Differential Power Processing Converter With Active Clamp Structure and Integrated Planar Transformer for Power Generation Optimization of Multiple Photovoltaic Submodules

In recent years, in order to improve the fuel efficiency of electric vehicles, a lot of research has been conducted to adopt photovoltaic (PV) systems to electric vehicles. In addition, various Differential Power Processing (DPP) converters that can mitigate power imbalance among submodules in PV systems are being studied and developed. However, most conventional DPP converters suffer from large volume, low-price competitiveness, and low efficiency. Therefore, in this paper proposes a novel DPP converter. The proposed DPP converter can significantly reduce the volume of converter by integrating multiple transformers into single transformer. Moreover, because of the active clamp structure, it is possible to eliminate secondary side circuits. Furthermore, more than 95% efficiency under 10-60% load and maximum 96.654% efficiency about 20% load were achieved. Thus, the proposed DPP converter can achieve high power density, high price competitiveness, as well as high efficiency. To support the validity of the proposed converter, a prototype of the proposed DPP converter was built, and experiments were proceeded with 8V input and 60W/30V output specifications.


I. INTRODUCTION
Photovoltaic (PV) power generation systems are being applied to improve the fuel efficiency of electric vehicles (EV), hybrid electric vehicles (HEV), and other xEVs [1]. Such this PV system is supposed to high voltage and sufficient power by connecting PV submodules in series and parallel to charge the battery system in xEVs. However, as shown in Fig. 1, in the PV system, variations in power and current difference occur among the submodules owing shading, pollution, aging, and failure. Moreover, since the PV submodules are connected in series, the output current of the PV system is limited to the smallest current of PV submodule The associate editor coordinating the review of this manuscript and approving it for publication was Zhehan Yi . resulting in degraded output power. As a result, the DPP converter has power generation efficiency is considerably degraded [2], and excessive power consumption, called hot spotting, occurs due to the reverse bias of the cells in the PV system [3].
To overcome the shortcomings of the PV system, various converters such as DC optimizer and cascaded converters have been introduced to overcome the shortcomings of the PV system [4], [5], [6], [7]. However, conventional converters have to handle the rated power of PV submodule or the whole system, large volume transformer and high number of components are required causing high loss and high price of converter [8]. Accordingly, for the reduced olume and price, the Differential Power Processing (DPP) converter has been proposed [9]. This DPP converter shows low volume, price, and high efficiency since it handles only power differences between submodules. This DPP converters have several architectures such as PV-PV [10], [12], [13], PV-Bus [14], [15], [16], and PV-isolated port (IP) [17], [18], [19]. The PV-Bus and PV-IP converter topologies use the bidirectional flyback converter, and the PV-PV converter topology commonly uses the bidirectional buck-boost converter.
In the PV-IP architecture shown in Fig. 2, the bi-directional flyback DPP converter is installed between the PV submodules and IP, and the current and power difference between these submodules is reduced by injecting surplus current of high power submodule to the low power submodule. As shown in Fig. 2(b), provided that the number of submodules of a conventional flyback DPP converter are N, N transformers, 2N switches, and 2N driving circuits are required. That is, the volume of the converter increases as the number of components increases. Moreover, a conventional flyback DPP converter has a magnetizing current offset because the magnetizing inductance stores and transmits energy. Thus, the volume of the transformer increases. As a result, it is difficult to use for applications that require low volume and high price competitiveness. To solve this problem, we aim to introduce a new DPP converter with an active clamp structure and a single integrated planar transformer. Since the proposed converter integrates multiple transformers of conventional DPP converter into a single integrated planar transformer with PCB windings, it is possible to minimize the volume and cost of transformer. In addition, the secondary side circuits can be removed because not only the power imbalance is compensated through the primary side of DPP converter, but also transformer is reset by the active clamp structure. Therefore, the proposed converter can provide high power density, high price competitiveness as well as high efficiency. These characteristics make the proposed converter promising to not only DPP converter for PV applications but also other power balancing circuits and applications such as the battery cell balancing [20], [21]. In this paper, the efficiency was improved by strengthening and supplementing the PCB pattern weight and width to reduce the conduction loss of the converter presented in [22].
In order to support the effectiveness of the proposed converter, the contributions of this work can be summarized as follows.
1) Section II describes the characteristics and operating principles of the proposed DPP converter. 2) Section III presents the design method for the proposed DPP converter. 3) Section IV shows experimental results verifying the validity and effectiveness of the proposed DPP converter. 4) The conclusions are presented in Section V.

