Considerations for DC motor drive design, simplified gate drive and edge delay drive circuit diagrams and design ideas for low-voltage drive circuits.
I. Design Goals of DC Motor Drive Circuit
In the design of DC motor drive circuits, the following points should be considered:
1. Function: Does the motor rotate in one or two directions? Is speed control required? For a unidirectional motor drive, a high-power transistor, MOSFET, or relay can directly drive the motor. When the motor needs to rotate in both directions, an H-bridge circuit consisting of four power components or a double-pole double-throw relay can be used. If speed control is not required, a relay is sufficient; however, if speed control is required, switching components such as transistors and MOSFETs can be used to implement PWM (Pulse Width Modulation) speed control.
2. Performance: The main performance indicators for PWM speed control motor drive circuits are as follows.
1) Output current and voltage range, which determines the power of the motor that the circuit can drive.
2) Efficiency: High efficiency not only means saving power but also reducing heat generation in the drive circuit. To improve circuit efficiency, one can start by ensuring the switching operation of power devices and preventing common-mode conduction (a problem that may occur in H-bridge or push-pull circuits, where two power devices conduct simultaneously, causing a short circuit in the power supply).
3) Impact on control input. The power circuit should have good signal isolation at its input to prevent high voltage and high current from entering the main control circuit. This can be achieved using high input impedance or optocouplers.
4) Impact on power supply. Common-mode conduction can cause a momentary drop in power supply voltage, resulting in high-frequency power supply pollution; large currents may cause ground potential fluctuations.
5) Reliability. The motor drive circuit should be designed to be as safe as possible, regardless of the control signal or passive load applied.
Different parts of a motor have different design requirements.
1. Input and level conversion section:
The input signal lines are introduced through the DATA pin; pin 1 is ground, and the rest are signal lines. Note that a 2K ohm resistor is connected to ground at pin 1. When the driver board and the microcontroller are powered separately, this resistor provides a path for the signal current to return. When the driver board and the microcontroller share a power supply, this resistor prevents large currents from flowing along the connection lines into the microcontroller's motherboard ground, causing interference. In other words, it effectively isolates the driver board's ground from the microcontroller's ground, achieving "single-point grounding."
The high-speed operational amplifier KF347 (or TL084) functions as a comparator, comparing the input logic signal with a 2.7V reference voltage from an indicator light and a diode, converting it into a square wave signal with an amplitude close to the power supply voltage. The KF347's input voltage range cannot approach the negative supply voltage, otherwise errors will occur. Therefore, a diode is added at the op-amp input to prevent voltage range overflow. Two resistors are used at the input: one for current limiting and the other to pull the input low when it is floating.
The LM339 or any other open-output comparator cannot be used to replace the op-amp because the output impedance of the high-level state of the open-output is more than 1,000 ohms, resulting in a large voltage drop, which will prevent the transistor in the next stage from being cut off.
2. Gate driving section:
The circuit consisting of transistors, resistors, and Zener diodes further amplifies the signal, drives the gate of the field-effect transistor, and uses the gate capacitance of the field-effect transistor itself (approximately 1000pF) for delay to prevent the field-effect transistors in the upper and lower arms of the H-bridge from conducting simultaneously ("common mode conduction"), which would cause a short circuit in the power supply.
When the op-amp output is low (approximately 1V to 2V, not reaching zero), the lower transistor is off and the field-effect transistor (FET) is on. The upper transistor is on, the FET is off, and the output is high. When the op-amp output is high (approximately VCC - (1V to 2V, not reaching VCC), the lower transistor is on and the FET is off. The upper transistor is off, the FET is on, and the output is low.
The above analysis is static; the following discussion focuses on the dynamic process of switching: The on-resistance of the transistor is much less than 2 kΩ, therefore, when the transistor switches from cutoff to on, the charge on the gate capacitance of the MOSFET can be released rapidly, and the MOSFET quickly cuts off. However, when the transistor switches from on to cutoff, the charging of the MOSFET gate through the 2 kΩ resistor takes a certain amount of time. Correspondingly, the speed at which the MOSFET switches from on to cutoff is faster than the speed at which it switches from cutoff to on. If the switching actions of the two transistors occur simultaneously, this circuit can allow the MOSFETs in the upper and lower arms to be turned off first and then on, eliminating the common-mode conduction phenomenon.
In reality, the op-amp output voltage requires a certain amount of time to change, during which time the output voltage is at the midpoint between the positive and negative supply voltages. During this time, both transistors conduct simultaneously, while the field-effect transistors (FETs) are simultaneously cut off. Therefore, the actual circuit is even safer than this ideal situation.
The 12V Zener diode at the gate of a field-effect transistor (FET) is used to prevent the FET gate from being overvoltage-damped. A typical FET gate has a withstand voltage of 18V or 20V; applying a 24V voltage directly will cause it to break down. Therefore, this Zener diode cannot be replaced by a regular diode, but it can be replaced by a 2kΩ resistor, which will still provide a 12V voltage divider.
