PCB Design Goal of DC Motor Drive Circuit
In the design of DC motor drive circuit, the following points should be considered:
1. Function: Does the motor rotate unidirectionally or bidirectionally? Don’t need speed regulation? For unidirectional motor drive, only a high-power triode or FET or relay can directly drive the motor. When the motor needs to rotate in both directions, an H-bridge circuit composed of four power elements or a double-pole double-throw relay can be used. If you don’t need speed regulation, just use a relay; However, if speed regulation is required, PWM (pulse width modulation) speed regulation can be realized by using switching elements such as triodes and field effect transistors.
2. Performance: For the motor drive circuit with PWM speed regulation, there are mainly the following performance indicators.
1) Output current and voltage range, which determines how much power the circuit can drive the motor.
2) Efficiency. High efficiency not only means saving power, but also reduces the heating of the driving circuit. To improve the efficiency of the circuit, we can start with ensuring the switching state of the power devices and preventing the common-mode conduction (a possible problem of H-bridge or push-pull circuit, that is, the power supply is short-circuited when two power devices are turned on at the same time).
3) Influence on control input. The input of the power circuit should be well isolated to prevent high voltage and current from entering the main control circuit, which can be isolated by high input impedance or photoelectric coupler.
4) Influence on power supply. Common-state conduction can cause the instantaneous drop of power supply voltage and cause high-frequency power supply pollution; Large current may cause the ground potential to float.
5) Reliability. The motor drive circuit should be done as much as possible, and the circuit is safe no matter what kind of control signal or passive load is applied.
1. Input and level conversion part:
The input signal line is led in by DATA, one pin is ground, and the rest are signal lines. Note that one pin is connected with a 2K ohm resistor to the ground. When the driving board and the single chip microcomputer are powered separately, this resistor can provide a path for the signal current to flow back. When the driving board shares a set of power supply with the single chip microcomputer, this resistor can prevent the large current from flowing into the ground wire of the single chip microcomputer motherboard along the connecting line and causing interference. In other words, it is equivalent to isolating the ground wire of the driving board from the ground wire of the single chip microcomputer, so as to realize “one point grounding”.
The high-speed operational amplifier KF347 (TL084 can also be used) is used as a comparator, which compares the input logic signal with the 2.7V reference voltage from the indicator lamp and a diode, and converts it into a square wave signal close to the voltage amplitude of the power supply. The input voltage range of KF347 cannot be close to the negative power supply voltage, otherwise an error will occur. Therefore, a diode is added at the input end of the operational amplifier to prevent the voltage range from overflowing. The two resistors at the input terminal are used to limit the current, and the other is used to pull the input terminal to a low level when the input is floating.
LM339 or any other comparator with open-circuit output can’t be used to replace the operational amplifier, because the high-level output impedance of the open-circuit output is above 1 kω, and the voltage drop is large, so the triode of the next stage will not be cut off.
2. Gate driving part:
The circuit composed of the rear triode, resistor and voltage regulator further amplifies the signal, drives the gate of the FET and delays by using the gate capacitance (about 1000pF) of the FET itself, so as to prevent the FET of the upper and lower arms of the H bridge from conducting at the same time (“common-state conduction”), resulting in power supply short circuit.
When the output of the operational amplifier is at a low level (about 1V to 2V, which can’t completely reach zero), the lower triode is turned off and the FET is turned on. The upper triode is turned on, the FET is turned off, and the output is high level. When the output of the operational amplifier is at a high level (about VCC-(1V to 2V), which can’t fully reach VCC), the lower triode is turned on and the FET is turned off. The upper triode is turned off, the FET is turned on, and the output is low.
The above analysis is static, and the dynamic process of switching is discussed below: the on-resistance of the triode is much less than 2 kω, so when the triode is switched from off to on, the charge on the gate capacitor of the FET can be quickly released, and the FET is quickly turned off. However, when the triode is switched from on to off, it takes a certain time for the gate of the FET to be charged through the 2 kω resistor. Accordingly, the speed of field effect transistor from conduction to cutoff is faster than that from cutoff to conduction. If the switching actions of the two triodes occur at the same time, this circuit can make the upper and lower arms of the field effect transistors break first and then turn on, thus eliminating the common-mode conduction phenomenon.
Actually, it takes a certain time for the output voltage of the operational amplifier to change, during which the output voltage of the operational amplifier is in the middle between the positive and negative power supply voltages. At this time, the two triodes are turned on at the same time, and the FET is turned off at the same time. So the actual circuit is safer than this ideal situation.
