Battery Charging PDF
Battery Charging PDF
Battery Charging PDF
This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
Received XX Month, XXXX; revised XX Month, XXXX; accepted XX Month, XXXX; Date of publication XX Month, XXXX; date of
current version XX Month, XXXX.
Digital Object Identifier 10.1109/OJPEL.2022.0817315
ABSTRACT This paper proposes and analyzes a two-stage dc-dc isolated converter for electric vehicle
charging applications, where high efficiency over a wide range of battery voltages is required. The proposed
conversion circuit comprises a first two-output isolation stage with CLLC resonant structure and a second
two-input buck regulator. The transformer of the first stage is designed such that its two output voltages
correspond, ideally, to the minimum and maximum expected voltage to be supplied to the battery. Then,
the second stage combines the voltages provided by the previous isolation stage to regulate the output
voltage of the whole converter. The first stage is always operated at resonance, with the only function
of providing isolation and fixed conversion ratios with minimum losses, whereas the second stage allows
output voltage regulation over a wide range of battery voltages. Overall, it is shown that the solution features
high conversion efficiency over a wide range of output voltages. This paper comprehensively describes
the solution, including modeling, analyses, design considerations for the main circuit components (e.g.,
magnetics, switches), and modulation choices. Experimental results are reported considering a converter
module prototype rated 10 kW, input voltage 800 V, and output range 250 V to 500 V, employing silicon-
carbide and gallium-nitride semiconductors.
INDEX TERMS Battery charger, CLLC, dc-dc converter, dc-transformer (DCX), fast-charging, gallium-
nitride, resonant LLC, post-regulator, silicon-carbide, soft-switching, two-input buck converter.
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME , 1
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
2 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
Sa1 Sb1
B. OPERATING PRINCIPLE
twin-bus buck
DB1 The two-output LLC resonant converter is designed for
ir Lr N1 : N2 iLo Lo
Vg
is1 +
a constant voltage conversion ratio, independent from the
SoH
vi V1 vsw Vo actual load. In such an operating condition the LLC behaves
Sa2 Sb2 vC Co1
SoL
as two-output DCX converter and its voltage gains can be
Cr
DB2 defined as follows:
is2 V1 N2
G1 = = = n1
Co2
V2 Vg N1
(1)
V2 N3
: N3 G2 = = = n2
Vg N1
FIGURE 2. Proposed two-stage topology: two-output dc-transformer &
post-regulator. where N1 , N2 , and N3 are the number of turns of the three
windings of the transformer, as indicated in Fig. 2.
The two-input post-regulator, herein referred to as twin-
bus buck (TBB) converter, is highlighted in Fig. 2 while its
and/or the use of multiple modules connected in parallel main waveforms are displayed in Fig. 3. It is based on a two-
may be required for field use, as typically done in actual input buck topology [61], [62], designed to operate in quasi-
implementations. square wave, that is, with a peak-to-peak inductor current
This paper extends the preliminary results presented in ripple higher than twice the average load current. This allows
[50]. In addition to the extended literature review discussed zero-voltage turn-on of both the switches SoH and SoL . The
in this Sect. I, Sect. II includes details on the operating TBB is responsible of the output voltage regulation of the
principle of the converter, with the main relations governing whole converter. The output voltage Vo is a function of the
the design of the proposed solution now highlighted and dis- TBB input voltages V1 and V2 , with V1 > V2 , and the duty
cussed; Sect. III is dedicated to the design of the DCX stage, cycle d of the upper switch SoH :
and it reports further details on the designed transformer;
Sect. IV presents a comparison with other representative Vo = d V1 + (1 − d) V2 (2)
topologies, which are herein also compared in terms of Therefore, the voltage gain of the converter in Fig. 2 results:
losses, including now a loss breakdown analysis for each
Vo
of the solution; extended experimental results are reported G= = d n1 + (1 − d) n2 (3)
in Sect. V considering a converter module prototype rated Vg
10 kW. Finally, Sect. VI concludes the paper. For fixed input voltages V1 and V2 , the minimum and
maximum output voltages can be defined as:
Vomin = dmin V1 + (1 − dmin ) V2
(4)
Vomax = dmax V1 + (1 − dmax ) V2
II. STRUCTURE AND OPERATING PRINCIPLE
A. TWO-STAGE CONVERTER CONFIGURATION
Several configurations of two-stage dc-dc converters exploit- dTs (1 − d)Ts
ing a voltage post-regulator are described in the literature SoH SoL
[39], [59], [60]. As shown in Fig. 2, the proposed two-stage 0 Ts t
V2 vsw (t)
converter consists of a first isolation stage based on an LLC
V1
resonant converter, and a second post-regulator stage based
on a buck converter. Such a post-regulator is responsible of 0 Ts t
the output voltage regulation and it is supplied by means of ILomax
a high-efficiency two-output DCX converter, with secondary iLo (t)
Io
voltages V1 and V2 . From Fig. 2, it is clear that the voltage
stress of the post-regulator, namely, V1 − V2 , is lower that 0
ILomin
Ts t
the output voltage Vo , which consequently allows switching
devices with smaller on-resistance as well as lower switching
losses. iSoH (t)
Io
It is worth remarking that the topology where the two
DCX-LLC outputs are connected in series can also be 0 Ts t
iSoL (t)
considered. This variant will be considered in future inves-
tigations. Preliminary studies, shown as potential advantage
in the transformer design but as disadvantage the additional FIGURE 3. Main waveforms of TBB stage shown in Fig. 2. In order: gate
losses due to the output current passing trough both the two driver signals including dead times, switching node voltage of TBB, Lo
diodes-rectifier bridges. inductor current and SoH , SoL switch currents.
