Optimizing ZVS Range: Clamp Capacitance & Dead-Time

The Core Condition for ZVS

Before we look at the individual parameters, let's recall the fundamental requirement for achieving ZVS on the main switch. There must be enough energy in the transformer's inductance to completely discharge the capacitance at the switch node during the dead-time. Conceptually, this is:

$\text{Inductive Energy}\ge \text{Capacitive Energy}$
$\frac{1}{2}L\cdot I_{\text{valley}}^2\ge \frac{1}{2}{C}_{\text{lump}}\cdot V_{\text{sw}}^2$

Here, $I_{\text{valley}}$ is the negative current flowing at the start of the transition, and $C_{\text{lump}}$ is the total parasitic capacitance at the switch node. All the design parameters we'll discuss are ways to ensure this inequality holds true over the desired range of operating conditions (input voltage, output load).

Parameter 1: Dead-Time - The "When"

Dead-time is the short duration when both the main and auxiliary switches are turned off. It's the most direct, and often adjustable, parameter for controlling ZVS. In an active clamp flyback, there are two critical dead-time intervals:

1. Main Switch OFF to Clamp Switch ON: This dead-time must be long enough for the primary current to charge the main switch's parasitic capacitance (C_oss) and raise the switch node voltage high enough to forward-bias the clamp switch's body diode. This transition is usually very fast.

2. Clamp Switch OFF to Main Switch ON: This is the most critical dead-time for achieving ZVS. It must be precisely long enough for the negative valley current ($I_{\text{valley}}$) to discharge the lumped switch node capacitance (C_lump) to zero.

The trade-offs are crucial:

• Too Short: The switch node voltage won't reach zero before the main switch turns on, resulting in hard switching and significant power loss.
• Too Long: After the voltage hits zero, the negative current will flow through the main switch's body diode. While this maintains the ZVS condition, body diodes have higher voltage drops and are less efficient than the MOSFET channel. Excessive conduction here leads to unnecessary losses. The timing needs to be just right.

This is why modern controllers often have adaptive dead-time control, sensing the switch node voltage to turn on the main switch at the optimal moment.

Parameter 2: Clamp Capacitance (C_clamp) - Shaping the Resonance

As we learned previously, the clamp capacitor (C_clamp) forms a resonant tank with the transformer's leakage inductance (L_leak). The value of this capacitor has a significant effect on the current waveforms and, consequently, the ZVS behavior, especially at high input voltages.

• How it works: C_clamp and L_leak resonate when the clamp switch is on. The resonant frequency, $f_{res}=1/(2\pi \sqrt{L_{leak}{C}_{clamp}})$, determines the shape and duration of the current that recharges the magnetizing inductance in the negative direction.
• Impact of value:
  • A larger C_clamp creates a lower resonant frequency. This results in a slower, more sinusoidal current waveform. This can reduce the RMS value of the current, which in turn lowers conduction losses and improves efficiency, particularly at high input voltages.
  • A smaller C_clamp leads to a faster, sharper resonance.

The choice is a trade-off between efficiency, component size, and the desired ZVS characteristics.

For a design, the capacitance can be calculated to ensure the resonant period aligns with the operating duty cycle. The onsemi paper "High-Density AC-DC Power Supplies..." gives a design equation for this:

Parameters 3 & 4: Magnetizing Inductance (L_mag) & Switching Frequency (f_sw) - Generating the Energy

To achieve ZVS, we need to generate that negative valley current ($I_{\text{valley}}$). The amount of current ripple, and therefore the depth of this valley, is primarily controlled by the magnetizing inductance and the switching frequency.

The fundamental inductor equation is $V=L\frac{di}{dt}$. Rearranged, the current ripple is $\Delta I=\frac{V\cdot t_{on}}{L_{mag}}$.

• Magnetizing Inductance (L_mag): A lower L_mag will result in a larger current ripple ($\Delta I$) for the same on-time. This makes it easier to achieve the required negative valley current for ZVS, especially under light load conditions. However, a larger ripple current also means higher conduction losses throughout the circuit.
• Switching Frequency (f_sw): As the switching frequency increases, the on-time (t_on) for a given duty cycle decreases. To maintain the same negative valley current, the inductance might need to be adjusted.

Modern controllers exploit this relationship by implementing frequency foldback or frequency modulation. As the load decreases or input voltage increases, the controller increases the switching frequency. This shortens the on-time, reducing the current ripple and keeping the negative valley current just sufficient for ZVS, thereby minimizing circulating current losses and maximizing efficiency.

Conclusion

You now have a map of the key "control knobs" for an active clamp flyback converter. You've seen how they are not independent but are all interconnected in the service of achieving the core ZVS energy-balance equation.

Key Takeaways:

• ZVS Range: This is the range of input voltages and output loads over which the converter can successfully maintain zero voltage switching.
• Dead-Time: Provides the time window for the resonant transition. It must be long enough to discharge the switch node capacitance but not so long as to cause excessive body diode conduction losses.
• Clamp Capacitance (C_clamp): Shapes the resonance with the leakage inductance, affecting current waveforms and efficiency. Larger values can soften the currents and improve high-voltage efficiency.
• Magnetizing Inductance (L_mag) & Switching Frequency (f_sw): These parameters together determine how much negative current is generated. Lower inductance or lower frequency leads to more negative current, extending the ZVS range to lighter loads but increasing circulating currents.

In our next module, we will broaden our perspective by examining a completely different ZVS topology: the Phase-Shifted Full-Bridge (PSFB) converter. You'll see how a different arrangement of switches and a different control strategy can also be used to achieve the same goal of eliminating switching losses, giving you another important pattern to recognize in high-power designs.