Learn Know-how

Optimization of Inverter Circuits

2025.01.29

table of contents

・Optimization of Inverter Circuits Using the ROHM Solution Simulator
・Effect of the Reverse Recovery Time trr in an Inverter Circuit
・Selection of Optimal Devices Through Loss Analyses of Inverter Circuits
・Features of Half-Bridge Inverter and Full-Bridge Inverter Circuits
・Comparison of Half-Bridge and Full-Bridge Inverter Circuit Operation

“Power Device Solution Circuits” for power devices are divided into three categories–AD-DC PFCs, DC-AC inverters, and DC-DC converters; various simulation circuits are available.

In this article, inverter circuits among the Power Device Solution Circuits are used to explain basic methods for adjusting different parameters and related know-how.

Optimization of Inverter Circuits Using the ROHM Solution Simulator

The ROHM Solution Simulator, which is a web-based simulation tool that can be used simply by registering at MyROHM, offers numerous Solution Circuits. “Power Device Solution Circuits” for power devices are classified into three categories, which are AD-DC PFCs, DC-AC inverters, and DC-DC Converters; various solution circuits are provided.

In this article, basic methods for adjusting different parameters and related know-how are introduced using inverter circuits among the Power Device Solution Circuits. Examples are presented in which simulations are used to pursue optimization with respect to three important matters requiring study in inverter circuit design: the influence of trr, selection of optimal devices based on losses, and features of half-bridge and full-bridge designs.

The “User Guide: Inverters” document on which this article is based can be downloaded from the following link:

https://fscdn.rohm.com/en/products/databook/applinote/discrete/common/web-sim-power-device-inverter_ug-e.pdf

List of Power Device Solution Circuits for Inverter Circuits

We begin by presenting a list of the Power Device Solution Circuits for inverter circuits that are currently available. Starting with general single-phase half-bridge and full-bridge circuits, in addition to 3-phase inverters and 3-level inverters, there are also circuits available for specific applications such as IH inverters and motor drives.

Category	Control No	Circuit Specifications
IH Inverter	B-1	IH Half-Bridge Inverter Po=10kW
IH Inverter	B-2	IH Full-Bridge Inverter Po=20kW
Half-Bridge Inverter	B-3	Half-Bridge Inverter Vo=200V Io=100A
Full-Bridge Inverter	B-4	Full-Bridge Inverter Vo=200V Io=100A
1-Phase 3-Wire Inverter	B-5	1-Phase 3-Wire Inverter Vo=100/200V Po=20kW
3-Phase Inverter	B-6	3-Phase 3-Wire Inverter Vo=200V Po=5kW
3-Phase Inverter	B-7	3-Phase 4-Wire Inverter Vo=115/200V Po=20kW
Motor Drive	B-8	Motor Drive 2-Phase-Modulation Po=10kW
	B-9	Motor Drive 3-Phase-Modulation Po=10kW
	B-10	Motor Drive Step-Modulation Po=5kW
3-level Inverter	B-11	3-level Inverter type-T Vo=200V Io=50A
3-level Inverter	B-12	3-level Inverter type-I Vo=200V Io=50A

Effect of the Reverse Recovery Time trr in an Inverter Circuit

In an inverter circuit, the trr (reverse recovery time) characteristic of a switching device has a large effect on losses. Here, we use a Power Device Solution Circuit of the ROHM Solution Simulator to perform simulations, to study the effect of trr in an inverter circuit.

Inverter Circuit Used in Simulations

An inverter circuit appearing in the list of Power Device Solution Circuits, the “B-6. 3-Phase 3-Wire Inverter Vo=200V Po=5kW“, is used in our example (Fig. 1). The switching devices in this inverter circuit (indicated by the yellow box) are changed and simulations are performed to determine the effect of trr.

Fig. 1: Power Device Solution Circuit inverter circuit B-6. 3-Phase 3-Wire Inverter Vo=200V Po=5kW

Importance of the trr Characteristic in an Inverter Circuit

Fig. 2 shows current paths during switching in the inverter circuit of Fig. 1.

In an inverter circuit, in order to adjust the supplied power, the high-side and low-side devices are switched on and off in alternation using PWM control, PFM control, or the like. In Fig. 2, ① to ⑤ show the circuit operation; this sequence is repeated.

Of interest here is the operation from ④ to ⑤; when the high side goes from off to on, a recovery current flows in the low-side internal diode, and so a shoot-through current, indicated in red, flows from the high side to the low side.

Fig. 2： Current paths during switching

The recovery current does not have a great effect on losses occurring in the free-wheeling device (low side) itself, but for the switching-side device (high side), as indicated in Fig. 3, the recovery current flows added onto the normal switching current prior to the VDS change, so that an extremely large turn-on loss results. Hence it is important to select devices with low trr values as the switching devices in an inverter circuit.