VOLUME , 3
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
Cl. Ref. Topology ac-FE vIN vOU T Prated ηmax S/D/T/I∗ Peculiarities
0 [10] HB-CLLC no 500 V 200 - 420 V 1 kW 96.5% 4/0/1/2 Bidirectional operation, limited controllability un-
(0.6 kW, der wide vOU T .
300 V-out)
0 [17] LLC no 380 - 420 V 400 V 6.6 kW 98% (3 kW, 4/4/1/0 High η, wide variation of fsw , narrow vIN , fixed
390 V-in) vOU T , η drops far from resonance.
1 [26] FB/HB- no 400 V 100 - 400 V 2 kW 96.36% 6/4/1/0 Wide controllable vOU T , high η over a wide load
LLC + (0.5 kW, range.
VD 200 V-out)
3 [51] switched no 200 - 400 V 1200 V 4 kW 97.71% 22/0/0/6 Low volume, no galvanic isolation.
tank (4 kW,
350 V-in)
0 [52] phase- no 700 - 800 V 350 - 700 V 20 kW 98.9% 4/4/1/1 High η, hard switching operated.
shift FB (13 kW,
700 V-in,
686 V-out)
0 [53] LLC yes 7 kV 400 V 350 kW 98.6% 6/0/1/1 10 kV SiC devices, 4.16 kV ac grid input, limited
(350 kW, controllability under the wide vOU T .
400 V-out)
3 [38] CLLC + no 750 V 314 - 450 V 18 kW 98.8% 20/0/2/3 High η and power density.
buck (18 kW,
375 V-out)
1 [20] interleaved no 390 V 230 - 440 V 1.3 kW 97.31% 6/6/2/0 Limited vOU T range.
LLC (1.3 kW,
440 V-out)
4 [42] ISOP no 1 - 2 kV 700 V 12 kW 93.7% 5/5/1/1 Modularity, high pre-regulator losses.
boost + (12 kW,
LLC 1.2 kV-in)
1 [24] LLC no 160 - 320 V 400 V 1 kW 95.2% 4/4/2/1 Wide vIN , simple control, fixed vOU T .
(1 kW,
160 V-in)
0 [54] CLLC no 400 V 250 - 450 V 1 kW 97.9% 4/0/1/2 Wide vOU T , low η far from resonance.
(1 kW,
325 V-out)
1 [23] interleaved no 390 V 10 - 420 V 1 kW 98.1% 4/4/2/0 Very wide vOU T .
LLC (0.82 kW,
420 V-out)
1 [27] LLC + no 390 V 250 - 420 V 1.3 kW 93.94% 5/6/1/1 Wide vOU T .
VQ (0.95 kW,
420 V-out)
0 [55] LLC yes 380 - 660 V 200 - 500 V 6.6 kW 98% 8/0/1/1 Bidirectional, low voltage-gain of the dc/dc stage,
(6.6 kW, 220 V ac input.
440 V-out)
2 [30] LLC with no 390 V 126 - 420 V 1 kW 97.18% 8/6/1/1 Smooth transitions without transients.
adj. transf. (0.7 kW,
420 V-out)
0 [4] CLLLC yes 650 - 900 V 214 - 413 V 11 kW 98.75% 8/0/2/2 Bidir., 380 V input PFC for voltage regulation.
(11 kW,
792 V-in,
330 V-out)
1 [56] active no 10 kV 700 V 30 kW 99.1% 10/0/1/1 High η and power density, fixed conversion ratio,
NPC DAB (10 kW, careful layout and integrated modules required.
700 V-out)
4 [44] 3-phase no 850 V 200 - 800 V 12.5 kW 97.72% 20/0/3/4 Ultra wide vOU T for 400 and 800 Vbattery sys-
CLLC + (12.5 kW, tems, low η over the wide range, high component
buck 800 V-out) count.
0 [3] interleaved yes 900 V 200 - 650 V 22 kW 99.51% 6/0/0/6 Very high η, no galvanic isolation, 480 V ac input.
buck (22 kW,
200 V-out)
3 [12] 3-port no 388 - 412 V 250 - 450 V 2.3 kW 98% 8/0/1/1 PPP, with low-voltage devices, and limited overall
CLLC + (2.3 kW, conduction losses, high component count and com-
PPP 450 V-out) plexity.
3 [40] 2-level no 1500 V 630 - 900 V 50 kW NA 14/8/2/4 Modularity, utilization of low voltage semiconduc-
LLC + tors.
PPP
0 [57] buck- yes 150 V 48 - 450 V 1.5 kW 95.6% 6/0/0/2 Bidir., non-isolated, limited η, 85-265 V ac input.
boost (1.5 kW,
110 V-in ac,
250 V-out)
0 [58] phase- yes 400 V 330 V 3.3 kW 97.2% 4/4/1/1 On-board charger, 230 V ac input.
shift FB (3.3 kW,
330 V-out)
0 [29] resonant no 800 V 150 - 500 V, 6.6 kW 98.2% 8/0/1/2 Wide vOU T , reconfigurable rectifier, phase shift
LCL-T 500 - 950 V (6.6 kW, modulation, resonant network for η optimization
580 V-out) during CC charging phase.
4 [21] Buck- no 800 V 250 - 500 V 5 kW 98% (3 kW, 4/4/1/1 Very wide voltage gain (potential), high losses at
Boost + 400 V-out) light-load.