3. Field-effect transistor output section:
High-power MOSFETs have internal diodes connected in reverse parallel between the source and drain. When configured as an H-bridge, this is equivalent to having four diodes connected in parallel at the output to eliminate voltage spikes, thus eliminating the need for external diodes. Connecting a small capacitor in parallel at the output (between out1 and out2) can help reduce voltage spikes generated by the motor, but it can also cause current spikes when using PWM, so the capacitance should not be too large. This capacitor can be omitted when using low-power motors. If this capacitor is added, it must be a high-voltage capacitor; ordinary ceramic capacitors may break down and short-circuit.
The circuit consisting of a resistor, a light-emitting diode, and a capacitor connected in parallel at the output terminal indicates the direction of motor rotation.
4. Performance Indicators:
The power supply voltage is 15~30V, the maximum continuous output current is 5A per motor, and it can reach 10A for a short time (10 seconds). The PWM frequency can be used up to 30KHz (generally 1 to 10KHz is used). The circuit board contains four logically independent power amplifier units, with their output terminals connected in pairs to form an H-bridge, which can be directly controlled by a microcontroller. This enables bidirectional rotation and speed regulation of the motor.
5. Wiring:
High-current circuits should be as short and thick as possible, and should avoid vias whenever possible. If vias are unavoidable, they should be made larger (>1mm), and a small ring of vias should be made around the pads, filling the gaps with solder during soldering; otherwise, they may burn out. Additionally, if a Zener diode is used, the source-to-power and ground traces of the MOSFET should be as short and thick as possible. Otherwise, under high current, the voltage drop across this trace may burn out the forward-biased Zener diode and the conducting transistor. In the initial design, a 0.15-ohm resistor was connected between the source and ground of the NMOS transistor to detect current; this resistor became the culprit for repeatedly burning out the board. Of course, replacing the Zener diode with a resistor eliminates this problem. In the 2004 Robocon competition, we primarily used this circuit for motor driving.
II. Simple Gate Drive in Low-Voltage Drive Circuits
The maximum gate-source voltage of a typical power MOSFET is around 20V. Therefore, in 24V applications, the gate-source voltage must not exceed 20V, increasing the complexity of the circuit. However, in 12V or lower voltage applications, the circuit can be greatly simplified.
The left diagram shows one side of a 12V driver bridge. The transistor section of the circuit above has been replaced by two diodes and two resistors. (Note that the logic is reversed compared to the previous diagram.) Due to the presence of the MOSFET's gate capacitance, charging the gate capacitance through R3 and R4 delays the MOSFET's turn-on; while directly discharging the gate capacitance through the diodes immediately turns off the MOSFET, thus avoiding common-mode conduction.
This circuit requires a square wave pulse with steep edges to be input at the IN terminal. Therefore, the control signal, after being input from the microcontroller or other open-output devices, must pass through a Schmitt trigger (such as a 555) or a push-pull output high-speed comparator before being connected to the IN terminal. If the input edge is too gentle, the diode delay circuit will become ineffective.
The selection of R3 and R4 is related to the rise and fall speed of the IN signal edge. The steeper the signal edge, the smaller R3 and R4 can be, and the faster the switching speed can be. In the boost circuit used in the Robocon competition (with a similar principle), a 555 timer is used before IN.
III. Edge Delay Drive Circuit
In the front-end logic circuit, intentionally delaying the falling edge controlling the PMOS and the rising edge controlling the NMOS before shaping them into square waves can avoid common-mode conduction of the MOSFETs. Furthermore, this simplifies the subsequent gate drive circuit, allowing for low-impedance push-pull gate drive without considering gate capacitance, thus better adapting to different MOSFETs. This drive circuit was used in the 2003 Robocon competition. The following diagram shows two edge delay circuits:
The following diagram shows the corresponding NMOS and PMOS gate drive circuits:
This gate drive circuit consists of two stages of transistors: the first stage provides the correct voltage required to drive the gate of the field-effect transistor, and the second stage is an emitter follower to reduce the output impedance and eliminate the effect of the gate capacitance. To ensure non-common-mode conduction, the input edge needs to be relatively steep, which can be achieved by the aforementioned delay-and-shape circuit.
IV. Other driving circuits
(The idea of relays + semiconductor power devices)
Relays offer advantages such as high current handling and stable operation, greatly simplifying drive circuit design. They can also be fully utilized in motor drive circuits requiring speed control. One approach is to use a relay to control the current direction to change the motor's direction, while using a single high-current MOSFET (such as the IRF3205, typically only available in N-type high-current configurations) to implement PWM speed control, as shown in the right figure below. This is a method for achieving very high current drive. The commutation relay should be a double-pole double-throw type, with wiring shown in the left figure below and coil wiring in the middle figure below.