Fig. 3： Turn-on waveform example of a switching-side device (high side) and the relationship of switching loss to the trr value

Comparison of Switching Losses for Different trr Characteristics

Fig. 4 shows simulation results for switching losses and switching waveforms when, as the switching device in the inverter circuit of Fig. 1, a R6047KNZ4, which is a general super-junction MOSFET for switching applications, is used, and when this is replaced with a R6050JNZ4, which is a PrestoMOS™ featuring an internal diode with a particularly fast trr (see the yellow box in Fig. 1).

Fig. 4： Comparison of switching losses and waveforms for switching devices with different trr characteristics (simulated)

As the simulation waveforms indicate, prominent differences in turn-on losses appear for different trr characteristics; for the R6050JNZ4, with the faster internal diode trr characteristic, the turn-on loss is reduced to about 1/5 of the loss for the R6047KNZ4. It should be noted that the internal diode trr for the R6047KNZ4 is 700 ns (typ.), versus 120 ns (typ.) for the R6050JNZ4, which is less than 1/5 the former value.

Moreover, upon analyzing losses for the switching devices (MOSFETs) over the entire inverter circuit operation, we find that losses due to the trr value have a large effect, as shown in Fig. 5.

Fig. 5： Analysis of losses for a general switching MOSFET and a fast trr MOSFET

From these results, we can conclude that it is important to choose a switching device for an inverter circuit that has a fast internal diode trr value.

Selection of Optimal Devices Through Loss Analyses of Inverter Circuits

When studying losses in an inverter circuit, selection of the power devices to be used is also important. Optimization must be performed to minimize losses while obtaining the desired circuit operation and characteristics. Here a method is described for selecting optimal devices by dividing power device losses into switching losses and conduction losses and performing analyses.

Inverter Circuit Used in Simulations

We will use as an example “B-9. Motor Drive 3-Phase-Modulation Po=10kW” from the inverter circuits in the Power Device Solution Circuit list.is used as an example (Figure 6). The contents of the yellow boxes in this inverter circuit are modified and simulations are performed to execute loss analyses, and the optimal device is selected.

Figure 6: “B-9. Motor Drive 3-Phase-Modulation Po=10 kW” inverter circuit in Power Device Solution Circuit list

Loss Analysis Method

To explain the loss analysis method, first an example of a DC-DC converter is shown. Figure 2 shows simulated waveforms during MOSFET switching of V_DS, ID, the loss (Pd), and the energy (E) obtained by time integration of the loss. In the ROHM Solution Simulator, a calculation function of the Waveform Analyzer included in the Waveform Viewer simulation results display tool can be used to integrate losses, so that energy waveforms can easily be output.

From the energy waveforms of Figure 7, the energy consumption during the switching sections (Eon, Eoff) and the conduction section (Econd) can be seen at a glance. And by reading off differences at the cursors, numerical values can be obtained.

When the input and output are constant, as in DC-DC converters, switching losses and conduction losses can each be calculated from the product of the energy over one cycle and the switching frequency.

Figure 7: MOSFET waveforms for a DC-DC converter circuit

However, because the load of an inverter circuit fluctuates as shown in Figure 8, losses in operation of the entire circuit cannot be calculated just by looking at some of the switching waveforms.

In such circuit operation in which device losses are not constant, the loss waveform can be divided into cases, and by extracting only selected parts, losses can be divided into conduction losses and switching losses and can be calculated.

Figure 8: MOSFET waveforms for an inverter circuit

The waveforms on the left side in Figure 9, similarly to those of Figure 8, are MOSFET waveforms for the inverter circuit. The waveforms on the right side are enlargements of the portions indicated by the dashed blue line in the waveforms on the left side. The yellow waveform that is visible on the right side is the extracted conduction loss.

In the case division performed here, in order to separate only the conduction loss, the power when “V_GS is high” and “power is at or below the maximum conduction loss” is extracted. When a waveform has been extracted, the average value over one cycle is determined to analyze the loss fraction.

In the case of Figure 9, the total loss is 29.5 W and the conduction loss is 20.5 W (from the average value shown in the waveform diagram), so that the switching loss is 9.0 W. We see that the loss fractions are 70% conduction loss and 30% switching loss.

Figure 9: Extraction of conduction loss from inverter circuit MOSFET waveform (right side)

In this example, losses are divided into switching losses and conduction losses only. By setting more detailed conditions for case division, however, losses can be further subdivided into turn-on losses, turn-off losses, recovery losses, parasitic diode losses, and the like.

Consideration of Optimal Device

Figure 10 shows the MOSFET loss analysis results when the MOSFET power device in the “B-9. Motor Drive 3-Phase-Modulation Po=10kW” circuit shown in Figure 6 is changed to other devices.

