Hello! Welcome to the first lesson in our new module on the practical design and optimization of ZVS converters.
In the previous modules, we explored the "how" behind several ZVS topologies—understanding their operating principles and comparing their strengths. Now, we shift our focus to the "what": the specific components that make these advanced circuits possible.
Your goal is to confidently review and optimize power converter PCBs. A crucial step in that process is looking at a component on a schematic, finding its datasheet, and judging if it's the right choice for the job. This lesson is designed to give you that capability. Our objective is to identify the key MOSFET and resonant component parameters critical for ZVS applications.
We will begin with the most important active component, the MOSFET, as its characteristics fundamentally define the performance and limitations of a ZVS converter. Then, we will examine the passive components that form the resonant tank.
The Central Role of the MOSFET
In a ZVS converter, the MOSFET isn't just a simple on/off switch. Its internal parasitic characteristics are actively used to eliminate switching losses. As we've discussed, the core principle of ZVS is to discharge the MOSFET's own output capacitance before the device is turned on, thereby preventing the simultaneous high voltage and high current that cause massive power loss and EMI in hard-switched converters.

To understand how to select a good ZVS MOSFET, we need to dive into the parameters that govern this process.
1. Output Capacitance (Coss) and Stored Energy (Eoss)
The output capacitance, , is the parasitic capacitance across the drain and source terminals of the MOSFET. In ZVS, we use the inductor current to discharge this capacitance during the dead time. However, is not a simple, fixed-value capacitor.
To understand this critical parameter, let's watch a detailed explanation.
Deciphering Coss of power MOSFETs
Dr. Sam Ben-Yaakov provides an excellent breakdown of MOSFET parasitic capacitances. This video explains what C_oss is, why it's highly non-linear, and how manufacturers represent its characteristics in datasheets.
Watch the video from the beginning to 19:16. Focus on these key points: The definition of C_iss, C_oss, and C_rss from the internal capacitances (0:00 - 1:35). The non-linear, voltage-dependent nature of these capacitances (1:35 - 2:03). How datasheet graphs show C_oss changing dramatically with voltage (6:54 - 9:52). The concepts of energy-related (Co(er)) and time-related (Co(tr)) equivalent capacitances, and the much more useful E_oss curve (9:52 - 19:16).
As the video explained, the key takeaways regarding are:
- It's Non-Linear: is very high at low and drops off significantly as voltage increases. This means it takes more current or time to discharge the last few volts across the MOSFET, a crucial factor in achieving full ZVS.
- Energy (Eoss) is What Matters: The energy stored in this capacitance, , is what creates turn-on loss. In a hard-switched converter, this loss is . In a ZVS converter, this energy is recycled to the load or source instead of being dissipated as heat in the MOSFET. Therefore, a MOSFET with lower is inherently better for high-frequency switching, as there's less energy to manage.
- Equivalent Capacitances (Co(er) and Co(tr)): To simplify calculations with the non-linear , datasheets provide equivalent linear capacitances.
- : An energy-related capacitance used to find the stored energy at a specific voltage (typically the MOSFET's max rated voltage).
- : A time-related capacitance used to estimate the time it takes to charge or discharge . This is extremely useful for calculating the required dead time in a ZVS converter. A MOSFET with a lower will require a shorter dead time to achieve ZVS, enabling higher efficiency.
For a practical perspective on how these parameters are used, let's look at an application note for LLC converters, which relies heavily on ZVS.
CoolMOS™ Primary Side MOSFET Selection for LLC ...
This Infineon application note explains the role of MOSFET selection in LLC converters. Its discussion on Coss, dead time, and selection criteria is broadly applicable to all ZVS topologies.
Please read the following sections: Section 2.1 'The effect of MOSFET Output Capacitance (Coss)': This reinforces how the non-linear Coss impacts the discharge time. Section 2.2 'The effect of dead time': Note the formula that directly links dead time to Co(tr). Section 4.1 'Qg and Coss of MOSFET': This summarizes the selection criteria, highlighting the importance of low Co(tr) and low Coss at low VDS.
2. The Body Diode: The Unsung Hero and Potential Villain
In any ZVS topology, the body diode of the MOSFET is forced to conduct during the dead time. This is what clamps the drain-source voltage to near-zero, completing the ZVS transition. However, this action comes with significant consequences.
- Conduction Loss: While conducting, the diode has a forward voltage drop (), typically around 0.7V-1V for silicon MOSFETs. This causes a conduction loss of . Minimizing dead time is key to reducing this loss.
- Reverse Recovery (Qrr): This is the most critical and dangerous characteristic. When the diode has been conducting and the opposing MOSFET in the bridge turns on, the diode must turn off and start blocking voltage. It cannot do this instantly. A stored charge of minority carriers, the reverse recovery charge (), must be swept out first. During this time, the diode effectively acts as a short circuit, leading to a large reverse current spike.
This reverse recovery process causes:
- Massive Power Loss: The energy loss per cycle is . At high voltages and frequencies, this can become a dominant loss mechanism and completely negate the benefits of ZVS.
- Destructive Failure: The large current spike and associated high can cause severe voltage ringing, EMI, and even destroy the MOSFET.
Let's watch a video that visualizes and explains the impact of .
