DSP solutions for power systems

    Abstract: This article introduces two DSP hardware platforms specially designed for power monitoring/monitoring systems, specifically explaining that DSP solutions are more efficient than traditional

    Keywords: power monitoring/monitoring RTU power protection DSP

Characteristics of power systems

A power monitoring/monitoring system needs to complete all or part of the following tasks:

① Collect voltage and current data of each phase at the same time, and calculate the effective value, power, active power, reactive power, apparent power and power factor of each phase voltage and current in real time;

② Based on certain fault criteria, determine whether a fault has occurred and record the fault;

③Monitor switch position;

④ Upload the monitored data to the central station according to the standard communication protocol;

⑤Receive the remote control data from the central station and issue the closing or opening command of the switch.

It can be seen from the above that compared with general industrial control systems, power monitoring/monitoring systems have two basic characteristics;

① Voltage/current are sampled simultaneously, and there is no phase difference between them, which facilitates the calculation of power and power factor.

② The amount of real-time data processing is large, requiring the use of high-speed processors.

Using DSP to replace MCU in power monitoring/monitoring systems is based on the second characteristic mentioned above.

Comparison between DSP and MCU

DSP is actually a special MCU. Compared with MCU, it has the following characteristics:

① There are multiple address, data and control buses on the chip, which can enable multiple control and computing components to work in parallel and improve the processing power of the CPU.

② There must be a hardware multiplier in the DSP, and the multiplication operation is completed in one instruction. Moreover, the multiplier is independent and can work in parallel with arithmetic components such as adders, which improves the digital processing capability of the CPU.

③There are some special instructions in DSP to accelerate digital processing. For example, the multiplication-accumulation (MAC) instruction completes multiplication and addition operations simultaneously in one instruction cycle.

④The main frequency is much higher than that of ordinary MCUs. From the instruction cycle point of view: low-end DSP is generally 50ns; mid-range DSP is generally 10ns; high-end DSP is generally 5ns. From a processing point of view; low-end DSP is generally 20MIPS; mid-range DSP is generally 100MIPS; high-end DSP is generally 1600MIPS.

Numerical operations, in the final analysis, are multiplication and addition operations, that is, ∑An∑×Xn. It can be seen from the above that the internal hardware structure of DSP is more suitable for digital signal processing than MCU.

The external hardware structure of DSP is the same as that of MCU. It consists of three buses: address, data and control. Therefore, the external hardware structure is roughly the same as that of MCU. However, the external bus of DSP is much faster than MCU. Therefore, when selecting external devices, be careful to choose high-speed devices. When making PCB boards, high-speed devices should generally be used. When making PCB boards, multi-layer boards should generally be used, so as to ensure the reliability and stability of the DSP system.

In terms of software development, compared with MCU, DSP better supports modular programming and is more convenient for engineering management. In terms of software/hardware debugging, there is a big difference between DSP and MCU. The MCU software/hardware calls are made in an alternative way. In other words, the MCU simulator is a complete MCU system. The simulation header of the MCU simulator is used to replace the MCU of the simulated target system. The memory on the simulated zero point can even be used to replace the memory of the target system. The disadvantages of this approach are: ① The hardware timing is the simulated hardware timing, which is somewhat different from the target system hardware timing. ②The emulator will occupy more or less certain hardware resources. ③As the number of MCU pins increases and the package becomes smaller, the simulation head becomes more difficult to produce. ④ As the main frequency of MCU increases, the length of the simulation cable will become shorter and shorter, making it more inconvenient to use. ⑤ Different MCU emulators have different hardware, so user development investment is a plus. DSP is simulated using an interface method. There are no DSP resources on the DSP simulator. All resources are on the DSP target system. The DSP simulator only provides a JTAG standard interface (IEEE standard) that is independent of the DSP. There is a dedicated interface on the DSP chip. For the signal pins of simulation debugging, the user only needs to make an interface (14-core double-row pin) on the DSP target board according to the JTAG standard, and connect the two to simulate and debug the DSP. The advantages of DSP simulator compared with MCU simulator are: ①Hardware timing is the target system hardware timing. ②The emulator does not occupy any DSP resources. ③The simulation interface has nothing to do with the number of DSP pins and packaging. ④The simulation interface has nothing to do with the DSP frequency. ⑤The emulator hardware has nothing to do with DSP. The hardware of different series of DSP emulators is the same. The only difference is the compilation software and debugging software, which saves users' development investment. Efficient compilation software and powerful scheduling software make it faster and more convenient for users to develop DSP systems.

Power system DSP solution

According to the characteristics of the power system and the cost-effective requirements of different applications, Beijing Hezhongda Company has launched two DSP hardware platforms for power systems: SEED-F206MS and SEED-C32MS. SEED-F206MS is suitable for power automation (such as distributed RTUs in stations, pole-mounted RTUs and power instruments/instruments) and low-voltage protection; SEED-C32MS is suitable for power high-voltage protection and fault recorders.

The schematic block diagram of SEED-F206MS is shown in Figure 1. It consists of the following parts:

① Processor: TMS320F206 16-bit fixed-point DSP is used as the main processor, processing 20MIPS. The F206 has 4.5×16-bit high-speed SRAM, 32K×16-bit high-speed Flash, a 16-bit timer, a pilot serial port, and a Synchronous serial port and three external interrupts.