Figure 10： Loss analysis results for different MOSFETs in the example inverter circuit

The third and fourth numerical digits of the MOSFET part number (“50” in the case of R6050JNZ4) indicate the current rating (ID). Thus the MOSFETs are in the same series but have ID values of, from the left, 50 A, 42 A, 30 A, and 20 A.

From the graph of Figure 10, we can confirm that the higher the current rating, the more the conduction loss falls; conversely, the lower the current rating, the more the switching loss declines. When selecting the optimal device for the circuit of Figure 1, we can determine that the R6030JNZ4, with the lowest total losses, is the optimal device.

Features of Half-Bridge Inverter and Full-Bridge Inverter Circuits

When designing circuits and verifying operation, the basic circuit configuration and features of the circuit operation must be understood. This article explains the respective features of half-bridge and full-bridge inverter circuits.

Inverter Circuits Used in Simulations

The following two circuits, appearing in the “List of Power Device Solution Circuits for Inverter Circuits”, are used.

B-3. Half-Bridge Inverter Vo=200V Io=100A　（Figure 11）
B-4. Full-Bridge Inverter Vo=200V Io=100A　（Figure 12）

As is clear from the Solution Circuit titles, B-3 is the half-bridge inverter circuit and B-4 is the full-bridge inverter circuit. The circuit diagrams appear in Figure 11 and Figure 12. The yellow boxes are places at which conditions are changed to perform comparative simulations of characteristics, described below.

Figure 11: ”B-3. Half-Bridge Inverter Vo=200V Io=100A” circuit in Power Device Solution Circuit list

Figure 12: ”B-4. Full-Bridge Inverter Vo=200V Io=100A” circuit in Power Device Solution Circuit list

Advantages and Disadvantages of Half-Bridge and Full-Bridge Inverter Circuits

Table 1 summarizes the features of half-bridge and full-bridge inverter circuits from the standpoint of advantages and disadvantages.

	Advantages	Disadvantages
Half-bridge	・Requires only two switching devices ・A single switch in a current path is advantageous for large-current applications	・Two DC voltage sources needed ・Because voltages of two power supplies are applied, switching devices must have a high withstand voltage ・Output up to only the voltage of one DC voltage source is possible
Full-bridge	・Requires only one DC voltage source ・Voltage of only one power supply applied to a switch	・Four switching devices needed ・Because there are two switches in a current path, switch conduction losses are high

Table 1: Advantages and disadvantages of half-bridge and full-bridge inverter circuits

While depending on operating conditions, it can be said that, considering the respective circuit characteristics, half-bridge circuits are better suited to low voltages and large currents, whereas full-bridge circuits are more suitable for high-voltage, high-power applications.

Comparison of Half-Bridge and Full-Bridge Inverter Circuit Operation

Figure 13 shows the results of comparison of module losses under initial conditions (Vin=500 V) in each of the circuits of Figure 11 (half-bridge) and Figure 12 (full-bridge). The output current Io is set to vary from 50 to 100 A (see the yellow boxes in the circuit diagrams). Here a module is a single half-bridge circuit, while the full bridge is basically composed of two half-bridge circuits.

Figure 13： Comparison of modules losses for half-bridge circuit and full-bridge circuits

Comparing the simulated losses, we find that the switching losses in the half-bridge circuit are large, greater than the per-module losses in the full-bridge circuit, because the voltages of two power supplies are applied to the switching devices. Upon considering the losses in the circuits as a whole, however, the conduction loss for the two modules of the full-bridge circuit is greater than the loss in the half-bridge circuit, so that losses are greater for the full-bridge circuit.

Next, a case is considered in which the half-bridge circuit has only a single V_IN =500 V power supply, and a simulation of the result when this is divided into two V_IN=250 V power supplies is performed (input voltage of the yellow box in Figure 1).

Figure 4 shows the output waveforms for the case of two V_IN =500 V power supplies and two V_IN =250 V power supplies. As the waveforms indicate, in the case of V_IN =250 V x 2, V_O hits a limit at 250 V. This reflects the disadvantage of a half-bridge circuit, indicated in Table 1, that “output up to only the voltage of one voltage source is possible”. The effective peak voltage is 282 V, and so V_IN =250 V x2 is insufficient to provide the output voltage setting of V_O =200 V.

Figure 14: Comparison of output waveforms from a half-bridge circuit for the cases of V_IN =500 V x 2 and V_IN =250 V x 2

In this way, half-bridge inverter circuits and full-bridge inverter circuits have their respective advantages and drawbacks, and a general statement as to which is superior is not possible. The features of each should be understood so that they can be used selectively according to the application.

Learn Know-how

Electrical Circuit Design

Thermal design

Switching Noise

Simulation