Brushless Motors: Waveforms in Winding Terminal Off Intervals, Freewheel Diodes, Power Supply Currents

Next up is my third talk, this time about “Actual Signal Waveforms in Brushless Motor Driving”. There are two things to know about these waveforms; the first of these is the waveforms in winding terminal off intervals.

Actual Signal Waveforms in Brushless Motor Driving

• ① Why is the U phase current waveform shaped this way?
  ▶Winding Currents and Induced Voltages
• ② I understand the UH and UL signals and waveforms, but when I zoom in, the UH signal has a pulsed shape.
  ▶Applied voltage pulses
• ③ Where the UH waveform is pulsed, the U phase voltage (the voltage across the U phase winding terminals) has a pulsed waveform, like the UH waveform, and where the UL signal is turned on, I understand why it is at GND level. But what is the diagonal waveform of the U phase voltage where UH and UL are both turned off?
  ▶Waveforms of Winding Terminal Off Intervals
  ▶Freewheel Diodes
• ④ What does the power supply current represent? This current also has a pulsed waveform.
  ▶Power Supply Currents

The above phenomena and waveforms are described in Actual Signal Waveforms in Brushless Motor Driving.

Brushless Motors: Waveforms in Winding Terminal Off Intervals

A “waveform in a winding terminal off interval” is a waveform that appears at winding terminals when both the upper and the lower transistors for the phase are turned off. This state is called an open or Hi-Z (high impedance) state; here I’ll call it “open”.

At the timing indicated by the broken line in the figure below on the left, the U phase is open, the V phase has a PWM voltage applied, and the W phase is fixed at low. At this time, the V phase and W phase winding terminal voltages are electrically fixed at the power supply voltage or at GND by the switches (transistors), resulting in the waveforms shown on the right.

The U phase is in the open state, and so a waveform appears with the neutral point in the figure as reference. The neutral point has the voltage division value of the V phase windings voltage and the W phase windings voltage. The Z∙I (voltage) values of the V and W phases are approximately the same, but because the induced voltages of the two phases are different, the neutral point voltage has a pulse waveform with a slope, as shown in the figure. Moreover, the U phase induced voltage is superposed on this neutral point voltage, so that the U phase terminal voltage has the waveform shown. This is the waveform that appears across the winding terminals of the phase in 120 degree conduction for which the upper and lower transistors are both turned off (open state).

In motor control, there is some value in using such phenomena in which a waveform that contains an induced voltage can be seen. As I mentioned before, the winding terminal voltage in an off interval (in the open state) is a waveform in which an induced voltage has been added to the neutral point voltage, and so by measuring the winding terminal voltage with the neutral point as reference, the induced voltage for the phase can be detected.

In the previous “The Principle of Induced Voltage Occurrence in a Brushless Motor”, I explained that an induced voltage is determined by the coil position and the relative position of the rotor (permanent magnet). This means that, conversely, one can use the induced voltage to infer the rotor position.

Methods in which this principle is used to find the rotor position without using Hall elements or other detectors are called senseless control methods; they are in wide use (I previously explained that detection of the rotor position is important for motor control). Incidentally, sensors in brushless motors are either magnetic pole detectors such as Hall elements or else rotor position detectors such as encoders; sensors are also needed to detect induced voltages.

However, there are places in the waveforms shown that are cause for concern. The U phase terminal voltage waveform in which an induced voltage appears does not fall below a certain value; this seems unnatural. I’ll explain the reason for this.

Brushless Motors: Freewheel Diodes

When we use an oscilloscope or some other instrument to check a winding terminal voltage, we find a waveform like that shown below in which the voltage falls a little below GND level, but does not fall any further. Put briefly, this is because there is a special diode, called a freewheel diode, in parallel with the power transistors (see the circuit diagram at the bottom of the following diagram).

Because of the presence of these diodes, even when the potential across the winding terminals tries to fall below GND potential by the amount of the diode forward-direction voltage (Vf), the diode turns on, and so the voltage does not fall further. For similar reasons, the potential does not rise above the power supply voltage plus Vf.

