What is the connection between microcontrollers, ARM, MCU, DSP, FPGA, and embedded systems?

A popular explanation of the complex relationship between MCU, ARM, MUC, DSP, FPGA and embedded systems!

First of all, "embedded" is a concept. There is no accurate definition. Different books have their own definitions. But the main idea is the same. Compared with general systems such as PCs, embedded systems are special systems with a streamlined structure. Only the necessary parts are retained in both hardware and software, and the unnecessary parts are cut off. Therefore, embedded systems generally have the characteristics of portability, low power consumption, and single performance.

Then, MCU, DSP, FPGA all belong to the category of embedded systems and are tools used to achieve a certain purpose.

MCU is commonly known as "single chip computer". After so many years of development, it is no longer limited to the Princeton structure 51, and its performance has been greatly improved. Because MCU must execute programs sequentially, it is suitable for control and is mostly used in industry. ARM is a company that specializes in designing MCUs. Due to its advanced technology and proper strategy, it has occupied a huge market share in the single chip microcomputer market in the past two years. There are many types of ARM single chip microcomputers, from low-end M0 (small appliances) to high-end A8 and A9 (mobile phones, tablets). So it is not necessary for ARM single chip microcomputers to be installed on the system. The key depends on the application scenario.

DSP is called digital signal processor. Its structure is different from MCU, which speeds up the operation and highlights the operation ability. It can be regarded as a super fast MCU. Low-end DSP, such as C2000 series, is mainly used in motor control, but TI seems to call it DSC (digital signal controller), which is something between MCU and DSP. High-end DSP, such as C5000/C6000 series, is generally used in video image processing and communication equipment, which require a lot of calculations.

FPGA is called Field Programmable Logic Array. It has no functions by itself, just like a blank sheet of paper. The functions it has are completely designed by programmers (all its processes are hardware, including VHDL and Verilog HDL programming, which are also hardware, generally referred to as programming "logic"). If you are good enough, you can turn it into an MCU or a DSP. Since the internal structures of MCU and DSP are designed, they can only be processed sequentially through software programming, while FPGA can process in parallel and sequentially, so it is the fastest in comparison.

So why do MCU, DSP and FPGA exist at the same time? That is because the internal structures of MCU and DSP are carefully designed by IC designers, and the power consumption and price are much lower than FPGA when completing the same functions. In addition, the development of FPGA itself is relatively complex, and it takes more manpower and financial resources to complete the same function. Therefore, each of the three has its own strengths and uses its own place. However, there is already a trend of integration among the three. ARM's M4 series has added a streamlined DSP core, TI's Da Vinci series itself is an ARM+DSP structure, and the newly launched FPGAs of ALTERA and XINLIX all contain ARM cores. Therefore, the relationship between the three is becoming more and more like three circles of three primary colors.

In a nutshell, "I am in you and you are in me."

Where to start learning to be a hardware engineer?

Single chip microcomputer: Usually has no operating system and is used for simple control, such as elevators, air conditioners, etc.

DSP: used for complex calculations, such as discrete cosine transform and fast Fourier transform. It is often used in image processing and is used in digital cameras and other devices.

ARM: A British chip design company that does not produce chips, but only sells intellectual property.

FPGA: Field Programmable Gate Array, a circuit design completed in hardware description language (Verilog or VHDL), can be quickly burned to the FPGA for testing after simple synthesis and layout, and is the mainstream technology for modern IC design verification.

Compared with desktop computers, embedded systems can be customized and have various shapes. They may be limited in size, power consumption, cost, and have high real-time requirements. For example, oscilloscopes, mobile phones, tablet computers, fully automatic washing machines, routers, and digital cameras. Although there is no desktop computer in these devices, one or more embedded systems are working.

Different options are provided according to the functional complexity and computational complexity of the object system. For simple home appliance control embedded systems, a simple 8-bit microcontroller is sufficient, which is inexpensive and good quality. For mobile phones and game consoles, 32-bit ARM and DSP chips must be used. FPGA is a more hardware-oriented implementation method.

So to become a hardware engineer through learning, you have to start with microcontrollers, and then learn ARM, DSP and the like.