II. THE PROPOSED DPP CONVERTER
A. CHARACTERISTICS OF THE PROPOSED DPP CONVERTER Fig. 2 shows the structure of the conventional DPP converter, and as shown in Fig. 2, the conventional DPP converter which VOLUME 11, 2023   is composed of various flyback converters adopts multiple transformers and secondary side for balancing the power. However, although the proposed DPP converter is an integrated converter, it is not a method of integrating converters by increasing the number of components, but a method of integrating converters by reducing the number of components. A proposed DPP converter shown in Fig. 3 shows single integrated transformer without any secondary side circuits. This is because that the excessive energy induced by the high power submodule is stored in the leakage inductor of the integrated transformer, and this energy is transferred to other submodules through the primary winding and switches. This means that the magnetizing inductor only transfers the excessive power to other submodules as the common transformer which has low magnetizing current offset. In addition, since the active clamp structure of the proposed converter resets the transformer, the integrated transformer shows zero-offset magnetizing current which can provide low volume of transformer and high efficiency by zero-voltage switching (ZVS) characteristics. Moreover, leakage inductance is reduced due to PCB windings of planar transformer, thereby it is possible to achieve low RCD snubber losses. As a result, due to the aforementioned features, high power density, high price competitiveness, and high efficiency can be achieved. Fig. 4 and Fig. 5 show the key waveforms and operational principles of the proposed DPP converter. For the convenience of analysis, two PV submodules were configured and the analysis of the proposed converter was performed concerning six modes. For the analysis, it was assumed that the power generation of the PV1 submodule was greater than that of the PV2 submodule, the turn ratio of a planar transformer was the same, and all leakage inductances of a planar transformer were equal.

B. OPERATION PRINCIPLE OF THE PROPOSED DPP CONVERTER
As illustrated in Fig. 4 and Fig. 5, the primary switches Q P1 and Q P2 are turned on and off at the same time, and the switch Q A operates opposite to Q P1 and Q P2 .
Mode 1 (t 0t 1 ): Both Q P1 and Q P2 remain turned on, and since the PV2 voltage (v PV 2 (t)) is lower than the PV1 voltage (v PV 1 (t)), the voltage applied to the magnetizing inductance L m (v Lm (t)) has a value between v PV 1 (t) and v PV 2 (t). The voltages of leakage inductance L lkg1 and L lkg2 are as follows. v Accordingly, the current of L lkg1 (i Llkg1 ) increases, and the current of L lkg2 (i Llkg2 ) decreases. The current of L lkg2 is as follows.
(1)-(3) Through, the magnetization inductor voltage v Lm (t) can be expressed by the following equation. v Mode 2 (t 1t 2 ): Both Q P1 and Q P2 are turned off, and the body diodes of Q A and Q P2 conduct. The magnetization current i Lm (t) is reduced because the clamp capacitor is charged through the active clamp. The voltages of L lkg1 and L lkg2 are as follows. v Accordingly, the current of L lkg1 (i Llkg1 ) decreases, and the current of L lkg2 (i Llkg2 ) increases. Mode 3 (t 2t 3 ): The Q P1 and Q P2 maintain a turn-off state, and the body diode of Q A conduct. In addition, the primary 5670 VOLUME 11, 2023 side of the PV1 and PV2 submodule are commutated, so the current flowing through the PV2 primary side is reduced, and the current flowing through the PV1 primary side is increased. The energy stored in L lkg2 on the primary side of the PV2 submodule is consumed by the RCD snubber, and the current flowing on the primary side of the PV1 submodule charges the clamp capacitor.
Mode 4 (t 3t 4 ): ZVS is possible because the Q A is turned on. Since clamp capacitor voltage -V Cc1 (t) is applied to the magnetization inductor voltage, the magnetization current reduces from a positive value to a negative value. Accordingly, the transformer is reset.
The Q A is turned off, and the body diodes of Q P1 and Q P2 conduct. In addition, the primary side of the PV1 and PV2 submodule are commutated, so the current flowing through the PV2 primary side is reduced, and the current flowing through the PV1 primary side is increased.
Mode 6 (t 5t 6 ): It starts when the primary side of the PV1 and PV2 submodules commutation is finished, and since the body diode of Q P1 and Q P2 conduct, ZVS turn-on is possible. Since the voltage applied to L lkg1 and the voltage applied to L lkg2 are the same as (1), i Lkg1 (t) increases and i Llkg2 (t) decreases. VOLUME 11, 2023

