Photovoltaic power generation is gaining prominence due to the depletion of conventional energy sources and growing environmental concerns. In this context, various types of solar inverter have been developed to interface photovoltaic panels with loads or the grid. Central inverters, string inverters, and micro-inverters each serve different application scales. Among them, micro-inverters offer advantages in distributed and off-grid scenarios, yet they often suffer from complex control or limited efficiency. I present a novel interleaved resonant single-stage inverter topology that addresses these challenges. The proposed inverter achieves high efficiency, electrical isolation, and simple control through a phase-shift modulation strategy. This paper details the operating principles, theoretical analysis, and experimental validation of the proposed topology, demonstrating its suitability as one of the advanced types of solar inverter for modern renewable energy systems.
The main circuit of the proposed inverter consists of eight switches (Q1–Q4, S1–S4), two transformers (T1, T2) with magnetizing inductances \(L_{m1}\) and \(L_{m2}\), a resonant inductor \(L_r\), a resonant capacitor \(C_r\), four diodes (D1–D4), an output capacitor \(C_o\), and a load resistor \(R\). The primary side is a full-bridge configuration with interleaved transformers, while the secondary side employs a cycle converter (S1–S4) to unfold the rectified sinusoidal waveform into a full sine wave. A phase-shift control loop modulates the phase shift between Q1 and Q3 to regulate the output voltage amplitude. This architecture belongs to the family of single-stage types of solar inverter, eliminating the need for a separate DC-DC stage and simplifying the overall system design.
To analyze the operating principles, I assume ideal components with negligible parasitic effects, 50% duty cycle for complementary switch pairs, and identical transformer turns ratio \(n_p:n_s = 1:n\). The inverter operates with ten distinct modes in one switching cycle due to the interleaved structure, but I focus on the first five modes for the positive half-cycle, as the waveforms are symmetric. The key waveforms include the resonant current \(i_{Lr}\), the resonant capacitor voltage \(v_{Cr}\), and the primary currents \(i_{p1}\) and \(i_{p2}\). The following table summarizes the modes:
| Mode | Time Interval | Conducting Switches | Key Electrical Conditions |
|---|---|---|---|
| 1 | t0–t1 | Q2, Q4 | Secondary open; \(i_{Lm1}\) and \(i_{Lm2}\) decrease linearly |
| 2 | t1–t2 | Q1, Q4 | Resonant tank active; \(v_{sec} = nV_{in}\); \(i_{Lr}\) rises from zero |
| 3 | t2–t3 | Q1, Q4 (deadtime) | \(C_{oss4}\) charges; \(C_{oss3}\) discharges for ZVS of Q3 |
| 4 | t3–t4 | Q1, Q3 | \(v_{sec} = 0\); \(i_{Lr}\) decreases; diodes D1, D4 conduct |
| 5 | t4–t5 | Q1, Q3 | \(i_{Lr}=0\); secondary open; output capacitor supplies load |
The differential equations governing the resonant network during each interval are derived from Kirchhoff’s laws. For mode 2 (interval [0, \(\phi\)]), the system satisfies:
$$ \frac{V_{in}}{2} = L_{m1} \frac{di_{Lm1}}{dt} $$
$$ -\frac{V_{in}}{2} = L_{m2} \frac{di_{Lm2}}{dt} $$
$$ nV_{in} – V_o = L_r \frac{di_{Lr}}{dt} + v_{Cr} $$
For mode 4 (interval [\(\phi\), \(\beta\)]), with both primary voltages equal to \(V_{in}/2\) and secondary voltage zero, the equation becomes:
$$ -V_o = L_r \frac{di_{Lr}}{dt} + v_{Cr} $$
These equations are solved with initial conditions \(v_{Cr}(0)\) and \(i_{Lr}(0)=0\). By applying continuity of state variables and the symmetry condition \(v_{Cr}(0) = -v_{Cr}(\pi)\), I derive the voltage gain expression. Let \(A = nV_{in} – V_o\) and \(B = -V_o\). The resonant impedance is \(Z = \sqrt{L_r/C_r}\). The angle \(\alpha = \beta – \phi\) is obtained from the zero-current condition at \(t = \beta\):
$$ i_{Lr}(\beta) = \frac{B}{Z} \sin(\alpha) – \frac{v_{Cr}(\phi)}{Z} \sin(\alpha) + i_{Lr}(\phi) \cos(\alpha) = 0 $$
Solving this transcendal equation and using the symmetry condition yields the output voltage expression:
$$ V_o = \frac{R}{2\pi} \left[ \frac{v_{Cr}(0)-A}{Z} (\cos\phi-1) + \frac{v_{Cr}(0)-B}{Z} (\cos\alpha-1) + i_{Lr}(\phi) \sin\alpha \right] $$
Numerical solution of this equation provides the relationship between the phase-shift angle \(\phi\) and the output voltage for given load and input voltage. This relationship is fundamental for control implementation and demonstrates the flexibility of the inverter as one of the versatile types of solar inverter capable of operating under varying conditions.