LLC
4 herein CLLC + no 800 V 250 - 500 V 10 kW 98.63% 8/8/1/2 Wide vOU T , simple control, high component
twin-bus (7 kW, count.
buck 500 V-out)
∗
S/D/T/I: number of active switches, diodes, transformers, and inductors, respectively.
4 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
with dmin and dmax the minimum and maximum duty and zero-voltage switching (ZVS) and zero-current switching
cycles of SoH , corresponding to Vomin and Vomax in Table 2, (ZCS) conditions always satisfied [37]. Notably, the DCX
respectively. Their value must guarantee the zero-voltage operation of the LLC does not require an external resonant
switching operation of SoH and SoL at the respective output inductor, because the conversion gain is fixed. An equivalent
voltage levels. Thus the needed input voltages V1 and V2 solution based on a resonant FB-LLC designed to operate
provided by the DCX stage, can be calculated from (4) as: over the same wide range of output voltages is expected
Vomax (1 − dmin ) − Vomin (1 − dmax ) to show higher losses than the LLC in permanent DCX
V1 = conditions, as demonstrated in Sect. IV-B.
dmax − dmin
min max
(5) From (1) and (5) and considering dmin = (1 − dmax ) =
V d − Vomax dmin
V2 = o 5%, the transformer turns ratio can be calculated as n1 =
dmax − dmin N2 /N1 = 0.642 and n2 = N3 /N1 = 0.295 to make the
and the voltage gains (1) of the DCX-LLC can be derived. LLC converter operate at the resonant frequency fs at input
By using (5), the maximum voltage stress of the switches voltage Vg = 800 V and output bus voltages V1 and V2 as
can be computed as: in (5).
Vomax − Vomin
V1 − V2 = (6)
dmax − dmin A. TRANSFORMER DESIGN
which is always lower than the voltage stress of the switches In the design of the main magnetic element, both winding
of a full-power converter that requires a supply voltage and core losses must be considered. The transformer design
higher than the maximum output voltage. In order to min- procedure adopted herein is based on [21], [47], [66].
imize such voltage stress, and the related switching loss, Once the magnetic core is selected, with given magnetic
the duty-cycle excursion dmax − dmin should be maximized; volume Vc , window winding area Wa , core cross-sectional
then, for example, by imposing dmin = (1 − dmax ) = 5%. area Ac , Steinmetz parameters Kc , α and β , and maximum
Consequently, the converter in Fig. 2 with voltage ratings window filling factor ku (typ., assume ku ≤ 40%), it is
Vomax = 2 Vomin = 500 V presents a voltage stress on the possible to calculate the winding losses as:
switching devices of V1 − V2 = 278 V, allowing the use of
devices of low rated-voltages, which typically implies lower P cond = RF (fs )ρw Vw ku J02 (8)
losses [37], [38]. where ρw is the copper resistivity, Vw is the total windings
Once the duty-cycle range of the TBB stage is defined, volume, RF (fs ) = Rac /Rdc is the resistivity factor for the
zero-voltage switching (ZVS) can be achieved with a proper selected litz wire at fundamental frequency fs [66] and J0
selection of the output inductor value and the switching is the current density. The last parameter is calculated as:
frequency, for the whole output voltage range. As shown in P
VA
Fig. 3, the TBB is operated in continuous conduction mode J0 = (9)
Kv fs kf Bmax ku Ap
(CCM) and the inductor current at switching instants can be P
computed as: where V A is the power rating of the transformer, Kv is
V1 − V2 the waveform factor, Bmax is the peak flux density, kf is
ILomax = Io + d (1 − d) core stacking factor, and Ap = Ac Wa is the area product of
2fs Lo
(7) the core.
V1 − V2
ILomin = Io − d (1 − d) The core losses can be estimated using the Steinmetz
2fs Lo
equation:
In order to achieve zero-voltage switching, these current
P core = Vc Kc fsα Bmax
β
(10)
values should satisfy the minimum switched current con-
ditions for ZVS (refer, e.g., to [13], [16], [21], [63]–[65]). where Kc , α and β are the Steinmetz parameters for the
The inductor current value and the switching frequency of considered material, while Vc is the core volume. The total
the TBB stage are key parameters for the converter design transformer dissipated power is then computed as the sum
and operation over the whole range of output voltages and of (8) and (10) and it must be lower than the thermal dissi-
powers. pation capability of the component at the desired operating
temperature, which can be estimated during the design phase.
III. DESIGN OF LLC STAGE OPERATED AS DCX Fig. 4 reports the results of the calculated transformer losses,
The converter structure is shown in Fig. 2. When the LLC showing a total loss of 24 W at nominal conditions, namely,
resonant tank is operated at the resonance frequency, the V1 = 514 V and V2 = 236 V, and Po = 10 kW. According
voltage conversion ratio becomes ideally independent from with Fig. 4, the selected design point is more conservative
the actual load. In other words, the LLC converter maintains in terms of core losses with respect to the optimal point,
a constant voltage conversion ratio and adjusts its current this is due to a trade-off between the desired magnetizing
automatically, according to the load conditions, behaving inductance and the conductor sections.
as a DCX. In this operating condition, the LLC shows its Fig. 5 depicts the winding layout of the designed trans-
maximum efficiency, with a minimum flow of reactive power former of Fig. 6a. The designed transformer presents turns
VOLUME , 5
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
L r1 1 : n 1 L r2
(P cond +P core )min
Design core i s1
cond +P Lr V1
1−d :
P
:d
L r3
P cond i s2 iL o
P core
+
V2
Vo
: n2
FIGURE 4. P -B plot for transformer design at Vo = 400 V and FIGURE 7. Equivalent circuit model for the estimation of resonant
Po = 10 kW. inductance at the primary side of the transformer, namely Lr in Fig. 2.