III. ANAYSIS OF PROPOSED DPP CONVERTER
For the analysis of the proposed DPP converter, the commercial PV system for solar loop of xEVs as shown in    Table 1 is adopted, and the analysis was conducted provided that the PV1 generates maximum power and other submodules cannot generate any power since this condition represents worst-case conditions that maximum power difference between submodule occurs.

A. BUS VOLTAGE AND CLAMP CAPACITOR VOLTAGE
The Bus voltage of the proposed converter is shown in Fig. 3. Since all PV submodules are connected in series, the bus voltage can be represented by (7). In addition, since the active clamp capacitor voltage can be expressed as a relationship between duty and voltage of the PV submodules, it can be expressed as (8).

B. PRIMARY SIDE CURRENT OF THE PROPOSED DPP CONVERTER
The primary current is equal to the switch current and is proportional to the RCD snubber loss and the switch turnoff loss. Fig. 6 shows the equivalent circuit when the primary switch turned on state and two PV submodules are connected in series. In the equivalent circuit, the magnetizing inductance was ignored because it is very large compared to the leakage inductance. In addition, it is shown in the equivalent circuit in consideration of the switch-on resistance (R ds_on ) and wire resistance (R wire ). The PV1 input capacitor voltage (v cpv1 (t)) can be obtained through the equivalent circuit in Fig. 6, and this is shown in (9).
In (9), ω = 1/ L eq C eq , Q = √ L eq / R eq √ C eq , β = 1 − 1/(4Q 2 ), θ = cos −1 (1/(2Q)), and v CPV1 (0), v CPV 2 (0), i Llkg1 (0), i Llkg2 (0) are initial values. In addition, C eq is the equivalent capacitance of C PV 1 and C PV 2 . The leakage inductor current is shown in (10).  It may be seen that the leakage inductor current in (10) is affected by a small change in the input capacitor voltage. Moreover, a small voltage is applied to the leakage inductor by switch-on resistance and wire resistance. Accordingly, the leakage inductor current performs a Quasi-resonant operation. As a result, the leakage inductor current is small at the switch turn-off, so the switch turn-off loss and RCD snubber loss are reduced.