②Analog input: 4 pieces of MAX125 are used to form an analog input part with 16 channels of simultaneous sampling, a sampling rate of 76KSPS per channel, a resolution of 14 bits, and an input range of ±5V. The current and voltage are converted into analog input quantities that meet the requirements through CT and PT. The actual analog input channels can reach 32 channels, but these 32 channels are not completely sampled at the same time, but 16 channels/16 channels are sampled at the same time. When the number of sampling channels is reasonably arranged, the same-phase voltage/current of the power system can be satisfied. Simultaneous sampling requirements.

③ Frequency measurement: The hardware circuit realizes the frequency measurement of the analog input signal. The hardware circuit converts the input analog signal into a digital square wave, counts the square wave signal with a 2.5MHz clock, and latches the count value into the frequency measurement register, F206 Read the frequency measurement register, divide 2.5MHz by the frequency measurement register value, which is the frequency of the signal being measured. The measured signal requirements meet: ±1V≤amplitude≤±10V, 39Hz≤frequency≤2.5MHz. The time resolution of frequency measurement accuracy is 400ns.

④Switch input: 16 switching signals are added to the photoelectric isolator after current limiting and debounce, and are buffered to F206 through data, and F206 monitors the switch displacement. The current limiting resistor is 3.6KΩ@1/4W, and the switching input range is: 18V~30V DC. The optoelectronic isolator uses Toshiba TLP121, with an isolation voltage of 2500V DC and a signal bandwidth of 10KHz.

⑤Switching output: 16 switching outputs are latched into the output register by F206, and then output through a Darlington type photoelectric isolator to drive external relays. The Darlington type photoelectric isolator uses Toshiba TLP127. The output stage serves as a switching node. The maximum withstand voltage is 40V DC, the maximum output current is 200mA, the isolation voltage is 2500V DC, and the signal bandwidth is 10KHz.

⑥External interface: Using 16C552 (2 serial/1 parallel) device, plus 1 pilot serial port and 1 synchronous serial port on the 1F206 chip, forming an external interface with 3 asynchronous serial ports, 1 synchronous serial port and 1 printer interface. Facilitates flexible application and expansion of the system. Among the 3 asynchronous serial ports, 2 are equipped with photoelectric isolation, and 1 can be configured as RS232/RS485/RS422, and the remaining 2 are RS232.

⑦ Others: watchdog circuit to improve the reliability of the system; real-time clock to provide a time reference; 2K×8-bit NVRAM that does not lose data during power-off, allowing users to store important parameters; 64K×16-bit external extended program or data memory , providing users with a wider range of applications.

The schematic block diagram of SEED-C32MS is shown in Figure 2, which consists of the following parts:

① Main processor: TMS320C32 32-bit floating point DSP is used as the main processor to process MFLOPS. The C32 chip has 0.5K×32-bit high-speed SRAM, two 32-bit timers, a synchronous serial port and four external interrupts.

②Analog input: Use 1 AD676 and 16 LF398M to form an analog input part with 16 channels of simultaneous sampling and time-sharing conversion, a sampling rate of 5KSPS per channel, a resolution of 16 bits, and an input range of ±5V. The current and voltage are converted into analog input quantities that meet the requirements through CT and PT. Analog input channels can also be expanded to 32 channels through the expansion port.

③ Frequency measurement: The hardware circuit realizes the frequency measurement of the analog input signal. The hardware circuit converts the input analog signal into a digital square wave, counts the square wave signal with a 2.5MHz clock, and stores the count value in the frequency measurement register. 1C32 reads Frequency measurement register, 2.5MHz divided by the frequency measurement register value is the frequency of the signal being measured. The measured signal requirements meet: ±1V≤amplitude≤±10V, 39Hz≤frequency≤2.5MHz. The time resolution of frequency measurement accuracy is 400ns.

④Switch input: 16 switching signals are added to the photoelectric isolator after current limiting and debounce, and are buffered to C32, and C32 monitors the switch displacement. The current limiting resistor is 3.6KΩ@1/4W, and the switching input range is: 18V~30V DC. The optoelectronic isolator uses Toshiba TLP121, with an isolation voltage of 2500V DC and a signal bandwidth of 10KHz.

⑤Switching output: 16 switching outputs are latched into the output register by `C32, and then output through Darlington type photoelectric isolation to drive external relays. The Darlington type photoelectric isolator uses Toshiba TLP127. The output stage serves as a switching node. The maximum withstand voltage is 40V DC, the maximum output current is 200mA, the isolation voltage is 2500V DC, and the signal bandwidth is 10KHz.

⑥External interface: Using 16552 device, it provides 2 asynchronous serial ports, one of which is photoelectrically isolated and configured as RS422/RS485; the other asynchronous serial port is photoelectrically isolated and converts the TTL level to 0~15V.

⑦Expansion interface: There is 1 digital expansion interface and 1 analog expansion interface. The digital expansion interface provides address, data and control buses for users to expand. The analog control interface can also control 16 analog inputs.

⑧Others: Watchdog circuit to improve system reliability; 256K×32-bit external expansion SRAM to provide users with the ability to store programs or data. 512K×32-bit external extended Flash provides users with the ability to store programs or important parameters.