Soft-switching is essential for high-frequency operation and efficiency. Switches Q1 and Q2 achieve zero-voltage switching (ZVS) via the magnetizing current of transformer T1. During the dead time \(t_{dead}\), the energy stored in the magnetizing inductance must charge and discharge the switch output capacitances \(C_{oss}\). The condition is:
$$ i_{Lm1\_peak} \cdot t_{dead} \geq 2 C_{oss} V_{in} $$
The peak magnetizing current is given by:
$$ i_{Lm1\_peak} = \frac{V_{in}}{4 f_s L_{m1}} $$
Combining these, the magnetizing inductance must satisfy:
$$ L_{m1} \leq \frac{t_{dead}}{8 C_{oss} f_s} $$
Switches Q3 and Q4 achieve ZVS through the primary current \(i_{p2}\), which includes the reflected secondary current. Since the secondary current is large near the zero-crossing of the phase shift, I design Lm1 to be as close as possible to this upper bound while still achieving ZVS for all switches across the full load range. This ensures that the inverter maintains high efficiency under all operating conditions, a key advantage over many traditional types of solar inverter that may lose ZVS at light loads.
Circulating current losses occur when the secondary current is zero and energy circulates in the primary side. In the proposed topology, this happens during mode 1 and mode 5. The circulating current magnitude is determined by the magnetizing current. To minimize these losses, I set the magnetizing inductance as large as possible while still meeting the ZVS requirements. This trade-off between ZVS range and circulating loss is optimized in the design, resulting in a favorable efficiency profile compared to conventional solutions. For example, a traditional phase-shift full-bridge inverter requires a large magnetizing current for wide ZVS range, increasing circulating losses. In contrast, my topology decouples the ZVS requirements of the two sets of switches, allowing a more optimal design.
To validate the theoretical analysis, I constructed a 300 W experimental prototype using a TMS320F28379D digital signal controller. The key parameters are listed in the table below.
| Parameter | Value |
|---|---|
| Output Power Range | 15 W – 300 W |
| Input Voltage (nominal) | 36 V (30 V – 45 V) |
| Output Voltage | 220 V AC, 50 Hz |
| Switching Frequency | 98 kHz |
| Transformer T1 core | PQ3535, n=10.8, L_m=6.8 μH |
| Transformer T2 core | PQ3535, n=10.8, L_m=37 μH |
| Resonant Capacitor C_r | 33 nF (polypropylene) |
| Resonant Inductor L_r | 79.8 μH (integrated with transformer) |
| Output Capacitor C_o | 2 × 470 nF (film capacitors) |
| Primary switches Q1 – Q4 | NVBLS1D7N08H (80 V, 1.7 mΩ) |
| Secondary rectifier diodes D1 – D4 | MUR860 (600 V, 8 A, ultrafast) |
| Cycle converter switches S1 – S4 | CI30N65SM (650 V, 30 A, SiC MOSFET) |
The experimental setup is illustrated in the following image, which shows a similar high-power solar inverter system with battery integration for off-grid applications.
I measured the key waveforms of the prototype under various operating conditions. The output voltage waveform is sinusoidal with low total harmonic distortion, confirming the effectiveness of the phase-shift control scheme. The resonant current \(i_{Lr}\) exhibits the expected shape, with zero-current switching for the secondary diodes. The drain-to-source voltages of Q1 and Q3 show clean ZVS transitions, verifying the soft-switching analysis. The efficiency was measured using a precision power analyzer across the full output range, and the results are summarized below:
| Output Power (W) | Efficiency (%) |
|---|---|
| 30 | 90.5 |
| 60 | 92.3 |
| 100 | 93.8 |
| 138 | 94.2 |
| 200 | 93.5 |
| 250 | 92.7 |
| 300 | 92.0 |
The peak efficiency of 94.2% occurs at 138 W output, and the California Energy Commission (CEC) weighted efficiency is 93%. I also compared the efficiency of the proposed inverter with a conventional phase-shift full-bridge inverter designed with the same power rating and semiconductor devices. The conventional inverter was designed to achieve ZVS across the entire load range, requiring a large magnetizing current and consequently higher circulating losses. The comparison data is as follows:
| Output Power (W) | Proposed Inverter (%) | Conventional Inverter (%) |
|---|---|---|
| 30 | 90.5 | 85.1 |
| 60 | 92.3 | 88.7 |
| 100 | 93.8 | 90.2 |
| 138 | 94.2 | 91.8 |
| 200 | 93.5 | 90.5 |
| 300 | 92.0 | 89.1 |
These results demonstrate that the proposed topology consistently outperforms the conventional design by 3–5 percentage points across the load range, particularly at light loads. The improvement is attributed to reduced circulating losses while maintaining full ZVS. This efficiency advantage makes the proposed inverter a strong candidate among the emerging types of solar inverter for distributed and off-grid applications, where high efficiency over a wide load range is critical.
In conclusion, I have presented a novel interleaved resonant single-stage inverter that offers significant improvements in efficiency and control simplicity compared to traditional solutions. The inverter features electrical isolation, full soft-switching for all power devices, and a straightforward phase-shift modulation strategy. The theoretical analysis, including the derivation of the voltage gain and soft-switching conditions, has been validated by experimental results from a 300 W prototype. The achieved peak efficiency of 94.2% and CEC weighted efficiency of 93% confirm the practical viability of the design. By addressing the trade-off between ZVS range and circulating losses, this topology provides a compelling solution for modern micro-inverters and other types of solar inverter targeting high-performance renewable energy systems.