...
... current to allow ZVS for all the switches of the main
...
... N1 = 24 converter. A classical design for a DCX-LLC with voltage
b = 14mm
... N2 = 15
ratings of Table 2 requires a magnetizing inductance of about
...
N3 = 7
200 µH (see, for example, [16], [21], [37]). The designed
... litz 420x50µm
... coil former transformer in Fig. 6a achieves the design target, with a
... kapton tape
magnetizing inductance of about 215 µH.
... The capacitive part of the resonant tank can be selected
core PQ65/60 on the basis of the desired resonant frequency (i.e., converter
switching frequency at DCX-LLC operation). The winding
w = 36.4mm
arrangement of the designed transformer in Fig. 4 is shown
FIGURE 5. Winding arrangement at the design point in Fig. 4. in Fig. 5. The interleaving of the primary and secondary-side
windings is an effective solution to limit the leakage induc-
ratio n1 = 0.625 and n2 = 0.292, current density J0 = tance and winding losses [67]. The experimental prototype in
5 A/mm2 , number of turns per winding N1 = 24, N2 = 15, Fig. 6a, which results from the design in Fig. 4 and winding
N3 = 7. arrangement in Fig. 5, presents values of leakage inductances
Lr1 = 795 nH, Lr2 = 445 nH, and Lr3 = 271 nH for the
B. RESONANT TANK DESIGN input, high-voltage, and low-voltage windings, respectively.
For what concerns the design of the resonant Lr Cr tank, The secondary windings leakage inductances Lr2 and Lr3
the transformer leakage inductance can be exploited for affect the overall resonance frequency proportionally to the
the implementation of the inductive part. Given the DCX normalized conduction interval of the respective diode bridge
operation mode of the LLC, low values of Lm can be used, rectifier. In fact, these inductances come into play only when
which is beneficial in terms of transformer design, losses, the corresponding rectifying diodes are conducting, and these
and resonant capacitor voltage stress. With the aimed DCX intervals are related to the duty-cycle of the TBB stage,
operation, the value of the magnetizing inductance Lm is as well as to the load current. The TBB stage imposes a
typically chosen to ensure a sufficiently high magnetizing strict relationship between the average charge transferred
through each output ports of DCX-LLC stage with respect
to the output voltage Vo and current Io . Then, the stage
can be modelled as shown in Fig. 7. The series-equivalent
inductance Lr of the resonant tank referred to the primary
side of Fig. 7 can be calculated as:
d2 Lr2 + (1 − d)2 Lr3
Lr = Lr1 + (11)
[n1 d + n2 (1 − d)]2
which is a function of the converter operating point, ac-
cording to (2). The validity of (11) is shown in Sect. IV
referring to a specific operating point. In order to remove
the dependence of the resonance frequency from the load,
(a) (b)
two additional resonant capacitors are connected in series
with the two output ports of the transformer, as shown
FIGURE 6. (a) Transformer prototype, and (b) thermography at the design
point in Fig. 4, namely, Vo = 400 V, Po = 10 kW, natural convection in Fig. 8. At resonance, the capacitive part of each of the
conditions. series-resonant impedances Lri Cri cancels out with the
6 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
: N3 C r3
Switching freq. of CLLC fs 200 kHz
Switching freq. of TBB fs o 50 - 400 kHz
FIGURE 8. Overall circuit schematic of the solution described herein,
composed of a DCX-CLLC stage plus two interleaved twin-bus buck Turns ratio N2 /N1 n1 0.625 -
stages.
