Four misunderstandings of DSP controller by developers
Time:2023-03-01
Views:1235
Motor control designers have recently encountered obstacles in the development of various applications such as household appliances and servo drives. They must choose between controller performance or expensive price. Most motor control applications are inherently low cost. One of the necessary conditions for market acceptance is that the product price should be attractive, which means that the cheapest controller that can achieve the work purpose and has limited other functions must be selected. The intelligent controller based on DSP is changing this situation, and the deadlock has been broken. The new controller based on DSP has achieved significant performance improvement at a relatively moderate price.
Simple control algorithms such as voltage Hertz constant and six step communication are difficult to achieve the performance required for high efficiency and optimization of fuselage size. The intelligent controller based on DSP has changed the above situation in two aspects.
First, it adds counting capability. This enables designers to implement control algorithms with higher performance, such as magnetic field-oriented control. Secondly, the algorithm with higher computational strength also enables designers to use more efficient motors. For example, we can replace AC induction motor with permanent magnet motor, which further improves efficiency and dynamic performance.
The so-called "problem" of vector control and other advanced technologies with high computational intensity is that multiplication and accumulation (MAC) operations occupy the majority of the algorithm. The standard 8, 16 or 32 bit microcontroller cannot handle the above operations because of the lack of appropriate bus architecture to achieve mathematical efficiency. In the end, this means that we must fundamentally change the design, not to DSP, but to 32-bit controller based on DSP.
It is not surprising that there are many common misunderstandings about the use of DSP controllers, such as:
*DSP controller does not have motor control peripheral;
*The code density problem makes the 32-bit system based on DSP hard to be accepted;
*DSP may be suitable for control algorithm, but it can not handle other control tasks well;
*DSP software is very difficult, and designers must endure the pain of no real-time operating system and good tool support.
Let‘s discuss the above misunderstandings.
Peripheral integration
When it comes to heavy-duty digital signal processing, the first idea in the minds of motor control engineers may be the DSP controller that is good at digital computing -- but how to deal with the universal integrated peripheral functions?
Today‘s motor control DSP optimizes the application of motor control, and integrates functions such as pulse width modulator (PWM), encoder interface, communication port and analog-to-digital converter (ADC). It also includes a large number of flash memory and RAM, which eliminates the need for external memory devices.
Code density
In fact, the traditional 32-bit microcontroller architecture has inherent weaknesses in code density, which is a key issue for low-cost applications with limited memory capacity.
The new 32-bit DSP architecture uses carefully selected 16-bit and 32-bit instructions to achieve the best code density. Use 32-bit instructions only when needed.
Control-oriented architecture
The best 32-bit DSP controller has atomic read-modify-write instructions and other functions, which can simplify programming. The interrupt waiting time is also short, less than 100 nanoseconds, thus achieving a highly responsive CPU.
Software support
DSP assembler has developed to an advanced level where it is rarely required to write assembly code. In addition, algorithm development has become much simpler. For example, Texas Instruments (TI) provides a huge library of commonly used algorithms that can be immediately run on its processor, as well as BIOS?, A real-time operating system optimized for DSP.
If you need to customize the algorithm, we will also provide help. In order to create their own algorithms, engineers can use IQMath? Tool, which can simplify the development of mathematical functions for fixed-point calculation.
TI is practicing Moore‘s Law and carrying out innovative work to make the task of motor control designers easier and simpler.
Disclaimer: This article is transferred from other platforms and does not represent the views and positions of this site. If there is any infringement or objection, please contact us to delete it. thank you! |