12V zener diode of FET grid is used to prevent overvoltage breakdown of FET grid. Generally, the withstand voltage of FET grid is 18V or 20V. If 24V voltage is applied directly, it will break down. Therefore, this zener diode can not be replaced by ordinary diode, but it can be replaced by a 2 kω resistor, and it can also get a 12V voltage division.
3. FET output part:
Inside the high-power FET, there are diodes in reverse parallel between the source and the drain. When connected into an H-bridge, it is equivalent to the fact that four diodes for eliminating voltage spikes have been connected in parallel at the output end, so there is no external diode here. Connecting a small capacitor (between out1 and out2) in parallel at the output is beneficial to reduce the peak voltage generated by the motor, but it has the side effect of peak current when PWM is used, so the capacity should not be too large. This capacitor can be omitted when low-power motors are used. If this capacitor is added, it must be high withstand voltage. The common ceramic capacitor may have breakdown and short circuit.
A circuit composed of resistor, LED and capacitor with parallel output terminals indicates the rotation direction of the motor.
4. Performance indicators:
The power supply voltage is 15~30V, the maximum continuous output current is 5A/ each motor, which can reach 10A in a short time (10 seconds), and the PWM frequency can be up to 30KHz (generally 1 ~ 10KHz). The circuit board contains four logically independent power amplification units, whose outputs are connected in pairs to form an H bridge, which can be directly controlled by a single chip microcomputer. Realize the bidirectional rotation and speed regulation of the motor.
5. Layout and wiring of 5.PCB design outsourcing painting:
The high-current line should be as short and thick as possible, and try to avoid passing through the via hole. If it must pass through the via hole, make the via hole bigger (> 1mm) and make a circle of small vias on the pad, and fill it with solder during soldering, otherwise it may burn out. In addition, if a voltage stabilizing tube is used, the source of the FET should be as short and thick as possible to the wires between the power supply and the ground, otherwise, the voltage drop on this wire may be burned down by the positive voltage stabilizing tube and the conducting triode under high current. In the initial design, a 0.15 ohm resistor was connected between the source of NMOS tube and the ground to detect the current, and this resistor became the chief culprit of constantly burning the board. Of course, if the voltage regulator is replaced by a resistor, this problem will not exist.
The PCB of motor drive circuit needs special cooling technology to solve the problem of power consumption. The substrate of printed circuit board (PCB), such as FR-4 epoxy resin glass, has poor thermal conductivity. On the contrary, copper has excellent thermal conductivity. Therefore, from the point of view of thermal management, increasing the copper area in PCB is an ideal solution. Thick copper foil (for example, 2 ounces (68 microns thick)) has better thermal conductivity than thinner copper foil. However, the cost of using thick copper foil is high, and it is difficult to realize fine geometry. Therefore, it is common to use 1 ounce (34 micron) copper foil. The outer layer usually uses an ounce to one ounce of copper foil. The solid copper surface used in the inner layer of multilayer circuit board has good heat dissipation. However, because these copper surfaces are usually placed in the center of the circuit board stack, heat will gather inside the circuit board. Increasing the copper area of the outer layer of PCB and connecting or “sewing” it to the inner layer through many through holes helps to transfer heat to the outside of the inner layer.
Heat dissipation of double-layer PCB may be more difficult due to the presence of traces and components. Therefore, it is necessary to provide as many solid copper surfaces as possible and achieve a good thermal connection with the motor driver IC. Adding copper-clad areas on both outer layers and connecting them with many through holes helps to dissipate heat among the areas divided by traces and components.
A, wiring width: the wider the better.
Due to the large current in and out of the motor driver IC (more than 10A in some cases), the PCB trace width in and out of the device should be carefully considered. The wider the line, the lower the resistance. The wiring size must be adjusted so that the wiring resistance will not consume too much power, and the wiring temperature will not rise. Too small a trace can actually be used as an electric fuse, and it is easy to burn out!
Designers usually use IPC-2221 standard to determine the appropriate trace width. This specification provides corresponding charts showing the cross-sectional area of copper for various current levels and allowable temperature rise, which can be converted into the trace width under the condition of a given copper layer thickness. For example, the trace carrying 10A current in 1 oz copper layer needs to be slightly wider than 7mm to achieve a temperature rise of 10℃. For 1-A current, the trace width only needs to be 0.3mm.