A detailed introduction to the seven major MCUs on the market

Microcontrollers are now everywhere and there are many types of them, which makes developers overwhelmed. The development is also quite rapid, from the 1980s, from the 4-bit and 8-bit ones to the various high-speed microcontrollers today.

Various manufacturers have also been up and down in speed, memory, and functions. At the same time, a large number of manufacturers with representative microcontrollers have emerged: Atmel, TI, ST, MicroChip, ARM... The domestic Hongjing STC microcontroller is also remarkable...

The following is a comparison of the advantages and disadvantages of 51, MSP430, TMS, STM32, PIC, AVR, and STC microcontrollers and their functional performance...

51 single chip microcomputer

The most widely used 8-bit microcontroller is of course also the easiest one for beginners to learn. It was first introduced by Intel. Due to its typical structure and centralized management of perfect bus-specific registers, numerous logic bit operation functions and rich control-oriented instruction system, it can be called a "classic" of a generation, laying the foundation for the development of other microcontrollers in the future.

The main reasons why the 51 single-chip microcomputer has become a classic and easy-to-use single-chip microcomputer are as follows:

characteristic:

1. From the internal hardware to the software, there is a complete bit-by-bit operating system, called a bit processor, which processes bits instead of words or bytes. It can not only process certain bits of special function registers in the chip, such as transfer, set, clear, test, etc., but also perform bit logic operations. Its functions are very complete and easy to use.

2. At the same time, a dual-function address range is specially opened in the on-chip RAM area, which is extremely flexible to use. This function undoubtedly provides great convenience to users.

3. Multiplication and division instructions, which also brings convenience to programming. Many 8-bit microcontrollers do not have multiplication functions, and when performing multiplication, you have to program a subroutine call, which is very inconvenient.

Disadvantages: (Although it is a classic, the disadvantages are still obvious)

1. AD, EEPROM and other functions need to be expanded, increasing the burden on hardware and software

3. The running speed is too slow, especially the dual data pointer. If it can be improved, it will bring great convenience to programming.

4. 51 has poor protection capability and can easily burn out the chip

Applications:

At present, it is widely used in teaching occasions and occasions with low performance requirements.

Most used devices: 8051, 80C51

MSP430 MCU

The MSP430 series microcontroller is a 16-bit ultra-low power mixed signal processor that Texas Instruments launched in 1996. Its biggest highlight is its low power consumption and high speed. Its assembly language is very flexible, it has many addressing modes, few instructions, and is easy to use. This is mainly because it integrates many analog circuits, digital circuits, and microprocessors on one chip to provide practical application needs.

"Single chip" solution. Its rapid development and expanding application scope are mainly due to the following characteristics...

characteristic:

1. Powerful processing capability, using a reduced instruction set (RISC) structure, with a variety of addressing modes (7 source operand addressing, 4 destination operand addressing), concise 27 core instructions and a large number of simulation instructions; a large number of registers and on-chip data memory can participate in a variety of operations; there are also efficient table lookup processing instructions; with a high processing speed, the instruction cycle is 125 ns under 8MHz crystal drive. These features ensure that a highly efficient source program can be compiled

2. In terms of operation speed, it can achieve a 125ns instruction cycle under the drive of an 8MHz crystal. The combination of 16-bit data width, 125ns instruction cycle and multi-function hardware multiplier (capable of multiplication and addition) can realize certain algorithms of digital signal processing (such as FFT, etc.)

3. Ultra-low power consumption: The reason why the MSP430 microcontroller has ultra-low power consumption is that it has its own unique features in reducing the chip power supply voltage and flexible and controllable operating clock. The power supply voltage is 1.8~3.6V. Therefore, when it runs under the clock condition of 1MHz, the current of the chip will be around 200~400uA, and the minimum power consumption in the clock shutdown mode is only 0.1uA.

shortcoming:

1. I personally feel that it is not easy to get started, not suitable for beginners, and there is relatively little information, so you can only go to the official website to find it

2. It takes up a lot of instruction space, because it is a 16-bit microcontroller, the program is in words, and some instructions actually take up 6 bytes. Although the program appears simple, it takes up a lot of space compared to the PIC microcontroller.