Where Does Switching Loss Come From? (And How to Get Rid of It)

Summary

This video dives into the complexities of switching losses in transistors, focusing on conduction and switching losses. It explains the four main types of switching losses—overlap loss, capacitive loss, inductive loss, and reverse recovery loss—along with techniques to minimize them. Key strategies like soft switching, including zero voltage turn-on and zero current turn-off, are explored. The video also touches on challenges with modern devices like gallium nitride (GaN) transistors, and their impact on efficiency and electromagnetic interference (EMI). The overall goal is to understand and reduce losses for improved power conversion efficiency.

Takeaways

• ⚡ All electronic components have losses, with transistors primarily experiencing conduction loss and switching loss.
• 🔹 Conduction loss is due to resistance and voltage drop, and depends on the current through the device.
• ⏱️ Switching loss occurs during transitions between on and off states and is composed of four types: overlap, capacitive, inductive, and reverse recovery losses.
• 🔌 Overlap loss happens when voltage and current waveforms overlap during switching and can be approximated by multiplying voltage, current, and switching time.
• 🔋 Capacitive loss arises from energy stored in the switch node capacitance and is independent of load current.
• 🌀 Inductive loss is caused by energy stored in loop inductance, dissipated during turn-off, and is predictable due to nearly constant inductance.
• ⚠️ Reverse recovery loss occurs in bipolar diodes as they briefly conduct in the reverse direction, creating high current spikes and EMI issues.
• 💡 Soft switching techniques aim to reduce switching losses by ensuring voltage or current is zero during transitions, including zero current turn-on, zero voltage turn-off, zero voltage turn-on, and zero current turn-off.
• 🔧 Zero voltage turn-on and zero current turn-off can recycle stored energy from capacitance and inductance, unlike methods that only reduce overlap loss.
• 🌐 Soft switching also reduces electromagnetic interference (EMI) because slower transitions lower voltage and current spikes.
• 🟢 MOSFETs with body diodes support zero voltage turn-on naturally, while GaN transistors lack effective body diodes, reducing reverse recovery losses but complicating synchronous rectification.
• 📐 Careful dead-time management is critical: too short causes hard switching, too long adds conduction losses through diodes.
• 💻 Practical demonstrations use synchronous buck converters and oscilloscopes to visualize switching waveforms, helping understand energy losses and soft-switching behavior.

Q & A

What are the two main forms of losses in transistors when current passes through them?

The two main forms of losses in transistors are conduction loss and switching loss. Conduction loss is a result of the resistance and voltage drop of the device, while switching loss occurs during transitions between on and off states of the transistor.

What is overlap loss and how is it calculated?

Overlap loss occurs when the voltage and current waveforms overlap during the switching transitions. It can be approximated by multiplying the voltage by the current to get the power, and then multiplying by the switching time to get the energy.

What is capacitive loss and how is it calculated?

Capacitive loss occurs when the energy stored in the capacitance of the switching node is dissipated when the transistor turns on. It can be approximated using the capacitor energy equation (1/2 * C * V^2), although the capacitance varies with voltage.

How does inductive loss occur during switching?

Inductive loss happens when the energy stored in the commutation loop inductance is dissipated, typically as a voltage spike during turn-off. The energy can be approximated using a similar equation to that for capacitive loss, but with inductance instead of capacitance.

What is reverse recovery loss in diodes, and which type of diodes exhibit this behavior?

Reverse recovery loss occurs when a diode continues to conduct in the wrong direction for a brief period before turning off. This is a characteristic of bipolar diodes and not of Schottky diodes. MOSFETs, which contain inherent bipolar diodes, are affected by this loss.

What are the quantities E_on, E_off, and ERR in device data sheets?

E_on represents energy dissipated during the turn-on transition, which includes capacitive and overlap losses. E_off represents energy dissipated during turn-off, including inductive and overlap losses. ERR stands for reverse recovery loss, which is specific to diodes.

Why does switching speed impact reverse recovery losses?

Switching speed can impact reverse recovery losses because turning the switch on too quickly allows the current to rise faster and further before the diode starts blocking, exacerbating reverse recovery effects.

What is soft switching, and how can it reduce switching losses?

Soft switching is a technique that aims to make voltage or current zero at the time of switching, which can reduce or eliminate switching losses. There are different strategies, such as zero voltage turn-on and zero current turn-off, that help achieve this.

What is zero voltage turn-on, and how is it achieved?

Zero voltage turn-on occurs when the voltage across a switch is already near zero before it turns on. It can be achieved using diodes in synchronous rectifiers or by using circuits with large switch node capacitance that limits the voltage rise.

What challenges exist in implementing zero current turn-off?

Zero current turn-off is challenging because it requires a topology where current naturally reaches zero before turning off the switch. This is typically seen in resonant converters, where the timing of the switch turn-off is crucial to avoid loss.