffield, UK. He is a
point of a three-phase inverter, in Proc. of the 50th IEEE Conference Distinguished Lecturer of IEEE Power Electronics Society and
on Decision and Control and European Control Conference, 2011, pp. the UK Representative to the European Control Association.
520525.
[31] Q.-C. Zhong, L. Hobson, and M. Jayne, Generating a neutral point for He serves on the University Technology Partnership (UTP)
3-phase 4-wire DC-AC converters, in Proc. of IEEE Compatibility in Board of Rolls-Royce Plc and served on the Scientific Ad-
Power Electronics (CPE), 2005, pp. 126133. visory Board of FREEDM Systems Center at North Carolina
[32] T. Hornik and Q.-C. Zhong, A current control strategy for voltage-
source inverters in microgrids based on H and repetitive control,
State University. He (co-)authored three research monographs:
IEEE Trans. Power Electron., vol. 26, no. 3, pp. 943952, Mar. 2011. Control of Power Inverters in Renewable Energy and Smart
[33] X. Cao, Q. Zhong, and W. Ming, Ripple eliminator to smooth DC- Grid Integration (Wiley-IEEE Press, 2013), Robust Control
bus voltage and reduce the total capacitance required, IEEE Trans. Ind. of Time-Delay Systems (Springer-Verlag, 2006), Control of
Electron., to be published, early Access.
[34] A. K. Ziarani and A. Konrad, A method of extraction of nonstationary Integral Processes with Dead Time (Springer-Verlag, 2010),
sinusoids, Signal Processing, vol. 84, no. 8, pp. 13231346, Apr. 2004. and a fourth, Completely Autonomous Power Systems (CAPS):
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/TPEL.2015.2414425, IEEE Transactions on Power Electronics
16
0885-8993 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.