[Episode 10] Beyond the Questions! What Engineer Ichinose Learns

Summary of the Previous Episodes

Last time, Ichinose learned about noise and reliability with the help of Tako and Dora.

Ichinose and Ninomiya were relieved to learn that even the super engineer, Sugiken, once had a lot he didn’t understand. This made them determined to absorb even more knowledge.

Character Introduction

• Ichinose (the protagonist) is a new engineer. He has been aiming to become an engineer since he was in middle school, and finally joined ROHM. He is so passionate that he forgets to eat and sleep while studying on motor drivers. Currently, only Ichinose can see Dora and Tako.
• Ninomiya is in the same year as Ichinose. Her grades are always at the top. She has a strong personality, but she is also a hard worker and has a high opinion of Ichinose. She is in secret a big fan of Sugiken.
• Teacher Sugiken is a super engineer at ROHM. He is usually kind, but he is passionate and takes pride in his work as an engineer. Sugiken used to be able to see Dora and Tako, but now he can’t.
• Dora is a motor driver fairy who loves people who are passionate about motors. He has a crush on Tako, but is always at the mercy of the insensitive Tako.
• Tako is a motor fairy and childhood friend of Dora. She is knowledgeable about motors, and her knowledge surpasses that of Dora. Although she is a reliable older sister, she is insensitive when it comes to love, and is unaware of Dora’s feelings.

Configuration of Blushless Motor Drivers

In the first through the ninth episodes of our Motor Driver Dojo, I explained the principles of motor rotation, the basic operation of motor drivers, and motor mechanism characteristics. I also discussed the performance demanded from motors, such as efficiency, quietness, and reliability. This is all basic knowledge that should be understood by anyone who sets out to develop an efficient, quiet, highly reliable motor.

In this tenth episode, I will explain the configuration and internal circuitry of motor drivers that are designed to make such motors possible.

Contents of Episode 10

• ・Motor Driver Configuration
• ・Controllers
• ・Power Transistors
• ・Level Shifters

Motor Driver Configuration

Circuitry to drive a brushless motor can be classified in terms of function into three blocks. These are a driving controller block that generates command signals to rotate the motor, a power transistor block that supplies power to the motor windings, and a level shifter block that assists in the exchange of signals (mainly by adjusting potential differences) between the other two functional blocks.

These functional blocks are packaged in ICs, which are provided with a single function or with multiple functions. When designing a motor driving circuit, these ICs can be combined as needed to obtain a circuit configuration like that shown above.

The ICs to combine to obtain a motor driver must be studied from the standpoints of ease of use, freedom to modify designs, package sizes, the number of traces on the board, the number of peripheral components, differences in withstand voltages between circuit blocks, and increases in temperature (heat dissipation properties). The optimum configuration must be selected in order to satisfy the most important requirements for the motor.

I’ll explain a number of configuration examples.

A full-function IC incorporates all functional blocks, and features ease of use. However, package sizes tend to be large, and there is the added limitation of reduced freedom of design.

In configurations in which single-function ICs are combined by selecting optimal functional ICs for use, motor driver specifications suited to the application can be obtained. When the driving control method is to be changed, all you have to do is simply modify the driving controller ICs, and also change the power transistor ICs and level shifter ICs if you need a different output, so there is a lot of design flexibility. There is also the advantage of being able to adjust the power transistor switching speed. However, this may result in an increase in the number of IC peripheral components and traces on the PCB compared with a full-function IC.

An IC provided with two different functional blocks can be combined with an IC having the other functional block. An IC with a driving controller and a level shifter gives you the freedom to choose power transistors to use with it, and can reduce the number of traces on the board, but it doesn’t let you use fewer components connected to the power transistors. A level shifter plus power transistor IC is as easy to use as a full-function IC, and lets you pick the driving controller to use with it. But the power transistor switching speed can’t be adjusted, and there is less design freedom (but there are also ICs that have internal circuits for speed adjustment).

So next I’ll talk about each of these functional blocks.

Controllers

Controller ICs generally have the internal structure shown below. Chips are connected by leads and wires, and both analog circuits and logic circuits are formed on the chip. The analog circuits exchange signals with devices outside the IC, and generate a power supply and a clock signal, and the logic circuits generate motor driving signals.

Next, I’ll talk about the controller IC leads (pins) and their relation to the analog circuits and the digital circuit blocks. I’ll use as an example a controller IC that follows Hall signal input specs.

The main pin functions and analog circuits are shown below.

Next, let’s look at the logic circuit. The logic circuit generates PWM pulses and FG signals based on the rotor position signal and duty commands.

A duty command for input to a controller conforms to specifications like those below. These signals are converted into digital signals by a duty command processing block. But I’ll mention that there are controllers that use the duty signal square waves as PWM pulses without digitizing the duty commands.

The output FG signal is a pulse signal that changes the frequency according to the motor rotation rate; the specs for this signal are shown below.