In view of this, it seems impossible for 10A current to pass through the micro IC board.
It should be understood that the suggested trace width in IPC-2221 is suitable for long-distance PCB traces with equal width. If a shorter PCB trace is used, it is possible to pass a much larger current without any adverse effects. This is because the short and narrow PCB wiring resistance is small, and any heat generated will be absorbed into the wider copper area, which acts as a heat sink.
Any heat generated by the narrower part of the trace is conducted to the wider copper area, so that the temperature rise of the narrower trace can be neglected.
The traces embedded in the inner layer of PCB can’t dissipate heat as well as the traces on the outer layer, because the insulation substrate has poor thermal conductivity. Therefore, the inner wiring should be designed to be about twice as wide as the outer wiring.
As a general guideline, the following table shows the recommended trace width for longer traces (more than about 2cm) in motor driver applications.
If space permits, the temperature rise and voltage drop can be minimized by using wider wiring or copper-clad wiring.
B. Hot through holes: use as many as possible.
A via is a small plated hole, which is usually used to pass a trace from one layer to another. Although thermal vias are made in the same way, they are used to transfer heat from one layer to another. Proper use of thermal vias is very important for PCB heat dissipation, but several technological issues must be considered.
The vias have thermal resistance, which means that when heat flows through the vias, there will be some temperature drop between the vias, and the measurement unit is℃/w. In order to minimize this thermal resistance and improve the efficiency of heat transfer through holes, large through holes should be used, and the holes should contain as much copper area as possible.
Although through holes can be used in the open area of PCB, through holes are often placed in the area of IC board to directly transfer heat from IC package. In this case, large through holes cannot be used. This is because the large plated through hole may lead to “tin penetration”, that is, the solder used to connect IC and PCB flows down into the through hole, resulting in poor quality of solder joints.
There are several ways to reduce tin penetration. One of them is to use a very small through hole to reduce the amount of solder penetrating into the hole. However, the thermal resistance of small vias is higher, so more vias are needed to achieve the same thermal performance.
Another technique is to “pitch a tent” for the through hole on the back of the board. This requires removing the gap in the solder mask on the back of the board so that the solder mask material covers the through hole. If the through hole is small, the solder mask will plug the through hole; Therefore, the solder cannot penetrate the PCB.
However, this may cause another problem: flux aggregation. After the through hole is plugged, flux (a component of solder paste) may accumulate in the through hole. Some flux formulations may be corrosive. If they are not removed, it will lead to reliability problems over time. However, most modern cleaning-free flux processes are not corrosive and will not cause problems.
Please note that hot air pads must not be used for thermal vias, they must be directly connected to the copper area.
It is suggested that PCB designers should check PCB assemblies with surface mount technology (SMT) process engineers to select the best via size and structure suitable for the assembly process, especially when thermal vias are placed in the IC board area.
C. arrangement of capacitors
The component layout guide of motor IC is similar to other types of power IC. The bypass capacitor should be as close to the power supply pin of the device as possible, and the large capacity capacitor should be placed next to it. Many motor driver ics use pilot and/or charge pump capacitors, which should also be placed near the IC.
Most signals are routed directly at the top layer. The power supply is routed from the bulk capacitor to the bypass and charge pump capacitors on the bottom layer, and multiple through holes are used at the transitions of each layer.
There is a large exposed IC board on the bottom layer of TSSOP and QFN packaged devices. The IC board is connected to the back of the chip for removing heat from the device. The IC board must be fully soldered to the PCB to consume power.
The mold opening used for depositing the solder paste of the IC board is not necessarily detailed in the IC data sheet. Usually, SMT process engineers have their own rules about how much solder should be deposited on the mold and which pattern should be used in the mold.
If a single opening similar to the size of IC board is used, a large amount of solder paste will be deposited. This may cause the device to be lifted due to the surface tension when the solder melts. Another problem is solder voids (cavities or gaps in solder areas). In the process of reflow soldering, when the volatile components of flux evaporate or boil, solder voids will appear. This may cause solder to be pushed out of the solder joint.
To solve these problems, solder paste is usually deposited in several small square or circular areas for IC boards with an area greater than about 2 square millimeters. Dividing the solder paste into smaller areas can make it easier for the volatile components of the flux to escape out of the solder paste without displacing the solder.
The solder mold of QFN package has four small openings for depositing solder paste on the central IC board.
SOT-23 and SOIC packages
Standard lead packages (such as SOIC and SOT-23 packages) are usually used in low-power motor drivers.
