MCU Introduction and Details
? The history of the emergence of microcontroller is not long, but the development is very rapid. Its generation and development and the generation and development of microprocessors are largely synchronized, since 1971, the United States Intel Corporation first introduced 4-bit microprocessor, its development so far can be roughly divided into five stages. The following development of Intel's microcontrollers as a representative to be introduced.
1971-1976
The primary stage of microcontroller development. In November 1971, Intel first designed a 4-bit microprocessor Intel 4004 with an integration degree of 2,000 transistors/chip, and equipped with RAM, ROM, and shift registers, which constituted the first MCS-4 microprocessor, and then launched an 8-bit microprocessor Intel 8008, as well as other 8-bit microprocessors launched by other companies one after another.
1976-1980
Low-performance microcontroller stage. In 1976 Intel launched the MCS-48 series as a representative, using the 8-bit CPU, 8-bit parallel I/O interface, 8-bit timer/counter, RAM and ROM, etc. integrated in a semiconductor chip on the monolithic structure, although its addressing range is limited (not greater than 4 KB), there is no serial I/O, RAM, ROM capacity is small, the interrupt system is also relatively simple, but the function can meet the general industrial control and intelligent control. Although its address range is limited (more than 4 KB), there is no serial I/O, RAM, ROM capacity is small, the interrupt system is also relatively simple, but the function can meet the general industrial control and intelligent instruments, meters and other needs.
1980-1983
High-performance microcontroller stage. This phase of the introduction of high-performance 8-bit microcontroller with a common serial port, multi-level interrupt processing system, more than 16-bit timer/counter. On-chip RAM, ROM capacity increased, and the addressing range up to 64 KB, individual chip with an A/D converter interface.
1983-80s
16-bit microcontroller stage. In 1983, Intel introduced the MCS-96 series of high-performance 16-bit microcontrollers, which utilized the latest manufacturing processes to enable chip integration of up to 120,000 transistors per chip.
The 1990s
Microcontrollers have evolved to a higher level of integration, functionality, speed, reliability, and applicability.
According to the characteristics of microcontroller, the application of microcontroller is divided into single-machine application and multi-machine application. In an application system, only use a piece of microcontroller is called single machine application. The scope of the microcontroller's single-machine application includes:
(1) Measurement and control system. With the microcontroller can constitute a variety of less complex industrial control systems, adaptive control systems, data acquisition systems, etc., to achieve the purpose of measurement and control.
(2) Intelligent instrumentation. The use of microcontrollers to transform the original measurement and control instruments, to promote the instrumentation to digital, intelligent, multi-functional, integrated, flexible direction of development.
(3) Mechatronics products. Combination of microcontroller and traditional mechanical products, so that the traditional mechanical products to simplify the structure, control wisdom.
(4) Intelligent interface. In the computer control system, especially in the larger industrial measurement and control system, with a microcontroller interface control and management, coupled with the parallel work of the microcontroller and the host, greatly improving the operating speed of the system.
(5) Intelligent civil products. Such as household appliances, toys, game consoles, audio and video equipment, electronic scales, cash registers, office equipment, kitchen equipment and many other products, the introduction of MCU controllers, not only to make the product features greatly enhanced, performance has been improved, but also to obtain a good use of the results.
MCU multi-machine system can be divided into functionally decentralized systems, parallel multi-machine processing and local network systems.
(1) Functional distribution system. Multi-functional centralized system is a multi-machine system set up to meet the requirements of multiple peripheral functions of the engineering system.
(2) Parallel multi-computer control system. Parallel multi-computer control system is mainly to solve the problem of rapidity of the engineering application system, in order to constitute a large-scale real-time engineering application system.
(3) Local network systems.
Microcontroller according to the scope of application can be divided into general-purpose and specialized. Specialized is designed for a particular product, such as thermometers for the microcontroller, microcontroller for washing machines and so on. In the general-purpose microcontroller, and can be divided into 4-bit word length, 8-bit, 16/32-bit, although the computer's microprocessor is now almost 32/64-bit world, 8-bit, 16-bit microprocessor has tended to atrophy, but the situation of the microcontroller is different, 8-bit microcontroller is low-cost, inexpensive, easy to develop, and its performance meets the majority of the needs only in the aerospace, automotive, robotics and other high-tech fields, the need for high-speed processing of large amounts of data. Only in aerospace, automotive, robotics and other high-tech fields, the need for high-speed processing of large amounts of data, only need to choose 16/32-bit, and in the general industrial sector, 8-bit general-purpose microcontroller, is still the most widely used set of microcontrollers.
