Characteristics of MCU

      Since its advent, the path that single-chip microcomputers have taken is different from that of microprocessors. Microprocessors have developed in the direction of high-speed computing, data analysis and processing capabilities, and large-scale capacity storage to improve the performance of general-purpose computers. Their interface is also designed to meet the needs of peripherals and network interfaces. Single-chip microcomputers are based on the characteristics of industrial measurement and control objects, environments, and interfaces, and are developed in the direction of enhancing control functions, improving reliability in industrial environments, and flexibly and conveniently forming the interface of application computer systems. Therefore, single-chip microcomputers have their own characteristics, mainly: [1] [2]
　 · Diverse varieties and models. Varieties and models are expanded year by year to meet various needs. This gives system developers a great deal of freedom of choice. CPUs range from 4, 8, 16, 32 to 64 bits, and some also use RISC technology;
　 · Improve performance and expand capacity. The integration level has reached more than 2 million transistors. The bus operating speed has reached tens of microseconds. The operating frequency has reached 30MHz or even 40MHz. The instruction execution cycle is reduced to tens of microseconds. The memory capacity RAM has developed to 1K, 2K, ROM has developed to 32K, 64K;
　 ·Increase control functions and extend to external interfaces. Integrate the functions of the original peripheral chips into the chip. Today's single-chip microcomputers have developed to a chip containing a CPU. In addition to embedded RAM, ROM memory and I/O interface, there are also A/D, PWM, U ART, Timer/Counter, DMA, Watchdog, Serial Port, Sensor, driver, display driver, keyboard control, function generator, comparator, etc., forming a complete and powerful computer application system;
　 ·Low power consumption. The power supply voltage is reduced from 5V to 3V, 2V or even 1V. The working current is reduced from mA level to μA level. In terms of production technology, CMOS replaces NMOS and transitions to HCMOS;
　 ·Application software matching. Provides a software library, including standard application software and demonstration design methods. It makes it faster and more convenient for users to develop single-chip microcomputer application systems. It is possible to develop a new application product in one week;
　 ·System expansion and configuration. There are three bus structures DB, AB, and CB for expanding external circuits to facilitate the construction of various application systems. According to the characteristics of single-chip network systems and multi-machine systems, single-chip serial buses are specially developed. In addition, sensors, human-computer dialogue, network multi-channel interfaces, etc. are specially configured to form networks and multi-machine systems.
1.2 Performance of single-chip microcomputers〔3〕 Single-chip
　 microcomputers are usually classified according to the number of bits of their microprocessor word length, such as 4, 8, 16, 32, and 64-bit single-chip microcomputers.
1.2.1 4-bit machine
　 In 1971, Intel first launched the 4-bit microprocessor chip 4004. Since then, various manufacturers have successively launched 4-bit machine products. Because the 4-bit machine can only process one bit of BCD code data at a time, it is only suitable for simple control occasions. However, due to its low price, it is still widely used in consumer products such as home appliances. In recent years, in order to resist the erosion of 8-bit machines, 4-bit machines have made great progress in structure and performance. The main ones are:
　 · Using <1μm CMOS technology to achieve an instruction execution speed of 1μs. ROM is 32~64kB, RAM is 4k×4 bits. These enhanced performances are comparable to 8-bit machines;
　 ·Integrate display drivers such as LED, LCD, VFD into a single chip. This enhanced I/O function is even stronger than that of ordinary 8-bit machines;　
 　·Reduce power consumption. It can operate at a voltage lower than 2.2V and a current of μA level. This is lower than that of ordinary 8-bit machines;
　 ·Adopt a structure similar to RISC. The performance of 4-bit machines is greatly improved;
　 ·Add specific functions for specific applications. Such as A/D, D/A, zero-crossing detection, comparator, counter, timer, graphic display, remote control, etc.
　 But even so, the market share of 4-bit machines is still small. The performance is listed in Table 1 (limited by space, only a few explanations are given).
   

1.2.2 8-bit machine
　 Since 8-bit machines can process one ASCII character at a time, they are widely used. Such as display, terminal keyboard, printing, word processing, industrial control, etc. The market share is more than 70%. It has rich functions, complete varieties and strong versatility. In 1972, Intel first launched the 8-bit microprocessor 8008, and then in 1976, it took the lead in launching the 8-bit machine MCS-48 series. In 1980, it launched the MCS-51 series, whose performance greatly exceeded that of the 48 series. For example, the computing speed was 10 times that of the 48 series, and the instruction cycle could be 1μs when the clock was 12MHz. Motorola did not launch its first single-chip computer MC6801 series until 1978. However, it had quite rich functions, such as embedded EEPROM, A/D, LED drive, PWM output, etc., becoming a powerful industrial controller. Zilog also launched the Z8 series of single-chip computers in 1978. It appeared in a new look from the beginning. It can be used not only as a single-chip computer, but also as a microprocessor in microcomputer systems. It is worth mentioning that the 89C series of single-chip computers launched by Atmel〔4〕 has the same core as the industrial standard 80C51, that is, the instruction set and pin distribution are compatible with the MCS51 products. Its characteristic is that the clock frequency is higher, which can reach 33MHz and up to 40MHz, so the operation speed is faster. One of the main application fields of single-chip microcomputers in China is industrial measurement and control. The single-chip microcomputers used in this field are mostly 8-bit machines, and the MCS-51 series and AT89C5 series are the most used. Many colleges and universities use this type of machine as the background machine for microcomputer principle or single-chip microcomputer principle courses. There
　 are quite a lot of manufacturers producing 8-bit machines, and there are many varieties and models. In China, the mainstream is the faction products of the above-mentioned companies. And each faction has multiple manufacturers and their various models of products. For example, the Intel faction includes Philips/Signetics, Siemens, AMD, OKI, MARTRA-MHS and other company models. The Motorola faction includes Hitachi, Mitsubish, Rockwell, WDC and other company models. The Zilog faction includes NEC, Hitachi, SGS-Thomson and other company models. Among these factions, Intel has the highest market share, Motorola is in the middle, and Zilog has the lowest. In order to occupy the market, Motorola also developed the 6804/68HC04 chip, which is a series between 4-bit and 8-bit machines. The CPU processes 8-bit data and runs 8-bit instructions, but the data is output serially one bit at a time. The design goal is the price of a 4-bit machine, which is very suitable for occasions with high quality and low price.
         

