Circulating Current Penalty vs. ZVS Range Trade-off

The Freewheeling Phase and Circulating Current

The PSFB's phase-shifting control method creates three main operational states within each half-cycle:

1. Power Transfer: A diagonal pair of switches is on, applying voltage across the transformer and sending power to the output.
2. ZVS Transition (Dead-Time): All switches are off, and stored inductive energy resonates with switch capacitances.
3. Freewheeling / Idling: Either the top two or bottom two switches are on, shorting the primary side of the transformer.

It is this third "freewheeling" state that is the source of the circulating current penalty. During this interval, the voltage across the transformer primary ($V_{AB}$) is zero, and no power is transferred to the load. However, the primary-side circuit still forms a closed loop, allowing current to flow. This current, which circulates in the primary without doing useful work, is what we call circulating current.

The following video provides an excellent intuitive explanation of this "idling stage."

This circulating current causes conduction losses ($I^{2}R$) in the MOSFETs and transformer windings. Since this happens while no power is being delivered, it is a pure loss, directly harming the converter's efficiency. The longer the freewheeling period and the larger the current, the greater the penalty.

The image below shows key waveforms in a PSFB converter. The shaded area clearly highlights the energy that is circulating in the primary during the freewheeling phase.

The Trade-off: ZVS Range vs. Circulating Losses

We now have two seemingly separate issues: the struggle to get ZVS at light load and the efficiency penalty from circulating current. In reality, they are two sides of the same coin.

As we learned, achieving ZVS, especially on the leading leg at light load, requires a minimum amount of primary current. If the reflected load current isn't enough, designers have a common trick: increase the transformer's magnetizing current. The magnetizing current is independent of the load, so it provides a "base" level of current that is always present to help with the ZVS transition. This is often done by adding a small air gap to the transformer core.

This solution, however, comes with a direct penalty. The magnetizing current flows in the primary at all times, including during the freewheeling phase. Therefore, increasing the magnetizing current to extend the ZVS range directly increases the magnitude of the circulating current.

This leads to the central trade-off of the PSFB converter:

• To improve light-load ZVS and reduce switching losses, you might increase the magnetizing current.
• But this increases the circulating current, which in turn increases conduction losses, especially at light load where the freewheeling period is longest.

Let's read a concise summary of this from a PhD thesis on the topic.

The reading confirms the dilemma: you can extend the ZVS range by gapping the transformer (increasing magnetizing current), but you do so "at the expense of increasing the circulating energy in the converter."

Quantifying the Penalty: When the Cure is Worse Than the Disease

Whether this trade-off is worthwhile depends on the specific operating conditions. You are trying to balance a reduction in switching loss against an increase in conduction loss. At high switching frequencies, switching losses can be dominant, so paying a conduction loss penalty might be acceptable. At lower frequencies, the reverse might be true.

The same thesis provides a fantastic real-world example of this trade-off in action. The authors designed a system and measured the power savings from achieving ZVS versus the extra losses introduced by the components used to achieve it.

As you can see from Table 2.1, at 200 kHz and 450V input, the design saved 10.4 W in switching losses by achieving ZVS. However, the additional core losses introduced to make this happen were 18.4 W. The net result is an 8 W increase in total power dissipation. In this case, the cure was worse than the disease!

This is a powerful illustration of the circulating current penalty. A designer must carefully balance the desire for a wide ZVS range against the inevitable increase in conduction and core losses that come with it.

Conclusion

This lesson completes our analysis of the PSFB converter's core operational trade-offs. You are now equipped to look at a PSFB design and understand not just how it works, but also its inherent limitations.

Key Takeaways:

• The PSFB's "freewheeling" or "idling" phase causes a circulating current that flows in the primary without transferring power, creating conduction losses.
• The duration of this freewheeling phase is the source of the duty cycle loss we discussed previously.
• To extend the ZVS range to lighter loads, designers often increase the transformer's magnetizing current.
• This creates a fundamental trade-off: a wider ZVS range (lower switching loss) comes at the cost of higher circulating current (higher conduction loss).
• The optimal design requires a careful balance of these competing loss mechanisms, which is dependent on factors like switching frequency and input voltage range.

In our next module, we will begin exploring the LLC Resonant Converter. You will see how its topology is fundamentally different and specifically designed to overcome some of the PSFB's main drawbacks, such as the circulating current penalty and the limited ZVS range on the leading leg. This will provide a great point of comparison and deepen your understanding of modern power converter design.