So far, China's microcontroller set and embedded systems development over the course of more than 20 years, with the embedded system gradually penetrate into all aspects of social life, the teaching of microcontroller courses from the traditional 8-bit processor platform to the 32-bit advanced RISC processor platform to change the trend, but the 8-bit machine is still difficult to be replaced. National economic construction, military and household appliances and other fields, especially cell phones, automotive automatic navigation equipment, PDA, intelligent toys, intelligent home appliances, medical equipment and other industries are in urgent need of microcontroller talent industry. Industry high-end more than 100,000 engineers engaged in microcontroller development set, but the face of the trend of industrialization of embedded systems and China's efforts to promote the construction of "embedded software factory" opportunities, China's embedded products to dissolve into the international market, the formation of industry, will be in urgent need of a large number of microcontroller set of talents, which provides great opportunities for higher vocational students to engage in this type of high-tech industry.
Categorized by use:
General-purpose: the development of resources (ROM, RAM, I/O, EPROM) and so on all provided to the user.
Specialized: Hardware and instructions are designed for a specific purpose, e.g., tape recorder controller, printer controller, motor controller, etc.
Specialized: The hardware and instructions are designed for a specific purpose.
Classified by the number of bits of data processed in its basic operation:
According to the width of the sink or data register, MCUs are divided into 1-bit, 4-bit, 8-bit, 16-bit, 32-bit, and even 64-bit microcontrollers.
The 4-bit MCUs are mostly used in calculators, automotive instrumentation, automotive immobilizers, pagers, and cordless phones, 4-bit MCUs are mostly used in calculators, car meters, car anti-theft devices, pagers, wireless phones, CD players, LCD driver controllers, LCD game consoles, children's toys, weighing scales, chargers, tire pressure gauges, temperature and humidity gauges, remote controls, and dummy cameras, etc.; 8-bit MCUs are mostly used in power meters, motor controllers, electric toy machines, inverter air conditioning, pagers, fax machines, callerID, telephone recorders, CRT monitors, keyboards, and USBs; 8-bit, 16-bit MCUs, 16-bit, 32-bit, and even 64-bit MCUs. 8-bit and 16-bit MCUs are mainly used in general control areas, generally not using the operating system, 16-bit MCUs are mostly used in mobile phones, digital cameras and camcorders, etc.; 32-bit MCUs are mostly used in Modems, GPS, PDAs, HPCs, STBs, Hubs, Bridges, Routers, workstations, ISDN phones, laser printers and color fax machines; 32-bit is used for color fax machines; 32-bit is used for color fax machines; and 32-bit is used for color fax machines; 32-bit is used for color fax machines; 32-bit is used for color fax machines; 32-bit is used for color fax machines.
Intel? Advanced Vector Computing
Intel? Advanced Vector Computing is a video interface standard designed for carrying uncompressed digital video data and to maximize the visual quality of digital display devices, such as LCD monitors, digital projectors, and digital projectors, using the latest technology and hardware.
8-bit MCUs operate at frequencies between 16 and 50MHz, emphasizing simple performance and low-cost adoption, and still have a certain status in the current MCU market, while many MCU vendors continue to develop frequency-regulated, energy-efficient designs for 8-bit MCUs in order to meet the demands of product development in a greener era.
16-bit MCUs, with 16-bit operation, 16/24-bit addressability, and 24~100MHz frequency are the mainstream specifications, and some 16-bit MCUs provide additional 32-bit special instructions for addition/subtraction/multiplication/division. Due to the continuous price reduction of 32-bit MCUs and the low price advantage of 8-bit MCUs, the 16-bit MCU market has been squeezed in the middle and has become the lowest shipping ratio.