I’ll note in passing that a freewheel diode is not a diode with certain special characteristics like a Zener diode or a light-emitting diode, but is instead a term used for a diode the purpose of which is to return or circulate current (there is probably no product with that name). Freewheel diodes are also called flyback diodes or feedback diodes.

The purpose of these diodes is to protect circuit devices when a transistor is turned off. The above diagram shows a current that is flowing when a transistor through which the current had been flowing is turned off. In a motor, currents flow through coils, and so if a transistor is turned off in order to stop the flow of a current, a voltage appears across the coil (the equation in the figure indicates the voltage that occurs when the current changes). If this voltage is too large, transistor failure can occur.

In order to prevent such failure, a diode is installed parallel to the transistor. When a diode is present, if a voltage lower than GND-Vf or higher than the power supply voltage +Vf occurs, current flows through the diode in a “clamping” action (which does not allow very low or very high voltages), so that the transistor can be protected.

This phenomenon similarly occurs even for upper-side transistors, indicated by PWM in the figure above. For the upper side, a lower-side diode becomes a current path to provide protection.
In motor control circuits, a phenomenon often occurs in which the voltage across winding terminals that are supposed to be in the open state are held close to the power supply voltage or to GND level (are “clamped”) by a flowing current. Keep in mind that some control designs make active use of this phenomenon.

As freewheel diodes, discrete diodes may be placed in parallel with each power transistor; but because in a MOSFET structure there is already an internal parasitic diode (called a “body diode”), if the characteristics of the parasitic diode are adequate for it to serve as a freewheel diode, it is generally used instead. To aid my explanation, I have shown you a circuit diagram that has no freewheel diodes, but in fact there are no MOSFET transistors that do not have freewheel diodes.

Next, I’ll talk about the fourth and last topic, on “power supply currents”, of this series on “Actual Signal Waveforms in Brushless Motor Driving”.

Brushless Motors: Power Supply Currents

One signal that should be checked when a motor is running is the power supply current. The power supply current is, as its name implies, the current that is flowing in the power supply. I explained earlier that when we think about the characteristics of a motor, it’s important that we understand the waveform of the current flowing in the windings for each phase, but it’s also important to know about the power supply current. The power supply current I’m talking about here is the current that flows from the power supply through the power transistors and the coils and then returns to the power supply, as shown in the figure. This does not include the controller IC current or the transistor gate driving currents.

Let me start by using the figure to explain what the power supply current is. The current flowing in the power supply changes depending on the on/off states of the power transistors. I’ll explain the waveform sections labeled a through e in the figure.

• a： This current flows from the U phase upper-side transistor through the windings, returning to the power supply from the W phase lower side. So the same waveform as that of the U phase winding current appears in the power supply current.
• b： When the upper-side transistor turns off, this current flows through the U phase lower-side freewheel diode. At this time, there is no longer a current heading toward the power supply.
• c： When the U phase PWM interval ends, the current that had been flowing in the U phase gradually diminishes as it flows through the diode to reach zero, so that this current does not appear in the power supply current.
• d： As with a, there is a current that flows from the V phase to the W phase and then to the power supply.
• e： Behavior similar to that of b.

So we see that the power supply current changes according to the on/off logic of the power transistors, and its waveform has a pulse shape (the same current amounts as the winding currents are flowing, or are not flowing).

An understanding of the power supply current is important because motor control uses information included in the power supply current relating to the winding currents. In motor control, a grasp of the winding current values is important, but if we want to detect the winding currents directly, we will need current sensors, isolation amplifiers, and other circuitry. But when using the power supply circuit, there is the advantage that we can detect winding currents by placing what is called a shunt resistor at positions such as that shown below and measuring the voltage across the resistor, with the potential on one side at GND level.

Detected values can be used for the following.

• ・Current limiting (keeping currents from becoming too large)
• ・Various kinds of control (inferring an induced voltage from a winding current value, detecting a current phase, etc.)

Information contained in power supply currents can be exploited for motor control.

Up to here, I have explained five aspects of the behavior of driving signal waveforms of actual motors that do not appear in timing charts and can’t be understood based only on previous explanations.

Next time, I will explain the output characteristics of brushless motors.