How to GaN 015 Understanding the Impact of Dead time, QRR, and COSS
This video from EPC clearly explains and illustrates the reverse recovery phenomenon and its impact on switching losses. While it promotes GaN FETs (which have zero Qrr), its explanation of the problem in silicon MOSFETs is excellent.
Watch from 04:30 to 11:55. Focus on the explanation of E_oss loss first (4:30-8:31), then pay close attention to the detailed breakdown of the reverse recovery (Qrr) process and its associated losses (8:31-11:55).
Because is so problematic, MOSFET manufacturers have developed devices with improved body diodes. When reviewing a design, especially a high-voltage one, look for:
- Low and : The lower, the better.
- "Fast Recovery" or "Freewheeling" Diodes: Some datasheets will explicitly mention that the MOSFET has an optimized body diode (e.g., Infineon's "CoolMOS™ CFD" or Vishay's "EF Series"). These devices have been processed to drastically reduce , making them suitable for ZVS applications, although often at the cost of slightly higher .
Beware of Zero Voltage Switching
This article from Vishay provides a great summary of the practical issues in ZVS, especially concerning the body diode.
Read the sections covering Figure 3a up to the end of the article. Focus on the discussion of Qrr vs. Qoss, the difference between hard and soft commutation of the body diode, and the introduction of MOSFETs with fast recovery diodes.
Test your understanding!
You are reviewing two 600V MOSFETs for a high-frequency ZVS application.
- MOSFET A: = 50 mΩ, = 8000 nC
- MOSFET B: = 65 mΩ, = 1000 nC
Which one is likely the better choice and why, even though its on-resistance is higher?
Show answer
MOSFET B is the much better choice. The reverse recovery charge () of MOSFET A is extremely high and would lead to massive switching losses and potentially destructive failure at high frequencies. MOSFET B is clearly optimized for fast recovery (it has 8x lower ). The small penalty in conduction loss from its slightly higher is a worthwhile trade-off to drastically reduce the much more severe reverse recovery losses.
3. On-State Resistance (Rds(on)) and Gate Charge (Qg)
While we've focused on ZVS-specific parameters, these fundamental MOSFET characteristics are still critical:
- On-State Resistance (): Determines the conduction loss (). As always, lower is better. However, be aware of the trade-off: for a given silicon technology, achieving a lower often means a larger chip area, which in turn leads to higher capacitances () and gate charge ().
- Gate Charge (): The charge required to turn the MOSFET on. A lower allows for faster switching transitions and reduces the power consumed by the gate driver. This is especially important in high-frequency converters.
Resonant Tank Components
The resonant inductor () and capacitor () form the tank circuit that stores and releases the energy needed to achieve ZVS.
The Resonant Inductor
The inductor (which can be a discrete component or the transformer's leakage/magnetizing inductance) is used to store energy from the DC bus and then use that energy to discharge the MOSFETs' output capacitances during the dead time.
Key parameters for inductor selection are:
- Inductance Value: This is the primary design choice. A smaller inductance allows a larger current to build up, providing more energy for ZVS. This makes it easier to maintain ZVS even at light loads. However, this larger current circulates in the primary, increasing conduction losses. This is a fundamental trade-off between ZVS range and light-load efficiency.
- Saturation Current (): The inductor's peak current must remain well below its saturation rating. If the core saturates, the inductance collapses, leading to a massive current spike that can destroy the switches.
- Core Material: At the high frequencies common in ZVS converters (hundreds of kHz to >1 MHz), core losses are a major concern. Ferrite materials are typically used. The specific material choice is a trade-off between permeability (which affects size), saturation flux density, and frequency-dependent core losses.
- DC Resistance (DCR): The resistance of the copper winding. Lower DCR means lower conduction loss.
The Resonant Capacitor
The resonant capacitor handles very high-frequency, high-amplitude currents.
- Capacitance Value: Chosen along with the inductor to set the resonant frequency of the tank.
- Capacitor Type: This is critical. The capacitor must have very low Equivalent Series Resistance (ESR) to handle the large circulating RMS currents without overheating. Film capacitors (like polypropylene) are the standard choice for this role. For lower power, C0G (NP0) ceramic capacitors are also excellent due to their stability and low loss. You would almost never use an X7R/X5R ceramic or an electrolytic capacitor here, as their high losses and poor stability with temperature and voltage make them unsuitable.
- Voltage Rating: Must be able to withstand the peak AC and DC voltages across it during operation.
Conclusion
You now have a framework for evaluating the most critical components in a ZVS converter. When you pick up a datasheet, you're no longer just looking at the voltage and current ratings. You are looking for the story behind the numbers.
Key Takeaways:
-
For a ZVS MOSFET, the most critical parameters are:
- Low and : To minimize the energy needed to achieve ZVS and allow for short dead times.
- Low Body Diode : To prevent severe switching losses and reliability issues. This is often the single most important parameter in high-voltage ZVS.
- Low and : To minimize conduction and gate drive losses, balanced against the trade-offs with capacitance and .
-
For Resonant Components, the focus is on managing high-frequency losses:
- Inductor: A value that balances ZVS range with conduction loss, made with a low-loss core material.
- Capacitor: A high-quality film or C0G/NP0 type with very low ESR.
In our next lesson, we will build on this knowledge. Now that we know what to look for in components, we'll explore optimization strategies for dead-time and resonant parameters to tune a ZVS converter for maximum efficiency and performance.