32-bit MCUs are the mainstay of the MCU market, with a single price point of $1.5 to $4, and an operating frequency of 100 to 350 MHz, which provides better performance, and a wide range of application types.
Intel? Advanced Vector Extensions (AVX)
Intel? Advanced Vector Extensions (AVX) Advanced Vector Extensions (AVX)
Auto Sync with Cloud
No manual interaction required to synchronize with cloud services.
Types of Embedded Program Memory
The following 51 microcontrollers as an example (MCS-51 series MCUs are the most used microcontrollers in our country) can be divided into the following basic types according to the type of internal memory:
1. No ROM type :8031
2. ROM type :8051
3. EPROM type :8751
4. EEPROM type :8951
5. Enhanced type :8032/8052/8752/8952/C8051F
MCU can be divided into two kinds of memory types according to the type of memory without on-chip ROM type and with on-chip ROM type. For the chip without on-chip ROM type, it must be connected to an external EPROM in order to apply (typical chip is 8031). Chips with on-chip ROM are divided into on-chip EPROM (typical 87C51), MASK on-chip mask ROM (typical 8051), and on-chip FLASH (typical 89C51), etc. Some companies have also introduced chips with on-chip One Time Programming (OTP) ( typical 97C51). MASKROM MCU price is cheap, but the program has been solidified in the factory, suitable for fixed program application occasions; FLASH ROM MCU program can be repeatedly erased, flexibility is very strong, but the price is higher, suitable for the price is not sensitive to the application occasions or development purposes; OTPROM MCU price is between the first two, but also has a one-time programmable ability, suitable for both OTPROM MCU price between the first two, but also has a one-time programmable ability, suitable for both the requirements of a certain degree of flexibility, but also requires a low-cost set of occasions, especially the function of continuous renovation, the need for rapid mass production of electronic products.
As MCUs emphasize maximum density and minimum chip area to achieve control functions with limited program code, most of today's MCUs use built-in MaskROM, OTP ROM, EEPROM, or Flash memory to store firmware code, and the capacity of the MCU's built-in Flash memory ranges from a low 4~64KB to a high 512KB~2MB. The newest addition to the MCU is a new generation of memory, which can be used to store up to 2MB of memory.
Memory Structures
MCUs can be categorized into Harvard and Von Neumann based memory structures. Von Neumann structure. Now the vast majority of microcontrollers are based on the von Neumann structure, this structure clearly defines the four basic parts necessary for embedded systems: a central processor core, program memory (read-only memory or flash memory), data memory (random memory), one or more timing/timer, as well as used to communicate with peripheral devices and extended resources, input/output ports, all of which are used to communicate with peripheral devices and extended resources, the input/output ports, all of which are used to communicate with peripheral devices and extended resources. Output ports, all of which are integrated on a single integrated circuit chip.
Instruction Structure
MCUs can be categorized into CISC (Complex Instruction Set Computer) and RISC (Reduced Instruction Set Computer) based on their instruction structure. Reduced Instruction Set Comuter (RISC)
Technical PrincipleThe MCU and the temperature sensor are connected via an I2C bus, which occupies two MCU input/output lines, and the communication between the two relies entirely on the software. The address of the temperature sensor can be set by two address pins, which makes it possible to connect 8 such sensors on one I2C bus at the same time. In this scheme, the 7-bit address of the sensor has been set to 1001000. when the MCU needs to access the sensor, it first issues an 8-bit pointer to the register, and then issues the address of the sensor (7-bit address, the lower bit is the WR signal). There are three registers in the sensor that can be used by the MCU, and the 8-bit register pointer is used to determine which register the MCU is actually going to use. In this scenario, the main program keeps updating the sensor's configuration registers, which causes the sensor to operate in single-step mode, measuring the temperature with each update.
To read the contents of the sensor's measured value register, the MCU must first transmit the sensor address and the register pointer.The MCU sends a start signal, followed by the sensor address, and then sets the RD/WR pin high to read the measured value register.