C. SWITCH LOSS AND RCD SNUBBER LOSS OF PROPOSED CONVERTER
The root-mean-square (RMS) current flowing through the conventional flyback DPP converter and a proposed DPP converter are shown in Table 2. In addition, the specifications are shown in Tables 1 and 3. The losses of the conventional DPP converter and a proposed DPP converter are shown in Table 2 and Fig. 8.
The proposed DPP converter reduced the number of components by adopting the active clamp structure, and as shown in Fig. 7, all switches perform ZVS turn-on, so switching loss can be reduced. That is, as shown in Table 2 and Fig. 8, the switching loss of a proposed DPP converter is less than that of the conventional DPP converter. The switching loss may be obtained by (11)- (14).
The RCD snubber design should be based on when maximum power is generated in only one of the PV2, PV3, or PV4 modules. the RCD snubber capacitor voltage V cc should be set in consideration of the leakage inductor current and the switch voltage of the submodule that is being generated, and the RCD snubber resistance can be obtained through (15).
The conventional flyback DPP converter has many windings, so its leakage inductance is large, and the number of RCD snubbers is large by the primary side and secondary side circuits. However, since the proposed DPP converter has a small number of windings and uses PCB windings, the leakage inductance is very small. Moreover, the number of RCD snubbers is small by integrating multiple transformers into a single planar transformer and applying the active clamp structure. In addition, since the leakage inductor current is similar to a Quasi-resonant, the peak value of the leakage inductor current decreases when switches are turned off.
Thus, a small RCD snubber loss occurs.

D. TRANSFORMER DESIGN AND LOSS
Generally, a conventional flyback DPP converter stores surplus energy in the magnetizing inductance of transformer, so there is magnetizing current offset in the transformer, and high flux density in the core can be induced. Therefore, to avoid saturation of the core, the high turn number of windings of the transformer is required despite the increased volume of the transformer. Moreover, since a lot of transformers must be used for power balancing, the conventional DPP converter suffers from considerably large volume transformers. On the other hand, the integrated planar transformer of proposed DPP converter only transfers surplus power to the low power submodules rather than stores this energy in the magnetizing inductor. In addition, the transformer reset is accomplished through the active clamp structure, and it is possible to remove the magnetizing inductor current offset showing low flux density in the core. Accordingly, it is possible to reduce the primary turn number of transformer to 4 as shown in Fig.9 by increasing flux variation ( B) appropriately, and the each primary winding can be implemented in one layer. Therefore, as shown in Fig. 9, by utilizing 6-layer PCB, all primary windings of the integrated transformer can be realized through PCB winding, this results in the daramatically reduced volume of the transformer in spite of the integration of several transformers. Table 3 shows the transformer area product (A P ) of the conventional flyback DPP converter and a proposed DPP converter. The transformer A P of the proposed converter can be represented as in (16) considering the number of multi-windings (N PV ,module ) on the primary side.
It can be seen from Table 3 that the transformer volume of a proposed DPP converter was reduced by 81% compared to the total transformer volume of the conventional flyback DPP converter. Although the proposed converter significantly minimize the volume of the transformer, the loss of the transformer shown in Fig. 8 is slightly increased because of the relatively large flux variation and high current density of the PCB windings. Nevertheless, the total loss of the proposed converter is noticeably reduced owing to the low switching and RCD snubber losses. In order to support the consistency of the design result, a FEM simulation based on ANSYS Maxwell was also conducted, and Fig. 10 shows simulation results about the flux density, current density at the peak value and RMS value of leakage inductor current, and transformer loss for one switching cycle. As shown in Fig. 10, the maximum flux density (B max ) and the current density are almost the same as the maximum flux density and the current density in Table 3. Moreover, the results of the transformer loss analysis of the proposed DPP converter in Fig. 8 are almost equal to the result of the FEM analysis in Fig. 10.
As aforementioned, the proposed integrated planar transformer have strength in that high power density, and high price competitiveness since the multiple cores and windings of the transformer integrated into single transformer and PCB windings. In addition, a proposed DPP converter can maximized power density by removing the secondary side and applying an active clamp structure. Moreover, the RCD snubber loss was reduced by negligible leakage inductance through the PCB windings. Since all switches perform ZVS turn-on by applying the active clamp structure, the switching losses also can be reduced. Therefore, a proposed DPP converter is a very promising converter for the PV system and other applications where power imbalance occurs.