Turns ratio N3 /N1 n2 0.292 -
Intermediate bus V1 V1 500 V
corresponding inductive part. Cr1 = 796 nF, Cr2 = 1.42 µF Intermediate bus V2 V2 234 V
and Cr3 = 2.34 µF are then calculated as proper values for Magnetizing inductance Lm 215 µH
the resonant capacitances in order to achieve a continuous Lr1 795 nH
resonant current operation, where the resonant frequency of Leakage inductances Lr2 445 nH
the CLLC stage becomes independent from the duty-cycle Lr3 271 nH
and the output current of the TBB stage, as otherwise shown TBB inductor Lo 30 µH
in (11). The proposed post-regulated converter is then shown Transformer Core: PQ65/60, Material: N87
in Fig. 8. Inductor Core: PQ40/40, Material: N97
Cr1 796 nF
IV. SIMULATION RESULTS
Resonant capacitances Cr2 1.42 µF
A. DCX RESONANT CURRENT OPERATION
Cr3 2.34 µF
A converter topology with parameters reported in Table 2
is considered for validation. Based on the considerations Sa1 , Sa2 , Sb1 , Sb2 G3R30MT12K, 1.2 kV SiC MOSFETs
reported in Sect. III, herein are reported the simulation results SoH , SoL LMG3422R030, 600 V GaN FET
focused on demonstrating the continuous resonant current Output Rectifier DB1 UJ3D06560KSD, 650 V SiC diodes
operation of the DCX stage. First, the operation of the Output Rectifier DB2 STTH100W04CW, 400 V Si diodes
converter in Fig. 2 is considered with a single resonant
capacitor Cr and, then, the operation of the proposed
converter in Fig. 8 is considered. Converters in Fig. 2 and presented in [21] (reported in Fig. 10 for reference), and the
Fig. 8 are simulated and the resonant currents are shown solution proposed herein. As discussed in Sect. I, these last
in Fig. 9. Let us consider different operating points at the two multi-stage topologies present higher efficiency than the
maximum output current Io = 25 A and minimum, nominal, classical frequency-modulated LLC converter. The consid-
and maximum output voltage Vo (i.e., 250, 400, 500 V). ered topologies are rated 10 kW and have been carefully
Figs. 9(a)-(c) shows the resonant currents ir , is1 , and is2 , designed in order to optimize their performances. The FB-
and the magnetizing current im of the circuit in Fig. 2 LLC is designed to allow the required output voltage range,
with resonant capacitance Cr = 174 nF. Such a value is while the FB-LLC of the proposed two-stage solution is
designed to have the desired resonance frequency fs , with designed for DCX operation. In principle, the FB-LLC for
Lr = 3.64 µH given by (11) at Vo = 250 V. Indeed, the DCX operation allows lower loss because it does not require
current is2 is resonant only in such an operating point. While, an inductor for accommodating a wide voltage gain range.
Figs. 9(d)-(f) shows the resonant currents considering the Fig. 11 reports and compares the efficiency curves of the
circuit in Fig. 8 with resonant capacitances Cri in Table 2. three topologies for Vo = 250 V, 400 V, and 500 V. The
Simulation results show that the resonance conditions are results are obtained by PLECS models tuned for accurately
satisfied, for the whole wide output voltage range, only in accounting for switching, magnetic, and conduction losses
this later case. Furthermore, conduction losses are minimized [21]. Such models were validated experimentally by means
only if the resonance conditions are satisfied. Based on the of thermal measurements in [21]. The efficiency perfor-
obtained results, the CLLC solution is considered for the mances of the topology in [21] is limited due to the high con-
investigations in the following. duction and switching losses in the pre-regulation stage. The
frequency-modulated FB-LLC converter has low-efficiency
B. PERFORMANCE COMPARISON performances in the minimum output voltage range due to
This section compares three different topologies, namely, the the limited voltage gain and the higher switching frequencies.
classical full-bridge LLC (FB-LLC), the buck-boost LLC Furthermore, the low Q-factor in the light-load region and
VOLUME , 7
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
current (A)
4 10
2
0 0
-2
-4 -10
-6
-8 -20
-10
-12 -30
is1 and is2 is1 and is2
40 40
30
i s1 30
i s1
i s2 i s2
20 20
current (A)
current (A)
10 10
0 0
-10 -10
-20 -20
-30 -30
-40 -40
11 12 13 14 15 16 17 18 19 20 21 22 2 3 time (µs) 11 12 13 14 15 16 17 18 19 20 21 22 2 3 time (µs)
(a) (b)
current (A)
10 4
2
0 0
-2
-10 -4
-6
-20 -8
-10
-30 -12
-14
is1 and is2 is1 and is2
60 40
i s1 30
i s1
40 i s2 i s2
20
current (A)
current (A)
20 10
0 0
-20 -10
-20
-40 -30
-60 -40
11 12 13 14 15 16 17 18 19 20 21 22 2 3 time (µs) 11 12 13 14 15 16 17 18 19 20 21 22 2 3 time (µs)
(c) (d)
current (A)
10
5
0 0
-5
-10 -10
-15 -20
-20
-25 -30
is1 and is2 is1 and is2
30 40
i s1 30
i s1
20 i s2 i s2
20
current (A)
current (A)
10 10
0 0
-10 -10
-20
-20 -30
-30 -40
11 12 13 14 15 16 17 18 19 20 21 22 2 3 time (µs) 11 12 13 14 15 16 17 18 19 20 21 22 2 3 time (µs)
(e) (f)
FIGURE 9. Simulation results for LLC with different resonant tank designs, (a)-(c) refer to Fig. 2 and (d)-(f) refer to Fig. 8. (a),(d) Vo = 250 V; (b),(e)
Vo = 400 V; (c),(f) Vo = 500 V. Io = 25 A. Converter parameters are reported in Table 2.
8 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
n:1
for evaluation, providing negligible benefits.
Cr
Fig. 12 shows the loss breakdown of the considered
(a)
topologies for the efficiency comparison in Fig. 11, using the
same methodology. In general, the major loss contribution
SaH SbH
vb Cb comes from the rectification stages; if active rectification
Lb ir Lr
Vg
ib is irct Io is implemented, the performances of all the considered
va vi Lm Co Vo
topologies will improve consequently, at the cost of higher
SaL SbL vC i
m
circuit complexity. Substantial magnetic losses are present in
Cr n:1 the frequency-modulated FB-LLC converter that affect the
(b) efficiency performances.