In order to read the 16-bit data in the sensor's measured value register, the MCU must communicate with the sensor twice with 8-bit data. When the sensor is powered up and operating, the default measurement accuracy is 9 bits with a resolution of 0.5 C/LSB (range -128.5 C to 128.5 C). This program uses the default measurement accuracy, and the sensor can be reset to increase the measurement accuracy to 12 bits as needed. If only a general temperature indication is required, such as a thermostat, then a resolution of 1 C is sufficient. In this case, the lower 8 bits of the sensor data can be ignored and only the upper 8 bits of data can be used to meet the design requirement of 1 C resolution. Since reading the register is in the order of high 8 bits first and then low 8 bits, the low 8 bits of data can be read or not read. The only benefit of reading the high 8-bit data is twofold: the first is that it can shorten the working time of the MCU and sensors, reducing power consumption; the second is that it does not affect the resolution index.
After the MCU reads the measured value of the sensor, the next step is to convert and display the result on the LCD. The whole process includes: determining the positive and negative sign of the display result, converting the binary code to BCD code, and transferring the data to the relevant temporary memory of the LCD.
After the data is processed and the result is displayed, the MCU sends a single-step instruction to the sensor.
Key DifferencesAmong the most lauded achievements of the 20th century were the development of integrated circuits and electronic computers, and the emergence of minicomputers in the 1970s led to far-reaching changes in the scientific and technological community. In the mid-1970s, the microcomputer family splintered into a small faction, the microcontroller. With the emergence of 4-bit microcontrollers followed by the introduction of 8-bit microcontrollers. the MCS48 series, especially the emergence of the MCS51 series of microcontrollers, the establishment of microcontrollers as microcontrollers (MCUs), causing a new change in the field of microcomputers. In today's world, microprocessors (MPUs) and microcontrollers (MCUs) form two branches with their own characteristics. They are different from each other, but they integrate and promote each other. Unlike microprocessors (MPUs), which are characterized by rapid development in terms of computing performance and speed, microcontrollers (MCUs) are marked by the continuous improvement of their control functions.
CPU (Central Processing Unit, Central Processing Unit) developed three branches, one is DSP (Digital Signal Processing/Processor, Digital Signal Processing), the other two are MCU (Micro Control Unit, Microcontroller Unit) The other two are MCU (Micro Control Unit, Microcontroller Unit) and MPU (Micro Processor Unit, Microprocessor Unit).
MCU integrated on-chip peripheral devices; MPU without peripheral devices (such as memory arrays), is a highly integrated general-purpose structure of the processor, is removed from the integrated peripherals of the MCU; DSP computing power, good at a lot of repetitive data operations, while the MCU is suitable for a variety of different sources of information, diagnostic data processing and computing, focusing on the control, the speed is not as good as DSP. The most important feature that distinguishes MCUs from DSPs is their versatility, which is reflected in the instruction set and addressing mode. the combination of DSPs and MCUs is DSC, which will eventually replace both chips.
1. Support for intensive multiplication
GPPs are not designed to do intensive multiplication tasks, and even some modern GPPs require multiple instruction cycles to do a multiplication. DSP processors, on the other hand, use specialized hardware to implement single-cycle multiplication, and DSP processors add accumulator registers to handle the sum of multiple products. The accumulator register is usually wider than the other registers, and extra bits called result bits are added to avoid overflow. Also, to fully realize the benefits of specialized multiply-accumulate hardware, almost all DSP instruction sets include explicit MAC instructions.
2. Memory Structures
Traditionally, GPPs have used von. Neumann memory structure. In this structure, only one memory space is wired to the processor core through a set of sinks (an address sink and a data sink). Typically, four memory accesses occur to do a multiplication, taking at least four instruction cycles.
Most DSPs use a Harvard architecture, which divides the memory space into two, storing program and data. They have two sets of sinks wired to the processor core, allowing simultaneous accesses to them. This arrangement doubles the bandwidth of the processor memory and, more importantly, provides both data and instructions to the processor core. With this layout, the DSP is able to implement single-cycle MAC instructions.
A typical high-performance GPP actually incorporates two on-chip high-speed caches, one for data and one for instructions, which are wired directly to the processor core to speed up runtime access. Physically, this on-chip dual-memory and sink architecture is almost identical to the Harvard architecture. Logically, however, there are important differences.
