Access Official Technical PDF

Share Summary:LinkedInEmail

Connecting a large photovoltaic (PV) plant to a medium-voltage grid typically requires a line-frequency transformer to raise the inverter output voltage. A modular medium-voltage converter offers another approach: series-connected power modules build the required grid voltage, while high-frequency transformers within the modules provide galvanic isolation and voltage conversion. This architecture is intended to reduce reliance on bulky line-frequency equipment while retaining a modular, scalable design.

The proposed system uses current-source, isolated buck-boost converters as submodules. Its performance depends not only on the overall architecture, but also on the selected submodule topology, operating mode, PV input voltage, and design priorities.

How a modular medium-voltage converter works

Each submodule accepts power from a PV array and produces an output voltage. Within each phase, the outputs of multiple submodules are connected in series, so their voltages combine to reach the level needed for grid connection. Adding modules can support changes in system voltage or power, and the modular arrangement can make it possible to replace an individual failed module rather than treating the converter as a single unit.

A high-frequency transformer in each submodule provides electrical isolation and contributes to voltage boosting. Operating the transformer at a higher frequency can reduce its size and weight compared with a line-frequency transformer. The transformer also helps mitigate electromagnetic interference. The proposed arrangement is designed to reduce common-mode and module voltage-imbalance concerns associated with PV systems.

At the PV input, the candidate converters have continuous current. This supports maximum power point tracking (MPPT) without requiring large input electrolytic capacitors. Smaller filtering requirements and the potential to avoid electrolytic capacitors are relevant design considerations because capacitor type and operating temperature affect converter reliability.

Four isolated converter options

The study evaluates four buck-boost submodule topologies: C5 (Ćuk), F5, G5 (SEPIC), and P5. Each can be coupled to a high-frequency isolation transformer, but their energy-transfer mechanisms and passive-component requirements differ.

  • C5, G5, and P5: Energy transfers between the transformer’s primary and secondary sides without being stored in its magnetic core during the transfer interval. This allows a smaller core than a design that relies on temporarily storing the transferred energy.
  • F5: Energy is temporarily stored in the transformer core and released later. The energy transferred is therefore related to the core’s capacity to store magnetic energy, which influences the core size.

All four candidates can provide continuous input current, but their inductance, capacitance, ripple, and stored-energy requirements are not identical. These differences affect component selection and the expected size of each submodule. They also mean there is no universally best topology: the suitable option depends on operating conditions and which design criteria matter most.

Balancing efficiency, size, and component count

The comparison considers power losses, voltage and current ripple, component stresses, and the energy stored in passive elements. The analysis finds that the preferred candidate can change with PV array input voltage and the number of modules. For the example system analyzed, C5 provides a compromise under the stated conditions. F5 shows the highest efficiency in that particular comparison, but requires the largest estimated submodule size. P5 has the lowest estimated stored energy, yet uses more passive elements.

These are design trade-offs, not fixed rankings for every PV installation. A designer may prioritize efficiency, physical footprint, reliability considerations, or a smaller component count. Input voltage and module count also affect current levels and losses. Component values must be chosen to control ripple and keep the input current suitable for MPPT operation.

The system can operate in two modes. In Mode 1, successive submodules are supplied by different PV arrays and include a DC offset in their output voltages. In Mode 2, successive submodules are supplied by the same PV array and each operates for half a cycle without that offset. At equal output-voltage peaks, Mode 2 can require twice the PV input current. The analysis reports lower semiconductor and copper losses for Mode 2, while Mode 1’s balanced waveforms are more compatible with conventional d-q-o frame control. Mode selection therefore involves both electrical and control considerations.

Validation and operation under partial shading

The proposed architecture was assessed using mathematical analysis, computer simulations, and a scaled-down three-phase prototype rated at 5 kVA and 1 kV, with four modules per phase. The prototype used a digital signal processor for control, and its configurable submodules allowed the different converter candidates to be tested. Experimental results demonstrated operation of the candidate systems and were compared with simulation results.

Partial shading is another important operating condition for PV plants. When some arrays receive less sunlight, their available power differs from that of unshaded arrays. The proposed system can adjust submodule duty ratios and output references to accommodate the difference and seek to extract the available power from shaded and unshaded modules. This approach depends on the controller having information about the PV array characteristics. The control design and a suitable MPPT system remain matters for further development.

Practical implications for PV system design

For engineers and system integrators, the central benefit of the architecture is the combination of series-connected voltage-building modules and high-frequency isolation at the submodule level. It provides a route to medium-voltage grid connection without relying on a bulky line-frequency transformer, while preserving modularity. However, choosing a submodule requires evaluating the complete design rather than selecting on efficiency alone.

Assess the PV input voltage, required module count, current and voltage ripple limits, passive-component footprint, and preferred operating mode together. Compare the four topologies against those priorities, and account for how partial shading and control requirements affect operation. The practical takeaway is that a modular medium-voltage converter can support scalable PV grid integration, but its system-level advantages depend on matching the topology and control approach to the installation’s electrical and operational requirements.

Access Official Technical PDF

About The Author