In summary, the proposed solution in Fig. 8 offers valuable
FIGURE 10. Additional topologies considered for efficiency and loss
comparisions in Sect. IV. (a) FB-LLC, (b) solution in [21]. efficiency improvements, at the cost of a small component in-
crement. The number of semiconductors is doubled and addi-
tional bus capacitances are needed for the proposed structure
99,0 in Fig. 8, as compared to the classical FB-LLC. Nevertheless,
98,5 excellent efficiency performances for the whole wide range
efficiency (%)
98,0
97,5 Fig. 13 shows the experimental validation of the consider-
97,0 ations discussed in Sect. IV. In particular, Figs. 13(a), (c), (e)
96,5 show the measured resonant currents at the same operat-
96,0 ing points of the simulated waveforms in Figs. 9(d), (e), (f),
95,5 respectively. It is possible to appreciate that the current
95,0 waveforms are very close to the continuous resonant cur-
0 2 4 6 8 10
FB-LLC [21] herein power (kW) rent operation of the DCX-CLLC. The measured waveform
amplitudes correspond to the expected values. The switch-
(b)
ing frequency is set to fs = 200 kHz and dead-time to
98,5 td = 260 ns. If needed, additional refinements to match the
98,0 true resonance frequency may be performed by adjusting
efficiency (%)
VOLUME , 9
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
FB-LLC 98.5% 4 (4x 1.2 kV SiC MOSFET) / Pros: Low component count; Cons: frequency modula-
(8 kW, 500 V-out)
(Fig. 10a) 4 (4x 650 V SiC Diode) / tion, limited controllability over the wide range, high
1 (150 µH, PQ65/60-N87) / magnetic losses.
2 (2x 16 µH, PQ40/40-N97).
BB-LLC 98.4% 4 (4x 1.2 kV SiC MOSFET) / Pros: very wide voltage gain (potential); Cons: high
(5 kW, 400 V-out)
(Fig. 10b) 4 (4x 650 V SiC Diode) / circulating reactive currents in light-load, higher losses.
1 (150 µH, PQ65/60-N87) /
2 (2x 8 µH, PQ40/40-N97).
CLLC+TBB 98.72% 8 (4x 1.2 kV SiC MOSFET + 4x 600 V GaN FET) / Pros: high efficiency in a wide range, simple control;
(8 kW, 500 V-out)
(Fig. 8) 8 (4x 650 V SiC Diode + 4x 400 V Si Diode) / Cons: high component count.
1 (200 µH, PQ65/60-N87) /
2 (2x 30 µH, PQ40/40-N97).
current mode charging when the battery is discharged (i.e., at maximum output voltage conditions. Such values are very
low battery voltages). In this condition, the charging current close to the estimations performed by the tuned simulation
is maximum and equals to the output current Iomax at the shown in Fig. 11 and previously presented in [50].
nominal output power (see, e.g., [70]), namely, 25 A at
10 kW in the considered case.
VI. CONCLUSIONS
Figs. 13(c), (d) show the converter waveforms at nominal
A two-stage converter, not previously documented for EV
output voltage Vo = 400 V and output current Io = 25 A.
battery charging applications, has been proposed, designed,
The duty cycle of TBB is set to 65% and the switching
and demonstrated in this paper. The conversion structure is
frequency of the TBB is fso = 73 kHz in order to achieve
composed of a DCX-CLLC and a post-regulator and features
ZVS. The conversion efficiency in such a point is about
high efficiency, a wide range of output voltage, and a simple
98.4%.
principle of operation. The DCX-CLLC converter always
Figs. 13(e), (f) show the converter waveforms at maximum
operates at its optimal operating point and the additional
output voltage Vo = 500 V and output current Io = 25 A.
post-regulator based on a two-input buck converter is used
The duty cycle of the TBB is set to 95%, the switching
to regulate the output voltage. In such a post-regulator, the
frequency to the lower limit of fso = 50 kHz. The conversion
stress of the switches is a fraction of the rated voltages.
efficiency in such a point is about 98.5%. For Vo = 500 V,
Hence, the efficiency of the proposed configuration, com-
ZVS conditions are not satisfied for output currents higher
pared with standard dc/dc converters processing full power,
than about 10 A. Remarkably, the loss of ZVS in heavy load
can be improved. The considered topology is presented, sim-
conditions and extreme duty-cycle is the direct consequence
ulation results of a resonant two-output CLLC are reported
of the selected inductance Lo . Indeed, the selected value
and an efficiency performance comparison is included. The
allows to achieve ZVS in light-load conditions and with a
reported analysis and the experimental characterizations and
switching frequency of the TBB limited to fso = 400 kHz.
tests were performed on a 10 kW prototype module based
Some ringing at the commutations of the input full-
on silicon-carbide (SiC) devices and gallium-nitride (GaN)
bridge current is visible in Figs. 13(a), (c), (e). The ringing
devices. Conversion performances covering the whole power
appears during the dead-times and is generated by resonances
and voltage ranges have been reported experimentally, show-
between the transformer leakage inductances and the devices
ing high efficiency over a wide range of operating conditions,
output capacitances. These resonances may bring partial
recording a peak efficiency of 98.63% at 500 V output
ZVS and ZCS conditions and eventually cause increased
voltage and 7 kW transferred power. In final applications,
switching losses. This aspect is investigated in [16], which
series or parallel connections of multiple modules can be
also proposes a method to reduce the related switching loss
considered for scaling the voltage or current ratings of the
based on switching frequency and dead-time perturbations.
final implementation, thanks to the isolated output. Future
Finally, Fig. 16 shows the converter efficiency measured
studies may include on-line controllers for optimal converter
at the minimum, nominal, and maximum output voltage.
modulation and procedures for the optimal design of the
Efficiency measurements were performed by means of a
components of the converter, like the output TBB inductors.