Power Transistors

The transistors used in a power transistor block may be bipolar transistors, MOSFETs, or IGBTs. Here, after comparing these three types of devices, I’ll explain MOSFET on/off operation.

Bipolar transistors are current-driven type of transistor that are put into the on state by passing a current in the base. The main losses in these transistors are the power that can be calculated by multiplying the VCE voltage by the collector current. However, a base current must be passed continuously in order to maintain the on state, and the power consumed by doing so is another drawback. And, because the switching speeds of these devices are slow, they are not suitable for high-frequency PWM driving.

MOSFETs are voltage-driven transistors that enter the on state when a voltage is applied to the gate. The main loss is power that is equal to the on-resistance (Ron) multiplied by the square of the drain current. Among their advantages are the fast switching speed and freedom from the need to continuously pass a current to the gate. But they have the drawback that losses increase exponentially as the drain current increases.

IGBTs are transistors that combine advantages of bipolar transistors and MOSFETs. They are voltage-driven so that there is not much gate current consumption, and on-state losses increase only as the first power of the collector current. Switching speeds are intermediate.

Losses are one criterion to use when deciding which transistor type to use. As I just explained, MOSFET on-resistance losses vary as the square of the current, so whether losses are greater when using MOSFETs or IGBTs may well depend on the size of the current. Often the point at which Ron times the current for a MOSFET is equal to VCE for an IGBT is taken to be a reference point: if the motor current is smaller than this value, MOSFETs are used, and for motors with a larger current than this, IGBTs are preferred.

Next, I’ll explain the on/off operation of an N-channel MOSFET. Let’s assume that the winding current is already flowing through the high-side transistor (the other transistor, which we’re not considering here), as in the diagram below on the left.

The state of a MOSFET changes according to the voltage at the gate (see the diagram below). When the gate voltage of a low-side transistor rises to exceed a threshold value relative to the source potential, current flowing in the windings gradually passes through the transistor and flows out. When all of the current that had been flowing in the windings passes through the low-side transistor, the D-S voltage falls, and the transistor enters the on state. Off-state operation is the opposite of this.

The waveforms in the diagrams below show the gate voltage G, drain current Id, and drain-source voltage Vds, with time plotted on the horizontal axis. You see that as the gate voltage changes, Id and Vds change gradually. The speed with which Id and Vds change (the switching speed) depends on the amount of change in the gate voltage. Because of this, if you wanted to adjust the switching speed, you could connect a gate resistor Rg and a gate capacitor Cg as shown below to adjust the gate voltage change. This switching speed is related to the switching loss in my eighth episode and to the electromagnetic noise in my ninth episode, and also to the dead time in my seventh episode. The faster the switching speed, the smaller the losses and the shorter the dead time, but electromagnetic noise increases; if the switching speed is slower, these are reversed. And so the switching speed is an essential parameter when designing a power transistor block.

Level Shifters

Each level shifter circuit unit in a level shifter block converts a gate command signal from the controller into a potential and current that can turn a power transistor on and off. As I explained earlier, MOSFETs are used as the power transistors, and they are turned on and off by changing the voltage at the gate terminal. This operation requires an appropriate voltage and charge/discharge current, and in general, the controller output circuit doesn’t have the power to supply an adequate potential and current. That’s why level shifter circuits are used to adjust the supplied levels.

Level shifter circuits have either of two kinds of configurations, according to the type (P channel or N channel) of the connected high-side power transistors.

Regardless of the high-side configuration, the low-side power transistors are N channel devices, and so the processing performed by level shifter circuits for the low side is the same in either case. That is, a signal received from the controller is converted into a signal with the voltage amplitude required to turn a power transistor on or off, and an adequate current is supplied.

The circuitry for high-side transistors is very different depending on whether the power transistors are PMOS or NMOS. If they are PMOS, transistors turn off when the gate signal output from the level shifter circuit is the same potential as the VM voltage, and turn on when the potential is below the threshold value. In the circuit shown on the upper left, when the transistor connected below the PMOS device gate turns on, the gate potential drops and the PMOS transistor turns on, and when the transistor below the gate turns off, the resistance causes the gate voltage to be raised to the VM potential, so that the PMOS transistor turns off.

This level shifter circuit does not have to create a power supply voltage above the VM voltage, and so it has the advantage of a relatively simple configuration. However, the level shifter circuit is often expensive compared with the case of NMOS power transistors, and the fact that the transistors on the high and the low sides have different characteristics complicates things.

When the high-side transistors are NMOS, a power supply voltage somewhat higher than VM is needed to turn on the high-side power transistors. This voltage is normally generated within the motor circuitry using voltage-boosting circuits such as charge pump or bootstrap circuits. Because the circuit connected to a high-side NMOS gate is a floating circuit that takes the winding voltage as reference, there has to be a circuit that transmits the HIN logic to the floating circuit, as shown in the circuit configuration on the upper right. In this way, the circuit configuration is rather complex when the high-side transistors are NMOS devices, but because both the high- and the low-side power transistors are NMOS, there is the advantage that they have the same characteristics.