GPPs use control logic to determine which data and instruction words are stored in on-chip high-speed caches, which the programmer does not specify (and may not even know). In contrast, DSPs use multiple on-chip memories and multiple sinks to guarantee multiple accesses to memory per instruction cycle. When using a DSP, the programmer has explicit control over which data and instructions are stored in the on-chip memory. The programmer must ensure that the processor is able to efficiently use its dual sinks when writing the program.
In addition, DSP processors almost never have high-speed caching of data. This is because the typical data for a DSP is a stream of data. That is, after a DSP processor does a calculation on each data sample, it discards it and hardly ever reuses it.
3. Zero-overhead loops
If one understands that one of the ***identical characteristics of DSP algorithms is that most of the processing time is spent executing smaller loops, it is also easy to understand why most DSPs have specialized hardware for zero-overhead loops. By zero-overhead loops, we mean that the processor doesn't have to spend time checking the value of the loop counter, conditionally shifting to the top of the loop, and decrementing the loop counter by 1 while executing the loop.
In contrast, GPP loops use softwares to implement them. Some high-performance GPPs use transfer forecasting hardwares to achieve almost the same effect as the zero-overhead loops supported by the hardwares.
4. Fixed-point computation
Most DSPs use fixed-point computation rather than floating-point. While DSPs must pay close attention to numerical precision in their routines, and it should be much easier to do so with floating-point, it is also important for DSPs to be inexpensive. Fixed-point machines are cheaper (and faster) than their floating-point counterparts. In order to be digitally accurate without using a floating-point machine, DSP processors support saturated computation, rounding, and shifting, both in the instruction set and in the hardware.
5. Specialized addressing modes
DSP processors often support specialized addressing modes that are useful for common signal processing operations and algorithms. For example, modular (cyclic) addressing (useful for implementing digital filter delay lines), and bitwise inverted addressing (useful for FFTs). These very specialized addressing modes are not often used in GPP and are only implemented in softwares.
6. Execution time prediction
Most DSP suites (e.g., cellular and datacom) are strictly real-time suites, where all processing must be completed within a specified time. This requires the programmer to determine exactly how much processing time is required for each sample, or, at least, to know how much time is required in the worst case. If one intends to use a low-cost GPP to accomplish the task of real-time signal processing, the prediction of execution time will probably not be much of an issue, since low-cost GPPs have a relatively straightforward structure that makes it relatively easy to predict execution time. However, most real-time DSP suites require processing power that low-cost GPPs cannot provide. The advantage of DSPs over high-performance GPPs at this point is that, even with DSPs that use high-speed caching, it is up to the programmer (not the processor) to decide which instructions will be put in, so it is easy to determine whether an instruction is being read from a high-speed cache or from memory.DSPs do not generally use dynamic features such as transfer prediction and inferred execution. Therefore, it is completely straightforward to predict the required execution time from a given piece of code. This allows the programmer to determine the performance limits of the chip.
7. The fixed-point DSP instruction set
The fixed-point DSP instruction set was designed with two goals in mind: to enable the processor to perform multiple operations per instruction cycle, and to increase the computational efficiency of each instruction cycle. Minimize the memory space required to store the DSP program (an issue that is particularly important in cost-sensitive DSP suites because memory has a significant impact on overall system cost). To achieve these goals, the instruction set of a DSP processor typically allows the programmer to specify several parallel operations within a single instruction. For example, a single instruction contains a MAC operation, i.e., one or two simultaneous data moves. In a typical example, a single instruction contains all the operations needed to compute one section of a FIR filter. The price paid for this high efficiency is that its instruction set is neither intuitive nor easy to use (compared to GPP's instruction set). GPP programmers usually do not care about the ease of use of the processor's instruction set because they generally use a high-level language such as C or C++. Unfortunately for DSP programmers, the main DSP applications are written (or at least partially optimized) in assembly language. There are two reasons for this:First, most widely used high-level languages, such as C, are not suitable for describing typical DSP algorithms. Second, the complexity of DSP architectures, such as multiple memory spaces, multiple sinks, irregular instruction sets, highly specialized hardware, etc., makes it difficult to write efficient compilers for them. Even if a compiler is used to compile the C raw code into DSP assembly code, the task of optimization remains heavy. Typical DSP suites have large computational requirements with tight overhead constraints, making program optimization essential (at least for the most critical parts of the program). Therefore, a key factor in considering the selection of a DSP is the existence of sufficient programmers who are well adapted to the DSP processor instruction set.