Keysight PA2203A power analyzer. The measured peak
efficiency at minimum output voltage is 97.8%, while at
nominal output voltage is 98.51%, which is very close to REFERENCES
the absolute maximum efficiency of 98.63% measured at [1] Global EV Outlook 2022, International Energy Agency, 2022. [On-
line]. Available: https://www.iea.org/reports/global-ev-outlook-2022
10 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
VOLUME , 11
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
vi (200 V/div)
[1 µs/div] [5 µs/div]
(a) (b)
vi (200 V/div)
fso = 73 kHz
is2 (4 A/div)
ir (5 A/div)
vsw (100 V/div) iLo (4 A/div)
is1 (10 A/div)
[1 µs/div] [4 µs/div]
(c) (d)
vi (200 V/div)
is1 (10 A/div)
ir (5 A/div)
is2 (5 A/div)
[1 µs/div] [5 µs/div]
(e) (f)
FIGURE 13. Experimental results of the proposed converter in Fig. 8 at Io = 25 A. (a),(b) Vo = 250 V; (c),(d) Vo = 400 V; (e),(f) Vo = 500 V.
(a), (c), (e) are the experimental validations of the simulations in Figs. 9(d), (e), (f), respectively.
12 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
VOLUME , 13
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
[40] A. Stanojevic, Y. E. Bouvier, and P. J. Grbovic, “Comparison of 2- medium-voltage extreme fast charger,” in 2022 IEEE Energy Conver-
stage isolated converters for fast EV charger, using partial power,” sion Congress and Exposition (ECCE). IEEE, oct 2022.
in IECON 2022 – 48th Annual Conference of the IEEE Industrial [57] H. Karneddi and D. Ronanki, “Universal bridgeless nonisolated battery
Electronics Society. IEEE, oct 2022. charger with wide-output voltage range,” IEEE Transactions on Power
[41] Z. Wu, Z. Wang, T. Liu, W. Xu, C. Chen, and Y. Kang, “High Electronics, vol. 38, no. 3, pp. 2816–2820, mar 2023.
efficiency and high power density partial power regulation topology [58] C. Hu, R. Wang, Y. Shi, X. Jia, and D. Xu, “A fixed frequency zero-
with wide input range,” IEEE Transactions on Power Electronics, voltage-switching on-board EV charger,” IEEE Open Journal of Power
vol. 38, no. 2, pp. 2074–2091, feb 2023. Electronics, vol. 3, pp. 75–83, 2022.
[42] F. Liu, G. Zhou, X. Ruan, S. Ji, Q. Zhao, and X. Zhang, “An input- [59] E. Abramov, Y. Schultz, M. Evzelman, and M. M. Peretz, “Analysis
series-output-parallel converter system exhibiting natural input-voltage and design of post-regulation stages for resonant capacitively-coupled
sharing and output-current sharing,” IEEE Transactions on Industrial wireless power systems,” in 2022 IEEE Applied Power Electronics
Electronics, vol. 68, no. 2, pp. 1166–1177, feb 2021. Conference and Exposition (APEC). IEEE, mar 2022.
[43] L. Zhang, X. Wu, and H. Chen, “1mhz LLC resonant DC-DC [60] I. Lopusina and P. Grbovic, “Comparative analysis of input-series-
converter with PWM output regulation capability,” in 2016 IEEE output-series partial power rated DC to DC converters,” in 2021 21st
8th International Power Electronics and Motion Control Conference International Symposium on Power Electronics (Ee). IEEE, oct 2021.
(IPEMC-ECCE Asia). IEEE, may 2016. [61] J. Sebastian, P. Villegas, F. Nuno, and M. Hernando, “High-efficiency
[44] F. Jin, A. Nabih, C. Chen, X. Chen, Q. Li, and F. C. Lee, “A and wide-bandwidth performance obtainable from a two-input buck
high efficiency high density DC/DC converter for battery charger converter,” IEEE Transactions on Power Electronics, vol. 13, no. 4,
applications,” in 2021 IEEE Applied Power Electronics Conference pp. 706–717, jul 1998.
and Exposition (APEC). IEEE, jun 2021. [62] A. Urtasun and D. D.-C. Lu, “Control of a single-switch two-input
[45] Y. Jeong, J.-K. Kim, J.-B. Lee, and G.-W. Moon, “An asymmetric buck converter for MPPT of two PV strings,” IEEE Transactions on
half-bridge resonant converter having a reduced conduction loss for Industrial Electronics, vol. 62, no. 11, pp. 7051–7060, nov 2015.
DC/DC power applications with a wide range of low input voltage,” [63] M. Kasper, R. Burkat, F. Deboy, and J. Kolar, “ZVS of power
IEEE Transactions on Power Electronics, vol. 32, no. 10, pp. 7795– MOSFETs revisited,” IEEE Transactions on Power Electronics, pp.
7804, oct 2017. 1–1, 2016.
[46] L. Lin, J. Xu, Y. Chen, X. Wang, and J. Cao, “Asymmetrical hybrid- [64] D. Costinett, R. Zane, and D. Maksimovic, “Circuit-oriented modeling
controlled half-bridge LCC resonant converter with low conduction of nonlinear device capacitances in switched mode power converters,”
loss and wide ZVS operation range,” Electronics Letters, vol. 53, in 2012 IEEE 13th Workshop on Control and Modeling for Power
no. 21, pp. 1422–1424, oct 2017. Electronics (COMPEL). IEEE, jun 2012.