In order to fully improve the power consumption capability of the lead package, a “flip chip lead frame” structure is adopted. Without using bonding wires, the chip is bonded to the metal leads by using copper bumps and solder, so that heat can be conducted from the chip to the PCB through the leads.
Flip-chip lead frame structure is helpful to fully improve the power consumption of lead package.
By connecting a larger copper area to a lead carrying a larger current, the thermal performance can be optimized. On the motor driver IC, the power supply, ground and output pins are usually connected to the copper area.
The following figure shows the typical PCB layout of “flip chip lead frame” SOIC package. Pin 2 is the device power supply pin. Note that the copper area is placed near the top device, and several thermal vias connect this area to the copper layer on the back of the PCB. Pin 4 is a grounding pin and is connected to the grounding copper-clad area of the surface layer. Pin 3 (device output) is also routed to the larger copper area.
Flip chip SOICPCB layout
Please note that there is no hot air pad on SMT board; They are firmly connected to the copper area. This is essential to achieve good thermal performance.
QFN and TSSOP packages
The TSSOP package is rectangular and uses two rows of pins. The TSSOP package of motor IC usually has a large exposed board at the bottom of the package to remove the heat from the device.
TSSOP packages usually have a large exposed board at the bottom to remove heat.
QFN package is a leadless package, with a board around the outer edge of the device, and a larger board at the center of the bottom of the device. This larger plate is used to absorb the heat in the chip.
To remove the heat from these packages, the exposed boards must be well soldered. The exposed board is usually grounded, so it can be connected to the ground layer of PCB.
Ideally, the thermal vias are located directly in the plate area. In the example of TSSOP package, an array of 18 through holes with a diameter of 0.38mm is used. The calculated thermal resistance of the via array is about 7.7 C/W
A TSSOP package PCB layout with an array of 18 thermal vias is adopted.
Generally, these thermal through holes use a drilling diameter of 0.4mm or less to prevent tin penetration. If the SMT process requires a smaller aperture, the number of apertures should be increased to keep the overall thermal resistance as low as possible.
In addition to the through holes located in the board area, the external area of the IC body is also provided with thermal through holes. In the TSSOP package, the copper area can extend beyond the package end, which provides another way for the heat in the device to pass through the top copper layer.
The boards around the edge of QF device package avoid using copper layer at the top to absorb heat. Thermal vias must be used to dissipate heat to the inner layer or the bottom layer of PCB.
QFN package PCB layout with 9 thermal vias
The PCB layout in the figure shows a small QFN(4×4mm) device. In the exposed board area, only nine thermal through holes are accommodated. Therefore, the thermal performance of this PCB is inferior to that of TSSOP package.
Flip chip QFN package
Flip-chip QFN(FCQFN) package is similar to the conventional QFN package, but its chip is directly connected to the board at the bottom of the device in a flip-chip manner instead of using bonding wires to connect to the package board. These boards can be placed on the opposite side of the heat-generating power devices on the chip, so they are usually arranged in long strips instead of small boards.
These packages are bonded to the lead frame by using multiple rows of copper bumps on the surface of the chip.
The FCQFN package uses multiple rows of copper bumps on the surface of the chip to bond to the lead frame.
The via hole can be placed in the board area, similar to the conventional QFN package. On multilayer boards with power and ground planes, through holes can directly connect these boards to each layer. In other cases, the copper area must be directly connected to the board in order to absorb the heat in the IC into the larger copper area.
The device shown in the figure below has a long power supply and ground plate, and three output ports. Please note that this package is only 4×4mm in size.
PCB layout of FCQFN packaged IC
The copper area on the left side of the device is the power input port. This larger copper area is directly connected to the two power boards of the device.
Three output boards are connected to the copper area on the right side of the device. Note that the copper area expands as much as possible after exiting the board. In this way, heat can be sufficiently transferred from the plate to the ambient air.
At the same time, pay attention to several rows of small through holes in the two boards on the right side of the device. These boards are grounded, and a solid grounding layer is placed on the back of PCB. The diameter of these through holes is 0.46mm, and the diameter of drilling holes is 0.25 mm The through hole is small enough to fit in the plate area.
To sum up, in order to use the motor driver IC to implement a successful PCB design, the PCB must be carefully laid out. Therefore, this paper provides some practical suggestions, hoping to help PCB design outsourcing artists achieve good electrical and thermal performance of PCB.