8. Development Tool Requirements
Because DSPs latch on to code that requires a high degree of optimization, most DSP vendors provide development tools to help programmers with their optimization work. For example, most vendors provide processor simulation tools to accurately simulate processor activity during each instruction cycle. These are useful tools both for ensuring real-time operation and for code optimization. GPP vendors typically do not provide such tools, primarily because GPP programmers do not typically need this level of detail, and the lack of simulation tools for GPPs that are accurate down to the instruction cycle is a big problem for DSP arbitrage developers: since it is almost impossible to predict the number of cycles that a high-performance GPP will need for a given task, it is impossible to show how to improve the performance of the code.
Leveraging ConferenceMCU Technology Innovation and Embedded Leveraging Conference is a forum for MCU technology exchanges and leveraging along with the Hi-Tech Fair. Ltd. by the Shenzhen Creative Times Exhibition Co., Ltd., the content of the arrangement is usually in the morning by domestic professionals on the mcu knowledge and innovation of the set of speeches, the afternoon forum, free to talk about the exchange of technology and industry trends.
Fourth
Time: August 21, 2012
Venue: Shenzhen Convention and Exhibition Center
Related Exhibitions: 2012 Embedded Systems Expo
Related Exhibitions: 2012 Embedded Systems Expo
Conference Name: The 4th MCU Technology Innovation and Embedded Adaptation Conference
Supporting Media: E-Exhibition.com
Keynote Speeches (Morning):
From MCU to SoC
From MCU to SoC
The newest and most innovative MCUs in the world, MCUs are the first in a long line of new technologies. MCU to SoC
Convergence, Openness and Innovation in MCU Technology
Seamlessly Upgrading Embedded Systems to 32-bit MCUs
Greener, More Reliable Embedded Designs, and More
Sub-Forums (Afternoon) Delve into More Embedded Suites:
Sub-Forums Forum 1: Home Appliances/Smart Home
Sub-Forum 2: HMI/IPC
Sub-Forum 3: Motor Control
The third
will focus on
Time: November 18, 2011
Venue: Shenzhen
Organizer: 13th Shenzhen Hi-Tech Fair Electronics Exhibition Organizing Committee
Underwriter: Creative Times Exhibition e-Exhibition.com
The keynote speeches (in the morning) will include:
The development trend of multicore MCUs
The keynote speech is about the development of the embedded computing industry. MCU Development Trends
MCU to SoC
Security and Reliability for Intelligent Systems
MCU!2011 Reaching into the Newest Application Markets (Afternoon):
Sub-forum 1: Household Appliances / Intelligent
Session 2: Intelligent Metering
Session 3: HMI/IPC
Session 4: Motor Control
Previously Reviewed
Previously Reviewed
Previously Reviewed
Previously Reviewed
Previously Reviewed
Over 460 professionals attended in 2009
606 professionals attended in 2010
2010 Professional Audience Analysis
MCU!MCU!2010 attracted participants from IBM, Siemens, Advantech, Emerson, TCL, Skyworth, Konka, Midea, ZTE, Lenovo, Foxconn, Flextronics, BYD and hundreds of other well-known domestic and foreign enterprises to participate in 606 technical and management personnel:
More than half of the professional audience of R & D technicians
Technology R & D personnel accounted for 52%, followed by middle and senior management accounted for 33%; a small portion of the audience for the 33%; a small portion of the market / marketing personnel, accounting for 13%, the other accounted for 2%
Professional audience engaged in the field of distribution Participants in the enterprise, consumer electronics accounted for 37%; industrial electronics accounted for 24%; embedded systems design accounted for 22%; medical electronics accounted for 19%; automotive electronics, embedded software development, each accounted for 15%; Home appliances 14%; cell phones and communications 11%; IT and network 10%; other 9%.
2010 professional audience engaged in the field distribution 2010 mcu professional audience analysis