Embedded design is a huge project. Today, let's talk about several points to note in hardware circuit design. First of all, let's get to know the embedded hardware architecture.
    We know that the CPU is the soul of the entire system, and all peripheral configurations are related to it. This also highlights a feature of embedded design: the hardware is scalable. When doing embedded hardware design, the following points need to be paid attention to.

01 Determination of power supply
    The role of power supply in embedded systems can be regarded as the effect of air on the human body, or even more important: the air we breathe contains oxygen, carbon dioxide, nitrogen, etc., but the content is stable. This is equivalent to various noises in the power supply system. We hope to obtain a pure and stable power supply that meets the requirements, but due to various constraints, it is only our dream. This should focus on two aspects:
a、Voltage
    Embedded systems require power supplies of various magnitudes, such as the common 5V, 3.3V, 1.8V, etc. To minimize the ripple of the power supply as much as possible, LDO devices are used in embedded systems. If DC-DC is adopted, not only is it large in size, but its ripple is also a very troublesome problem.
b、Electric current
    The normal operation of an embedded system not only requires a stable and sufficient power supply but also a sufficient current. Therefore, when choosing power supply devices, the load needs to be considered, and generally, a 30% margin should be left.
    If it is a multi-layer board, the power supply section needs to be divided during layout. At this time, attention should be paid to the division path, and a certain amount of power supply should be placed together as much as possible. If it is a double-sided board, the width of the wire routing needs to be noted. It should be widened as much as possible if the board allows. The appropriate decoupling capacitor should be placed as close as possible to the power pin.

02 Determination of Crystal oscillator
    The crystal oscillator is equivalent to the heart of an embedded system, and its stability directly affects its operating status and communication performance. Common oscillators include passive crystal oscillators and active crystal oscillators. First, determine their oscillation frequency, and second, determine the type of crystal oscillator.
a、Passive crystal oscillator
    The selection of its matching capacitors and matching resistors is generally based on the reference manual. In the design of single-chip microcomputers, plug-in crystal oscillators are often used in combination with ceramic capacitors. In ARM, to reduce space and facilitate wiring, four-corner passive crystal oscillators are often used in combination with surface mount capacitors. Although we are quite familiar with the matching circuit of fixed crystal oscillators, to ensure absolute safety, it is still necessary to refer to the reference manual to determine details such as the size of the capacitor and whether a matching resistor is needed.
b、Active crystal oscillator
     It has a better and more accurate clock signal, but in comparison, it is more expensive than an unconnected crystal oscillator. Therefore, this is also a cost that needs to be considered in hardware circuit design. When designing a circuit board, it is necessary to pay attention to keeping the crystal oscillator traces as close to the chip as possible and keeping the key signals away from the clock traces. Add grounding protection rings when conditions permit. If it is a multi-layer board, the key signals should also be kept away from the traces of the crystal oscillator.

03 Reserve the test IO port
    During the embedded debugging stage, when there are abundant pin resources, I usually reserve an IO port to connect an led or a speaker, laying the groundwork for the next step of software programming. During the operation of the embedded system, the IO interface should be appropriately controlled to determine whether the system is operating normally.

04 External expansion memory
    If an embedded system has a power supply, a crystal oscillator and a CPU, then this is the smallest system we are familiar with. If the embedded system needs to run a larger operating system, not only does the CPU need to have an MMU, but the CPU also needs to be externally connected to SDRAM and NANDFLASH.
    If the cpu is equipped with SDRAM and NANDFLASH controllers, there is no need to overly consider the use of address lines in the hardware design. If there is no relevant controller, then attention should be paid to the use of address lines.
    This part is a key point during LAYOUT. The reason for this is to make the relevant signal lines of equal length to ensure that the signal delays are equal, as well as the differential signal line routing of the clock and DQS. When wiring, various wiring techniques need to be used in combination, such as symmetrical distribution with the cpu, Daisy chain wiring, and T-shaped wiring. All these should be selected based on the number of memory units. Generally speaking, the more memory units there are, the more complex the wiring will be. However, knowing the key points will make everything easy to solve.

05 Functional Interface
    The most important thing for an embedded system is to control the peripheral modules through various interfaces to achieve the preset purpose of the designer. Commonly used interfaces include serial ports (which can be used to connect modules such as Bluetooth, wifi and 4G), USB interfaces, network interfaces, JTAG interfaces, audio and video interfaces, HDMI interfaces, etc.
    Since these interfaces are connected to external modules, doing a good job in electromagnetic compatibility design is an important task. In addition, pay attention to the use of differential lines when LAYOUT.

06 Screen
    This function is listed separately because it is dispensable. If an embedded system merely serves as a connector to connect peripheral device modules, connects to the computer host through relevant interfaces or is directly hung on the network, then a screen is not necessary.
    However, if what is produced is a consumer product that interacts frequently with users, it is necessary to say a few words. Capacitive screens are the main components of embedded screens. In circuit design, attention should be paid to the layout of touch screen connection lines and display screen connection lines.
    During the wiring process, try to approach the main control cpu as short as possible. At the same time, pay attention to using differential lines for paired signals and equal lengths for RGB control signals. The spacing of various signal traces follows the 3W rule to avoid mutual interference. In the design of the screen, it is essential to ensure power and prevent interference to avoid the occurrence of screen flickering and screen distortion.