The problem we set out to solve
Renewable sources such as photovoltaic panels and fuel cells typically produce relatively low DC voltage, while a grid, inverter, or DC bus often requires approximately 380–400 V. High step-up DC-DC converters are designed to close that gap.
In existing high-gain converters, we repeatedly encountered several tradeoffs:
- Some designs provide wide, adjustable gain only by adding stages or components as the target voltage increases.
- Many voltage-multiplier-cell converters use a large number of parts without achieving proportionately high gain.
- Several two-switch converters require two carefully timed gate signals, complicating the controller.
- Some bidirectional or soft-switched converters limit switch stress, but do not provide enough step-up gain for demanding high-voltage applications.
We wanted a converter with genuinely high voltage gain, simple control, and effective recycling of stored energy in parasitic and passive elements—without a large collection of extra components.
The core idea
The converter combines one three-winding coupled inductor with two switched inductors and a passive clamp.
- A three-winding coupled inductor adds two independent design parameters. Primary, secondary, and tertiary windings provide two turns-ratio variables, widening and shaping the gain range without requiring extra conversion stages.
- Both switches share one gate pulse. The power switches turn on and off together, avoiding a pair of independently timed signals and making the controller straightforward to implement.
- Stored energy is reused. Multiple capacitive and inductive transfer paths move energy from passive and parasitic elements toward the load instead of allowing it to circulate or dissipate. The passive clamp also limits switch-voltage spikes.
- Two series-connected output ports share the voltage. Dividing the output across two capacitor ports keeps the stress on each output diode and capacitor manageable even when the combined output voltage is very high.
- Analysis and control remain simple. The converter has only two main continuous-conduction operating modes. Its boundary and two possible discontinuous-conduction cases were also analyzed to support design across the full operating range.
What we built and tested
A 300 W prototype was built with a 24 V input, an 8:1 output-per-port winding configuration (N1 = 1, N2 = 3), a duty cycle of 0.48, and a switching frequency of 50 kHz. An STM32F746 microcontroller generated the synchronized switching pulses.
- The theoretical model predicted an output near 382 V, while the measured output was approximately 370 V—the expected difference once real component losses are included.
- Measured efficiency reached 95% at 263 W output and remained in the mid-90% range across the tested load span.
- Both switch-voltage waveforms and all three internal-diode waveforms closely matched the predicted voltage stresses and showed only small spikes, confirming the passive clamp’s effectiveness.
- With a 30% load step at two operating points, the converter returned cleanly to its target output voltage after a short transient.
Why this matters
In comparison with a broad set of non-isolated high step-up converters, many designs with equally few switches produced lower gain. Other two-switch converters approaching this gain still required independently timed gate signals rather than one shared command. Above a duty cycle of approximately 0.5, this converter delivered the highest voltage gain in the comparison while retaining only two operating modes and a relatively simple controller.
Ultra-high voltage gain usually comes at the price of more components, more complex control, or greater switch stress. This design minimizes those costs, making it relevant to renewable-energy and DC-microgrid applications where reliability and simple control are as important as the gain itself.