[65] D. Costinett, D. Maksimovic, and R. Zane, “Circuit-oriented treatment
[47] N. Zanatta, T. Caldognetto, G. Spiazzi, and P. Mattavelli, “A two-
of nonlinear capacitances in switched-mode power supplies,” IEEE
stage isolated resonant DC-DC converter for wide voltage range
Transactions on Power Electronics, vol. 30, no. 2, pp. 985–995, feb
operation,” in 2021 IEEE International Conference on Environment
2015.
and Electrical Engineering and 2021 IEEE Industrial and Commercial
[66] W. Wölfle and W. Hurley, Transformers and Inductors for Power
Power Systems Europe (EEEIC / I&CPS Europe). IEEE, sep 2021.
Electronics: Theory, Design and Applications. Wiley, Apr. 2013.
[48] Q. Liu, Q. Qian, B. Ren, S. Xu, W. Sun, and L. Yang, “A two-
[67] P. L. Dowell, “Effects of eddy currents in transformer windings,” IEE
stage buck–boost integrated LLC converter with extended ZVS range
Proceedings, vol. 113, no. 8, pp. 1–16, 1966.
and reduced conduction loss for high-frequency and high-efficiency
[68] V. Sankaranarayanan, Y. Gao, R. W. Erickson, and D. Maksimovic,
applications,” IEEE Journal of Emerging and Selected Topics in Power
“Online efficiency optimization of a closed-loop controlled SiC-based
Electronics, vol. 9, no. 1, pp. 727–743, feb 2021.
bidirectional boost converter,” IEEE Transactions on Power Electron-
[49] X. Sun, J. Qiu, X. Li, B. Wang, L. Wang, and X. Li, “An improved ics, vol. 37, no. 4, pp. 4008–4021, apr 2022.
wide input voltage buck-boost + LLC cascaded converter,” in 2015 [69] J. Min and M. Ordonez, “Unified bidirectional resonant frequency
IEEE Energy Conversion Congress and Exposition (ECCE). IEEE, tracking for CLLC converters,” IEEE Transactions on Power Elec-
sep 2015. tronics, vol. 37, no. 5, pp. 5637–5649, may 2022.
[50] N. Zanatta, T. Caldognetto, D. Biadene, G. Spiazzi, and P. Mattavelli, [70] L. Wang, Z. Qin, T. Slangen, P. Bauer, and T. van Wijk, “Grid Impact
“Analysis and design of a partial-power post-regulator based dc/dc of Electric Vehicle Fast Charging Stations: Trends, Standards, Issues
converter for automotive applications,” 2022 IEEE 13th International and Mitigation Measures - An Overview,” IEEE Open Journal of
Symposium on Power Electronics for Distributed Generation Systems Power Electronics, vol. 2, pp. 56–74, 2021.
(PEDG), 2022.
[51] Z. Ni, Y. Li, J. Johnson, M. Wei, C. Liu, X. Lyu, and D. Cao, “A
GaN switched tank converter with partial power voltage regulation for
electric vehicle applications,” in 2019 IEEE Applied Power Electronics NICOLA ZANATTA (Student Member, IEEE) re-
Conference and Exposition (APEC). IEEE, mar 2019. ceived the M.S. degree in electronic engineering
[52] C. O. C. K. G. K. R. Ramachandran, M. Nymand, “High efficiency from the University of Padova, Padua, Italy, in
20kw sic based isolated dc-dc converter for batterycharger applica- 2020. He is currently working toward the Ph.D.
tions,” in 2018 20th European Conference on Power Electronics and degree in mechatronics and product innovation en-
Applications(EPE’18 ECCE Europe). Piscataway, New Jersey: IEEE, gineering with the Department of Management and
2018, pp. 1–9. Engineering, University of Padova, Vicenza, Italy.
[53] G. V. Bharath, S. K. Voruganti, V. T. Nguyen, V. U. Pawaskar, and His main research interests include the analysis,
G. Gohil, “Performance evaluation of 10kv SiC-based extreme fast design, and optimization of dc–dc converters for
charger for electric vehicles with direct MV AC grid interconnection,” ultra-fast charging applications.
in 2020 IEEE Applied Power Electronics Conference and Exposition
(APEC). IEEE, mar 2020.
[54] J. Min and M. Ordonez, “Bidirectional resonant CLLC charger for
wide battery voltage range: Asymmetric parameters methodology,”
IEEE Transactions on Power Electronics, vol. 36, no. 6, pp. 6662–
6673, jun 2021.
[55] H. Li, S. Wang, Z. Zhang, J. Zhang, W. Zhu, X. Ren, and C. Hu,
“A bidirectional synchronous/asynchronous rectifier control for wide
battery voltage range in SiC bidirectional LLC chargers,” IEEE Trans-
actions on Power Electronics, vol. 37, no. 5, pp. 6090–6101, may
2022.
[56] H. Feng, J. Won, X. Liang, S. Srdic, and S. Lukic, “Design and
implementation of DC-transformer using 10 kV SiC MOSFET for
14 VOLUME ,
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in IEEE Open Journal of Power Electronics. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/OJPEL.2023.3271227
VOLUME , 15
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/