1.2.3 16-bit machine
　 In 1978, Intel first launched the 16-bit microprocessor 868 series, followed by Motorola's M68000 and Zilog's Z8000, which became the three major series of 16-bit microprocessors at the time. Due to the wide application of 8-bit machines and their ability to solve problems, 16-bit single-chip microcomputers entered the market relatively late. Intel did not launch the MCS-96 series until 1988. This machine has high-speed computing, high-speed processing and control capabilities, a 16-bit CPU, an 8-bit external bus (hence it is also called a quasi-16-bit machine), a rich and efficient instruction system, and excellent performance-price ratio. Its price is only slightly higher than that of an 8-bit machine. There are A/D, PWM, Watchdog and a flexible interrupt system on the chip. The instruction execution time is 1 to 2μs at an operating frequency of 12MHz. Due to its good performance and functions, it attracted widespread attention from the industry as soon as it appeared. In my country, MCS-96 was the most widely used in the early years. In recent years, Intel80C196, 80C251, 80C51XA, etc. have been widely used. Other manufacturers of 16-bit machines include Motorola, MATRA-MHS, Mostek (Thomson), NEC, OKI, Phil ips/Signetic, Siemens, etc. However, the total output is still small, far lower than that of 8-bit machines. However, due to measures such as enhancing functions, improving performance, diversifying varieties and continuously reducing prices, it has developed rapidly in recent years and is expected to catch up with 8-bit machines in the late 1990s. The measures taken are mainly:
　 · Enhance computing power and increase capacity. There are sound multiplication and division instructions on the chip, the RAM capacity is increased to 2k, the ROM to 64k, and it can directly support C and Forth languages;
　 · Improve data processing and transmission capabilities. Generally, DMA transmission and fast I/O functions are added;
　 · Increase speed. For example, the operating frequency of 80C51XA reaches 30MHz, and the HPC series reaches 40MHz. The instruction cycle is reduced to several μs.
　 · Adopt RISC structure. For example, Waterscale's PSC1000 and Harris's RTX2000 all adopt RISC structure.
         

   At present, the production of 16-bit computers is the largest in Intel's 8096 series, and the quasi-16-bit computer 8098 has been discontinued. Among the CMOS-processed microcontrollers, National Semiconductor's HPC series has the highest performance-price ratio, and Hitachi's H8/500 is the microcontroller with the best overall performance.
1.2.4 32-bit computers
　 With the application and development of high technology in the fields of intelligent robots, optical disks, laser printers, real-time image and data processing, complex real-time control, network servers, etc., 16-bit computers have become powerless, and 32-bit computers are needed to meet the requirements. In the late 1980s, a variety of 32-bit computer products were launched. For example, Motorola's 68300 and Intel's 80960 were both produced in 1989. Their common characteristics are:
　 · Addressing capacity is above GB level (storing and processing color images requires a large memory);
　 · High instruction execution speed. M instructions per second. For example, Intel's i960A has a speed of 66MIPS;
　 · Fast computing power. Some are embedded with floating-point computing components, which greatly enhances computing power;
　 Directly support high-level languages ​​and real-time multi-task execution. For example, support C and Forth languages. Embed real-time multi-task operating systems. Most use RISC structure. Except for Motorola's MC68332 and National Semiconductor's NS32CG160, which still use CISC, all use RISC. 
            

1.2.5 64-bit machine
　  64-bit machine is used in engine control, intelligent robot control, disk control, voice/image communication, algorithm-intensive real-time control and other occasions. However, there is still no application in China. Here is a product example.
    Transputer T800 of Inmos Company in the UK is a 64-bit high-performance machine. It integrates processor, cache, 64-bit floating-point arithmetic unit, storage controller, serial interface, and is suitable for ultra-high-speed parallel processing.
    The central processing unit is 32 bits, and its floating-point computing speed reaches 1.2 billion times/second. RAM 4k, I/O link channel 4 groups 20Mbit/s, clock frequency 25MHz, data transfer rate 100MB/s, addressable external storage space 4GB. External memory transfer rate 33MB/s.