IV. EXPERIMENTAL RESULTS
As shown in Fig. 11, the prototype was tested using power analyzer (WT5000), DC power supply (DSP1500WS),  TMS320F28379D LaunchPad and the PV Simulator function of DC power supply. Fig. 12 illustrates a prototype of the proposed DPP converter, which is designed to solve the power imbalance between PV submodules when PV submodules are connected in series. The design specifications of the proposed DPP converter and conventional flyback DPP converter are shown in Tables 2 and 3. To prove the performance of the proposed DPP converter, experiments were conducted considering the worst-case condition as shown in Fig. 13(a) and Fig. 13(b). The worst-case condition means that the PV1 submodule VOLUME 11, 2023   generates 60 W power, and other submodules generate 0 W power. Therefore, the DPP converter supplies the 45 W      Fig. 14, a surplus current of the PV1 submodule is injected into other submodules to reduce power imbalance. The proposed DPP converter has a small voltage variation of the leakage inductor caused by the input capacitor voltage ripple. In addition, since a small voltage is applied to the leakage inductor current performs a Quasi-resonance operation. Therefore, since the peak leakage inductor current decreases at switch turn-off, the switching loss and the RCD snubber loss are reduced. As a result, the proposed DPP converter can provide higher efficiency than the conventional flyback DPP converter. Fig. 15 shows the magnetizing current waveform. As illustrated in Fig. 15, the magnetizing current shows zero offset current due to the current second balance of the clamp capacitor. In addition, since the magnetizing current is zerooffset, and thus, even if several transformers are integrated, the transformer volume significantly reduces. Fig. 16 shows the ZVS characteristics of worst-case conditions. Since the magnetizing current shows zero offset, as the load increases, the conduction period of the body diode of the switch reduces. Therefore, the 100% load condition is the worst-case condition for the ZVS operation. As illustrated in Fig. 16, it can be seen that ZVS is possible under full load conditions, so it is possible to satisfy ZVS under all conditions. Fig. 17 show the experimental waveforms of the proposed DPP converter with Maximum Power Point Tracking (MPPT) operation, and Fig. 18 shows a PV simulator that simulates a PV1 submodule under worst-case conditions. As shown in Fig. 17, before the DPP converter operates, the voltage imbalance among submodules has occurred, and the output voltage is 10V in the PV1 submodule and 6V in the PV2-PV4 submodules. When DPP is on, soft-start-up is performed to prevent an inrush current, and an output power voltage is charged. Therefore, the imbalance between modules is reduced. After DPP on, MPPT control was performed using the Solar DC Converter (SDC), as shown in Fig. 13, so that maximum power generation could be achieved in the PV system. Fig. 19 shows the efficiency of the proposed DPP converter and the conventional DPP converter. The efficiency of the proposed DPP converter is considerably higher than that of the conventional flyback DPP converter, and it can be seen that the proposed DPP converter is about 7% higher under full load conditions. In the case of the proposed converter, more than 95% efficiency under 10-60% load. In addition, the maximum efficiency achieved 96.654% efficiency about 20% load. Therefore, since the proposed converter provides high efficiency and power density, because of these characteristics, it can be applied not only to DPP converters but also to battery cell balancing and others.

V. CONCLUSION
In this paper, a novel DPP converter was developed by adopting and integrated planar transformer and an active clamp structure. This proposed DPP converter significantly reduced the transformer volume by applying an integrated transformer. In addition, due to the active clamp structure of the proposed converter, it is possible to eliminate the secondary side resulting further improvement of power density. Moreover, not only low RCD snubber losses but also low switching losses are obtained due to low leakage inductance of PCB windings and ZVS feature of active clamp respectably. Since all switch gate signals except the clamp switch are driven by a single PWM signal, control and driving circuits may be simplified. Thus, due to these features, the proposed converter can be a very promising candidate for the DPP converter in PV applications and other balancing circuits when the power difference occurs between modules.